diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,93606 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 19.997756086615055,
+  "eval_steps": 500,
+  "global_step": 66840,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.4561903178691864,
+      "learning_rate": 4.9999999309637996e-05,
+      "loss": 1.7761,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6195425391197205,
+      "learning_rate": 4.999999723855201e-05,
+      "loss": 1.6886,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6293871402740479,
+      "learning_rate": 4.9999993786742175e-05,
+      "loss": 1.6967,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6493249535560608,
+      "learning_rate": 4.999998895420865e-05,
+      "loss": 1.8346,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6501443982124329,
+      "learning_rate": 4.999998274095173e-05,
+      "loss": 1.7955,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5174015760421753,
+      "learning_rate": 4.999997514697176e-05,
+      "loss": 1.8013,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7737800478935242,
+      "learning_rate": 4.999996617226913e-05,
+      "loss": 1.6485,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2001756429672241,
+      "learning_rate": 4.999995581684437e-05,
+      "loss": 1.3865,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1627835035324097,
+      "learning_rate": 4.9999944080698024e-05,
+      "loss": 1.2725,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5961304306983948,
+      "learning_rate": 4.999993096383076e-05,
+      "loss": 1.5426,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7442084550857544,
+      "learning_rate": 4.999991646624329e-05,
+      "loss": 1.5626,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9939375519752502,
+      "learning_rate": 4.9999900587936426e-05,
+      "loss": 1.427,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8643019199371338,
+      "learning_rate": 4.9999883328911034e-05,
+      "loss": 1.8313,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8557299375534058,
+      "learning_rate": 4.999986468916807e-05,
+      "loss": 1.4777,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.45538774132728577,
+      "learning_rate": 4.9999844668708574e-05,
+      "loss": 1.5458,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6336299777030945,
+      "learning_rate": 4.9999823267533627e-05,
+      "loss": 1.3938,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.778620719909668,
+      "learning_rate": 4.9999800485644445e-05,
+      "loss": 1.4802,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8213117718696594,
+      "learning_rate": 4.9999776323042255e-05,
+      "loss": 1.351,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6166107058525085,
+      "learning_rate": 4.999975077972841e-05,
+      "loss": 1.4811,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6074905395507812,
+      "learning_rate": 4.999972385570432e-05,
+      "loss": 1.4381,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7994800209999084,
+      "learning_rate": 4.999969555097146e-05,
+      "loss": 1.4045,
+      "step": 105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5199000239372253,
+      "learning_rate": 4.99996658655314e-05,
+      "loss": 1.479,
+      "step": 110
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5505474209785461,
+      "learning_rate": 4.999963479938577e-05,
+      "loss": 1.5272,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.027380108833313,
+      "learning_rate": 4.999960235253631e-05,
+      "loss": 1.4791,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7694772481918335,
+      "learning_rate": 4.99995685249848e-05,
+      "loss": 1.4485,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9029272794723511,
+      "learning_rate": 4.99995333167331e-05,
+      "loss": 1.3591,
+      "step": 130
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46942275762557983,
+      "learning_rate": 4.999949672778316e-05,
+      "loss": 1.3628,
+      "step": 135
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8350025415420532,
+      "learning_rate": 4.999945875813701e-05,
+      "loss": 1.4336,
+      "step": 140
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6389840841293335,
+      "learning_rate": 4.999941940779673e-05,
+      "loss": 1.4277,
+      "step": 145
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.38392823934555054,
+      "learning_rate": 4.99993786767645e-05,
+      "loss": 1.4192,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.229735255241394,
+      "learning_rate": 4.999933656504257e-05,
+      "loss": 1.514,
+      "step": 155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3870686292648315,
+      "learning_rate": 4.9999293072633273e-05,
+      "loss": 1.4909,
+      "step": 160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.531494617462158,
+      "learning_rate": 4.9999248199539006e-05,
+      "loss": 1.4567,
+      "step": 165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5871274471282959,
+      "learning_rate": 4.9999201945762244e-05,
+      "loss": 1.326,
+      "step": 170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35784173011779785,
+      "learning_rate": 4.999915431130554e-05,
+      "loss": 1.4567,
+      "step": 175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.901470422744751,
+      "learning_rate": 4.9999105296171535e-05,
+      "loss": 1.5553,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3752151727676392,
+      "learning_rate": 4.9999054900362915e-05,
+      "loss": 1.389,
+      "step": 185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1029325723648071,
+      "learning_rate": 4.9999003123882494e-05,
+      "loss": 1.4993,
+      "step": 190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.255204200744629,
+      "learning_rate": 4.999894996673311e-05,
+      "loss": 1.5035,
+      "step": 195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.002902865409851,
+      "learning_rate": 4.9998895428917704e-05,
+      "loss": 1.3108,
+      "step": 200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3146979808807373,
+      "learning_rate": 4.9998839510439286e-05,
+      "loss": 1.5487,
+      "step": 205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5592719316482544,
+      "learning_rate": 4.999878221130095e-05,
+      "loss": 1.3763,
+      "step": 210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5583714246749878,
+      "learning_rate": 4.999872353150586e-05,
+      "loss": 1.4305,
+      "step": 215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0968598127365112,
+      "learning_rate": 4.999866347105725e-05,
+      "loss": 1.4064,
+      "step": 220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7956498265266418,
+      "learning_rate": 4.999860202995844e-05,
+      "loss": 1.3986,
+      "step": 225
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1131305694580078,
+      "learning_rate": 4.999853920821283e-05,
+      "loss": 1.4269,
+      "step": 230
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5098061561584473,
+      "learning_rate": 4.999847500582388e-05,
+      "loss": 1.5158,
+      "step": 235
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8009808659553528,
+      "learning_rate": 4.999840942279514e-05,
+      "loss": 1.4029,
+      "step": 240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5857890248298645,
+      "learning_rate": 4.999834245913023e-05,
+      "loss": 1.4703,
+      "step": 245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9343911409378052,
+      "learning_rate": 4.9998274114832854e-05,
+      "loss": 1.358,
+      "step": 250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.430106520652771,
+      "learning_rate": 4.999820438990678e-05,
+      "loss": 1.3883,
+      "step": 255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5393354296684265,
+      "learning_rate": 4.999813328435586e-05,
+      "loss": 1.4375,
+      "step": 260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7708584666252136,
+      "learning_rate": 4.999806079818403e-05,
+      "loss": 1.4298,
+      "step": 265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8902773261070251,
+      "learning_rate": 4.999798693139528e-05,
+      "loss": 1.4823,
+      "step": 270
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7855858206748962,
+      "learning_rate": 4.99979116839937e-05,
+      "loss": 1.3869,
+      "step": 275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.864590585231781,
+      "learning_rate": 4.9997835055983436e-05,
+      "loss": 1.3566,
+      "step": 280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4891940355300903,
+      "learning_rate": 4.999775704736873e-05,
+      "loss": 1.3024,
+      "step": 285
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1072434186935425,
+      "learning_rate": 4.9997677658153885e-05,
+      "loss": 1.4124,
+      "step": 290
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1894373893737793,
+      "learning_rate": 4.999759688834329e-05,
+      "loss": 1.5361,
+      "step": 295
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6644517183303833,
+      "learning_rate": 4.99975147379414e-05,
+      "loss": 1.4811,
+      "step": 300
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1441906690597534,
+      "learning_rate": 4.999743120695275e-05,
+      "loss": 1.4017,
+      "step": 305
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.748852550983429,
+      "learning_rate": 4.999734629538197e-05,
+      "loss": 1.3883,
+      "step": 310
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8531980514526367,
+      "learning_rate": 4.999726000323373e-05,
+      "loss": 1.3784,
+      "step": 315
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.746377170085907,
+      "learning_rate": 4.99971723305128e-05,
+      "loss": 1.4622,
+      "step": 320
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8359838128089905,
+      "learning_rate": 4.999708327722402e-05,
+      "loss": 1.2369,
+      "step": 325
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8505661487579346,
+      "learning_rate": 4.999699284337232e-05,
+      "loss": 1.6082,
+      "step": 330
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.608825922012329,
+      "learning_rate": 4.99969010289627e-05,
+      "loss": 1.5367,
+      "step": 335
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0551233291625977,
+      "learning_rate": 4.99968078340002e-05,
+      "loss": 1.4654,
+      "step": 340
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0916029214859009,
+      "learning_rate": 4.999671325848999e-05,
+      "loss": 1.585,
+      "step": 345
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6593181490898132,
+      "learning_rate": 4.9996617302437296e-05,
+      "loss": 1.4262,
+      "step": 350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.575879693031311,
+      "learning_rate": 4.999651996584741e-05,
+      "loss": 1.4462,
+      "step": 355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3045742511749268,
+      "learning_rate": 4.999642124872571e-05,
+      "loss": 1.492,
+      "step": 360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6474262475967407,
+      "learning_rate": 4.999632115107764e-05,
+      "loss": 1.3665,
+      "step": 365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9724390506744385,
+      "learning_rate": 4.999621967290874e-05,
+      "loss": 1.2883,
+      "step": 370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.917503297328949,
+      "learning_rate": 4.999611681422461e-05,
+      "loss": 1.5095,
+      "step": 375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5174071788787842,
+      "learning_rate": 4.999601257503093e-05,
+      "loss": 1.5043,
+      "step": 380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2477240562438965,
+      "learning_rate": 4.999590695533345e-05,
+      "loss": 1.4741,
+      "step": 385
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0032185316085815,
+      "learning_rate": 4.9995799955138025e-05,
+      "loss": 1.3719,
+      "step": 390
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0506250858306885,
+      "learning_rate": 4.999569157445054e-05,
+      "loss": 1.4187,
+      "step": 395
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5169631242752075,
+      "learning_rate": 4.999558181327699e-05,
+      "loss": 1.3557,
+      "step": 400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.166700839996338,
+      "learning_rate": 4.9995470671623446e-05,
+      "loss": 1.5298,
+      "step": 405
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9644590616226196,
+      "learning_rate": 4.999535814949603e-05,
+      "loss": 1.4766,
+      "step": 410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.849117636680603,
+      "learning_rate": 4.999524424690097e-05,
+      "loss": 1.3298,
+      "step": 415
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5561323761940002,
+      "learning_rate": 4.9995128963844545e-05,
+      "loss": 1.5718,
+      "step": 420
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6733776330947876,
+      "learning_rate": 4.9995012300333134e-05,
+      "loss": 1.3693,
+      "step": 425
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8910932540893555,
+      "learning_rate": 4.999489425637317e-05,
+      "loss": 1.5194,
+      "step": 430
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7125866413116455,
+      "learning_rate": 4.9994774831971184e-05,
+      "loss": 1.6596,
+      "step": 435
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8103092312812805,
+      "learning_rate": 4.999465402713376e-05,
+      "loss": 1.4965,
+      "step": 440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.902019739151001,
+      "learning_rate": 4.999453184186757e-05,
+      "loss": 1.3179,
+      "step": 445
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9942508935928345,
+      "learning_rate": 4.999440827617938e-05,
+      "loss": 1.3365,
+      "step": 450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8688194751739502,
+      "learning_rate": 4.9994283330075983e-05,
+      "loss": 1.3126,
+      "step": 455
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.632107675075531,
+      "learning_rate": 4.9994157003564314e-05,
+      "loss": 1.4769,
+      "step": 460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.341265082359314,
+      "learning_rate": 4.999402929665133e-05,
+      "loss": 1.3436,
+      "step": 465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.463435173034668,
+      "learning_rate": 4.999390020934408e-05,
+      "loss": 1.3653,
+      "step": 470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5955095291137695,
+      "learning_rate": 4.9993769741649707e-05,
+      "loss": 1.3658,
+      "step": 475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7632045149803162,
+      "learning_rate": 4.999363789357541e-05,
+      "loss": 1.2948,
+      "step": 480
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9410882592201233,
+      "learning_rate": 4.9993504665128474e-05,
+      "loss": 1.4352,
+      "step": 485
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.489148736000061,
+      "learning_rate": 4.999337005631625e-05,
+      "loss": 1.5849,
+      "step": 490
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0961707830429077,
+      "learning_rate": 4.9993234067146186e-05,
+      "loss": 1.4885,
+      "step": 495
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9856806993484497,
+      "learning_rate": 4.999309669762577e-05,
+      "loss": 1.6123,
+      "step": 500
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6104403734207153,
+      "learning_rate": 4.999295794776261e-05,
+      "loss": 1.4599,
+      "step": 505
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5283727645874023,
+      "learning_rate": 4.9992817817564366e-05,
+      "loss": 1.5948,
+      "step": 510
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.664371967315674,
+      "learning_rate": 4.9992676307038765e-05,
+      "loss": 1.5192,
+      "step": 515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.813569962978363,
+      "learning_rate": 4.999253341619363e-05,
+      "loss": 1.3386,
+      "step": 520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8261412382125854,
+      "learning_rate": 4.999238914503686e-05,
+      "loss": 1.3626,
+      "step": 525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0505256652832031,
+      "learning_rate": 4.999224349357641e-05,
+      "loss": 1.3934,
+      "step": 530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3349155187606812,
+      "learning_rate": 4.999209646182033e-05,
+      "loss": 1.3489,
+      "step": 535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8616823554039001,
+      "learning_rate": 4.999194804977674e-05,
+      "loss": 1.4238,
+      "step": 540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6977237462997437,
+      "learning_rate": 4.9991798257453834e-05,
+      "loss": 1.4843,
+      "step": 545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.441572666168213,
+      "learning_rate": 4.9991647084859894e-05,
+      "loss": 1.3559,
+      "step": 550
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1105183362960815,
+      "learning_rate": 4.9991494532003255e-05,
+      "loss": 1.5466,
+      "step": 555
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4936366081237793,
+      "learning_rate": 4.9991340598892355e-05,
+      "loss": 1.256,
+      "step": 560
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0292248725891113,
+      "learning_rate": 4.999118528553569e-05,
+      "loss": 1.4012,
+      "step": 565
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9438468217849731,
+      "learning_rate": 4.9991028591941846e-05,
+      "loss": 1.4391,
+      "step": 570
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2015315294265747,
+      "learning_rate": 4.9990870518119456e-05,
+      "loss": 1.2896,
+      "step": 575
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.021628737449646,
+      "learning_rate": 4.9990711064077276e-05,
+      "loss": 1.4336,
+      "step": 580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3274457454681396,
+      "learning_rate": 4.999055022982409e-05,
+      "loss": 1.3766,
+      "step": 585
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7466582655906677,
+      "learning_rate": 4.99903880153688e-05,
+      "loss": 1.4709,
+      "step": 590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1090950965881348,
+      "learning_rate": 4.999022442072036e-05,
+      "loss": 1.505,
+      "step": 595
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0768952369689941,
+      "learning_rate": 4.999005944588778e-05,
+      "loss": 1.4515,
+      "step": 600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0834194421768188,
+      "learning_rate": 4.998989309088021e-05,
+      "loss": 1.5076,
+      "step": 605
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.119468092918396,
+      "learning_rate": 4.998972535570682e-05,
+      "loss": 1.3812,
+      "step": 610
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5594128370285034,
+      "learning_rate": 4.9989556240376864e-05,
+      "loss": 1.3452,
+      "step": 615
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9768519401550293,
+      "learning_rate": 4.9989385744899705e-05,
+      "loss": 1.5971,
+      "step": 620
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7751976847648621,
+      "learning_rate": 4.9989213869284734e-05,
+      "loss": 1.5723,
+      "step": 625
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8533837795257568,
+      "learning_rate": 4.998904061354146e-05,
+      "loss": 1.3348,
+      "step": 630
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8224301338195801,
+      "learning_rate": 4.9988865977679454e-05,
+      "loss": 1.4717,
+      "step": 635
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4539473056793213,
+      "learning_rate": 4.998868996170835e-05,
+      "loss": 1.4803,
+      "step": 640
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3216567039489746,
+      "learning_rate": 4.9988512565637866e-05,
+      "loss": 1.4922,
+      "step": 645
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8094632029533386,
+      "learning_rate": 4.998833378947781e-05,
+      "loss": 1.521,
+      "step": 650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1426763534545898,
+      "learning_rate": 4.9988153633238065e-05,
+      "loss": 1.2602,
+      "step": 655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7161766886711121,
+      "learning_rate": 4.998797209692856e-05,
+      "loss": 1.3155,
+      "step": 660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38354432582855225,
+      "learning_rate": 4.998778918055933e-05,
+      "loss": 1.4266,
+      "step": 665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6859254837036133,
+      "learning_rate": 4.9987604884140485e-05,
+      "loss": 1.3925,
+      "step": 670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.218139886856079,
+      "learning_rate": 4.9987419207682186e-05,
+      "loss": 1.4437,
+      "step": 675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8014891743659973,
+      "learning_rate": 4.99872321511947e-05,
+      "loss": 1.398,
+      "step": 680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.764227271080017,
+      "learning_rate": 4.998704371468835e-05,
+      "loss": 1.4238,
+      "step": 685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8913390636444092,
+      "learning_rate": 4.998685389817355e-05,
+      "loss": 1.5147,
+      "step": 690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5515832901000977,
+      "learning_rate": 4.998666270166078e-05,
+      "loss": 1.5834,
+      "step": 695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7237977385520935,
+      "learning_rate": 4.998647012516061e-05,
+      "loss": 1.3329,
+      "step": 700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.647252082824707,
+      "learning_rate": 4.998627616868366e-05,
+      "loss": 1.3771,
+      "step": 705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8865868449211121,
+      "learning_rate": 4.9986080832240646e-05,
+      "loss": 1.4632,
+      "step": 710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0670599937438965,
+      "learning_rate": 4.998588411584236e-05,
+      "loss": 1.3726,
+      "step": 715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.742981493473053,
+      "learning_rate": 4.998568601949968e-05,
+      "loss": 1.3809,
+      "step": 720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7563424110412598,
+      "learning_rate": 4.998548654322351e-05,
+      "loss": 1.5463,
+      "step": 725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2141450643539429,
+      "learning_rate": 4.998528568702491e-05,
+      "loss": 1.4768,
+      "step": 730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1518967151641846,
+      "learning_rate": 4.998508345091494e-05,
+      "loss": 1.5653,
+      "step": 735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5695226788520813,
+      "learning_rate": 4.998487983490478e-05,
+      "loss": 1.3969,
+      "step": 740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9854166507720947,
+      "learning_rate": 4.998467483900568e-05,
+      "loss": 1.5679,
+      "step": 745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0108509063720703,
+      "learning_rate": 4.998446846322897e-05,
+      "loss": 1.3145,
+      "step": 750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8769714832305908,
+      "learning_rate": 4.998426070758603e-05,
+      "loss": 1.2757,
+      "step": 755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4470043182373047,
+      "learning_rate": 4.998405157208833e-05,
+      "loss": 1.4435,
+      "step": 760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0755447149276733,
+      "learning_rate": 4.9983841056747446e-05,
+      "loss": 1.3629,
+      "step": 765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8289944529533386,
+      "learning_rate": 4.998362916157498e-05,
+      "loss": 1.4345,
+      "step": 770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8254856467247009,
+      "learning_rate": 4.998341588658265e-05,
+      "loss": 1.4909,
+      "step": 775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4825249910354614,
+      "learning_rate": 4.998320123178223e-05,
+      "loss": 1.3529,
+      "step": 780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.111685037612915,
+      "learning_rate": 4.998298519718557e-05,
+      "loss": 1.292,
+      "step": 785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8751310706138611,
+      "learning_rate": 4.998276778280461e-05,
+      "loss": 1.2957,
+      "step": 790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0307741165161133,
+      "learning_rate": 4.998254898865134e-05,
+      "loss": 1.364,
+      "step": 795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2933695316314697,
+      "learning_rate": 4.998232881473787e-05,
+      "loss": 1.4821,
+      "step": 800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1037958860397339,
+      "learning_rate": 4.998210726107635e-05,
+      "loss": 1.3466,
+      "step": 805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1049686670303345,
+      "learning_rate": 4.9981884327679005e-05,
+      "loss": 1.3488,
+      "step": 810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.055496335029602,
+      "learning_rate": 4.9981660014558165e-05,
+      "loss": 1.4615,
+      "step": 815
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8814008831977844,
+      "learning_rate": 4.99814343217262e-05,
+      "loss": 1.4898,
+      "step": 820
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.712348222732544,
+      "learning_rate": 4.998120724919559e-05,
+      "loss": 1.4881,
+      "step": 825
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7310431003570557,
+      "learning_rate": 4.9980978796978865e-05,
+      "loss": 1.3357,
+      "step": 830
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5869240164756775,
+      "learning_rate": 4.998074896508865e-05,
+      "loss": 1.4089,
+      "step": 835
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.99127197265625,
+      "learning_rate": 4.9980517753537634e-05,
+      "loss": 1.5001,
+      "step": 840
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.083500623703003,
+      "learning_rate": 4.998028516233859e-05,
+      "loss": 1.334,
+      "step": 845
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7491528391838074,
+      "learning_rate": 4.998005119150436e-05,
+      "loss": 1.5227,
+      "step": 850
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5255687832832336,
+      "learning_rate": 4.997981584104788e-05,
+      "loss": 1.3318,
+      "step": 855
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6178213357925415,
+      "learning_rate": 4.997957911098212e-05,
+      "loss": 1.4365,
+      "step": 860
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9565240144729614,
+      "learning_rate": 4.997934100132018e-05,
+      "loss": 1.3325,
+      "step": 865
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6852719187736511,
+      "learning_rate": 4.99791015120752e-05,
+      "loss": 1.4526,
+      "step": 870
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6580244302749634,
+      "learning_rate": 4.99788606432604e-05,
+      "loss": 1.585,
+      "step": 875
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2035952806472778,
+      "learning_rate": 4.9978618394889097e-05,
+      "loss": 1.3965,
+      "step": 880
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3621057271957397,
+      "learning_rate": 4.9978374766974664e-05,
+      "loss": 1.4035,
+      "step": 885
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.326550006866455,
+      "learning_rate": 4.997812975953056e-05,
+      "loss": 1.6006,
+      "step": 890
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5781381726264954,
+      "learning_rate": 4.997788337257031e-05,
+      "loss": 1.376,
+      "step": 895
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.650931715965271,
+      "learning_rate": 4.997763560610752e-05,
+      "loss": 1.4043,
+      "step": 900
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7104997038841248,
+      "learning_rate": 4.997738646015588e-05,
+      "loss": 1.3388,
+      "step": 905
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.88402259349823,
+      "learning_rate": 4.997713593472915e-05,
+      "loss": 1.2928,
+      "step": 910
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.100692868232727,
+      "learning_rate": 4.997688402984116e-05,
+      "loss": 1.3306,
+      "step": 915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7148817777633667,
+      "learning_rate": 4.997663074550584e-05,
+      "loss": 1.2836,
+      "step": 920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2554770708084106,
+      "learning_rate": 4.9976376081737154e-05,
+      "loss": 1.365,
+      "step": 925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8165137767791748,
+      "learning_rate": 4.9976120038549187e-05,
+      "loss": 1.3327,
+      "step": 930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1443816423416138,
+      "learning_rate": 4.997586261595606e-05,
+      "loss": 1.6129,
+      "step": 935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6165381669998169,
+      "learning_rate": 4.997560381397201e-05,
+      "loss": 1.3688,
+      "step": 940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8930396437644958,
+      "learning_rate": 4.997534363261132e-05,
+      "loss": 1.2127,
+      "step": 945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0518968105316162,
+      "learning_rate": 4.997508207188836e-05,
+      "loss": 1.407,
+      "step": 950
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8033773303031921,
+      "learning_rate": 4.997481913181758e-05,
+      "loss": 1.4512,
+      "step": 955
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7851698398590088,
+      "learning_rate": 4.99745548124135e-05,
+      "loss": 1.4645,
+      "step": 960
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4588683843612671,
+      "learning_rate": 4.997428911369072e-05,
+      "loss": 1.3892,
+      "step": 965
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6735519170761108,
+      "learning_rate": 4.99740220356639e-05,
+      "loss": 1.4797,
+      "step": 970
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6732616424560547,
+      "learning_rate": 4.997375357834781e-05,
+      "loss": 1.4322,
+      "step": 975
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9135339856147766,
+      "learning_rate": 4.997348374175727e-05,
+      "loss": 1.489,
+      "step": 980
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7825419902801514,
+      "learning_rate": 4.9973212525907176e-05,
+      "loss": 1.4497,
+      "step": 985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4693922996520996,
+      "learning_rate": 4.997293993081251e-05,
+      "loss": 1.4971,
+      "step": 990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7837210893630981,
+      "learning_rate": 4.997266595648834e-05,
+      "loss": 1.4497,
+      "step": 995
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0610207319259644,
+      "learning_rate": 4.997239060294978e-05,
+      "loss": 1.3558,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9275992512702942,
+      "learning_rate": 4.997211387021204e-05,
+      "loss": 1.3527,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9326786994934082,
+      "learning_rate": 4.997183575829042e-05,
+      "loss": 1.4296,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0663551092147827,
+      "learning_rate": 4.997155626720026e-05,
+      "loss": 1.4676,
+      "step": 1015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5471872091293335,
+      "learning_rate": 4.997127539695701e-05,
+      "loss": 1.2381,
+      "step": 1020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2285401821136475,
+      "learning_rate": 4.997099314757617e-05,
+      "loss": 1.3078,
+      "step": 1025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5964838862419128,
+      "learning_rate": 4.9970709519073334e-05,
+      "loss": 1.4322,
+      "step": 1030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.244211196899414,
+      "learning_rate": 4.997042451146417e-05,
+      "loss": 1.3203,
+      "step": 1035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8410915732383728,
+      "learning_rate": 4.997013812476442e-05,
+      "loss": 1.6386,
+      "step": 1040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.518509030342102,
+      "learning_rate": 4.996985035898989e-05,
+      "loss": 1.4518,
+      "step": 1045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8771889209747314,
+      "learning_rate": 4.996956121415648e-05,
+      "loss": 1.3745,
+      "step": 1050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0236855745315552,
+      "learning_rate": 4.996927069028016e-05,
+      "loss": 1.3879,
+      "step": 1055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9780511856079102,
+      "learning_rate": 4.996897878737697e-05,
+      "loss": 1.3488,
+      "step": 1060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8157287836074829,
+      "learning_rate": 4.9968685505463036e-05,
+      "loss": 1.2988,
+      "step": 1065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7686500549316406,
+      "learning_rate": 4.9968390844554556e-05,
+      "loss": 1.5909,
+      "step": 1070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0402336120605469,
+      "learning_rate": 4.9968094804667805e-05,
+      "loss": 1.3257,
+      "step": 1075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8256892561912537,
+      "learning_rate": 4.9967797385819135e-05,
+      "loss": 1.3697,
+      "step": 1080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6995580792427063,
+      "learning_rate": 4.9967498588024956e-05,
+      "loss": 1.3315,
+      "step": 1085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6209663152694702,
+      "learning_rate": 4.996719841130179e-05,
+      "loss": 1.3986,
+      "step": 1090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9525293707847595,
+      "learning_rate": 4.996689685566621e-05,
+      "loss": 1.5316,
+      "step": 1095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5040790438652039,
+      "learning_rate": 4.9966593921134863e-05,
+      "loss": 1.5408,
+      "step": 1100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3051776885986328,
+      "learning_rate": 4.996628960772449e-05,
+      "loss": 1.4065,
+      "step": 1105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1322470903396606,
+      "learning_rate": 4.996598391545189e-05,
+      "loss": 1.3907,
+      "step": 1110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1070497035980225,
+      "learning_rate": 4.996567684433395e-05,
+      "loss": 1.4401,
+      "step": 1115
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0832529067993164,
+      "learning_rate": 4.996536839438762e-05,
+      "loss": 1.3753,
+      "step": 1120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2609474658966064,
+      "learning_rate": 4.996505856562995e-05,
+      "loss": 1.4671,
+      "step": 1125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5162214040756226,
+      "learning_rate": 4.996474735807805e-05,
+      "loss": 1.3594,
+      "step": 1130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6403785347938538,
+      "learning_rate": 4.9964434771749105e-05,
+      "loss": 1.441,
+      "step": 1135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0005476474761963,
+      "learning_rate": 4.996412080666036e-05,
+      "loss": 1.323,
+      "step": 1140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7727034687995911,
+      "learning_rate": 4.996380546282919e-05,
+      "loss": 1.4661,
+      "step": 1145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1217765808105469,
+      "learning_rate": 4.996348874027298e-05,
+      "loss": 1.4096,
+      "step": 1150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5037316083908081,
+      "learning_rate": 4.9963170639009246e-05,
+      "loss": 1.3398,
+      "step": 1155
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.459764838218689,
+      "learning_rate": 4.996285115905554e-05,
+      "loss": 1.5387,
+      "step": 1160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.856709897518158,
+      "learning_rate": 4.996253030042951e-05,
+      "loss": 1.4499,
+      "step": 1165
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6828152537345886,
+      "learning_rate": 4.996220806314888e-05,
+      "loss": 1.3039,
+      "step": 1170
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1198612451553345,
+      "learning_rate": 4.996188444723144e-05,
+      "loss": 1.356,
+      "step": 1175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6652525067329407,
+      "learning_rate": 4.996155945269507e-05,
+      "loss": 1.3835,
+      "step": 1180
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4609639644622803,
+      "learning_rate": 4.996123307955773e-05,
+      "loss": 1.4632,
+      "step": 1185
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3598580360412598,
+      "learning_rate": 4.996090532783742e-05,
+      "loss": 1.4897,
+      "step": 1190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.24635910987854,
+      "learning_rate": 4.996057619755225e-05,
+      "loss": 1.417,
+      "step": 1195
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0729330778121948,
+      "learning_rate": 4.996024568872042e-05,
+      "loss": 1.3506,
+      "step": 1200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6681000590324402,
+      "learning_rate": 4.9959913801360156e-05,
+      "loss": 1.3517,
+      "step": 1205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8434261679649353,
+      "learning_rate": 4.995958053548979e-05,
+      "loss": 1.6053,
+      "step": 1210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7810892462730408,
+      "learning_rate": 4.9959245891127745e-05,
+      "loss": 1.4025,
+      "step": 1215
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6281015276908875,
+      "learning_rate": 4.995890986829249e-05,
+      "loss": 1.2792,
+      "step": 1220
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.185341477394104,
+      "learning_rate": 4.9958572467002586e-05,
+      "loss": 1.5974,
+      "step": 1225
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9366752505302429,
+      "learning_rate": 4.995823368727667e-05,
+      "loss": 1.3961,
+      "step": 1230
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3230159282684326,
+      "learning_rate": 4.995789352913345e-05,
+      "loss": 1.5296,
+      "step": 1235
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.690661907196045,
+      "learning_rate": 4.995755199259171e-05,
+      "loss": 1.522,
+      "step": 1240
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1200650930404663,
+      "learning_rate": 4.995720907767031e-05,
+      "loss": 1.6149,
+      "step": 1245
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8390517830848694,
+      "learning_rate": 4.9956864784388204e-05,
+      "loss": 1.4568,
+      "step": 1250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49734947085380554,
+      "learning_rate": 4.9956519112764385e-05,
+      "loss": 1.4331,
+      "step": 1255
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3497544527053833,
+      "learning_rate": 4.995617206281797e-05,
+      "loss": 1.4477,
+      "step": 1260
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4694572687149048,
+      "learning_rate": 4.9955823634568105e-05,
+      "loss": 1.3934,
+      "step": 1265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5619333982467651,
+      "learning_rate": 4.9955473828034045e-05,
+      "loss": 1.5304,
+      "step": 1270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7059439420700073,
+      "learning_rate": 4.99551226432351e-05,
+      "loss": 1.548,
+      "step": 1275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1901280879974365,
+      "learning_rate": 4.995477008019067e-05,
+      "loss": 1.5823,
+      "step": 1280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7577235698699951,
+      "learning_rate": 4.9954416138920235e-05,
+      "loss": 1.3385,
+      "step": 1285
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0513741970062256,
+      "learning_rate": 4.995406081944333e-05,
+      "loss": 1.5494,
+      "step": 1290
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9030594229698181,
+      "learning_rate": 4.995370412177959e-05,
+      "loss": 1.5816,
+      "step": 1295
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6190074682235718,
+      "learning_rate": 4.99533460459487e-05,
+      "loss": 1.4034,
+      "step": 1300
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6334771513938904,
+      "learning_rate": 4.995298659197045e-05,
+      "loss": 1.4598,
+      "step": 1305
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5323703289031982,
+      "learning_rate": 4.9952625759864694e-05,
+      "loss": 1.3749,
+      "step": 1310
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3500525951385498,
+      "learning_rate": 4.9952263549651346e-05,
+      "loss": 1.38,
+      "step": 1315
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.885956883430481,
+      "learning_rate": 4.995189996135042e-05,
+      "loss": 1.4463,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.774631917476654,
+      "learning_rate": 4.9951534994981994e-05,
+      "loss": 1.24,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.670936644077301,
+      "learning_rate": 4.9951168650566226e-05,
+      "loss": 1.348,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5310854315757751,
+      "learning_rate": 4.995080092812335e-05,
+      "loss": 1.4768,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7614613771438599,
+      "learning_rate": 4.995043182767368e-05,
+      "loss": 1.4268,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5375897884368896,
+      "learning_rate": 4.995006134923759e-05,
+      "loss": 1.3634,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8461021780967712,
+      "learning_rate": 4.994968949283554e-05,
+      "loss": 1.4866,
+      "step": 1350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7309428453445435,
+      "learning_rate": 4.994931625848808e-05,
+      "loss": 1.5075,
+      "step": 1355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.151624321937561,
+      "learning_rate": 4.994894164621581e-05,
+      "loss": 1.4178,
+      "step": 1360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.812978208065033,
+      "learning_rate": 4.9948565656039434e-05,
+      "loss": 1.2951,
+      "step": 1365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8754581212997437,
+      "learning_rate": 4.99481882879797e-05,
+      "loss": 1.4593,
+      "step": 1370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1998158693313599,
+      "learning_rate": 4.994780954205747e-05,
+      "loss": 1.3958,
+      "step": 1375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9774079918861389,
+      "learning_rate": 4.994742941829364e-05,
+      "loss": 1.3993,
+      "step": 1380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7645953297615051,
+      "learning_rate": 4.9947047916709224e-05,
+      "loss": 1.4456,
+      "step": 1385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7388116717338562,
+      "learning_rate": 4.994666503732528e-05,
+      "loss": 1.4401,
+      "step": 1390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9941197633743286,
+      "learning_rate": 4.994628078016296e-05,
+      "loss": 1.4012,
+      "step": 1395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6217386722564697,
+      "learning_rate": 4.994589514524347e-05,
+      "loss": 1.5009,
+      "step": 1400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7688075304031372,
+      "learning_rate": 4.9945508132588134e-05,
+      "loss": 1.332,
+      "step": 1405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.620936393737793,
+      "learning_rate": 4.994511974221831e-05,
+      "loss": 1.4864,
+      "step": 1410
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8650867938995361,
+      "learning_rate": 4.9944729974155444e-05,
+      "loss": 1.5008,
+      "step": 1415
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.145591139793396,
+      "learning_rate": 4.994433882842108e-05,
+      "loss": 1.242,
+      "step": 1420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6311044692993164,
+      "learning_rate": 4.9943946305036806e-05,
+      "loss": 1.6072,
+      "step": 1425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.777138888835907,
+      "learning_rate": 4.9943552404024303e-05,
+      "loss": 1.4589,
+      "step": 1430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3018752336502075,
+      "learning_rate": 4.994315712540533e-05,
+      "loss": 1.5466,
+      "step": 1435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3640024662017822,
+      "learning_rate": 4.994276046920172e-05,
+      "loss": 1.5863,
+      "step": 1440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0018880367279053,
+      "learning_rate": 4.994236243543537e-05,
+      "loss": 1.3074,
+      "step": 1445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5440133810043335,
+      "learning_rate": 4.994196302412827e-05,
+      "loss": 1.5295,
+      "step": 1450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.098235845565796,
+      "learning_rate": 4.994156223530248e-05,
+      "loss": 1.4089,
+      "step": 1455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.157474398612976,
+      "learning_rate": 4.994116006898013e-05,
+      "loss": 1.4596,
+      "step": 1460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6245862245559692,
+      "learning_rate": 4.994075652518344e-05,
+      "loss": 1.3593,
+      "step": 1465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1995525360107422,
+      "learning_rate": 4.994035160393469e-05,
+      "loss": 1.4744,
+      "step": 1470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7864271402359009,
+      "learning_rate": 4.993994530525624e-05,
+      "loss": 1.3689,
+      "step": 1475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8403075337409973,
+      "learning_rate": 4.9939537629170544e-05,
+      "loss": 1.4856,
+      "step": 1480
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5901763439178467,
+      "learning_rate": 4.99391285757001e-05,
+      "loss": 1.3932,
+      "step": 1485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9339320063591003,
+      "learning_rate": 4.993871814486751e-05,
+      "loss": 1.5516,
+      "step": 1490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3036901950836182,
+      "learning_rate": 4.993830633669544e-05,
+      "loss": 1.4868,
+      "step": 1495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6879803538322449,
+      "learning_rate": 4.9937893151206626e-05,
+      "loss": 1.38,
+      "step": 1500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7101956009864807,
+      "learning_rate": 4.99374785884239e-05,
+      "loss": 1.3969,
+      "step": 1505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.220004916191101,
+      "learning_rate": 4.9937062648370154e-05,
+      "loss": 1.4985,
+      "step": 1510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5632555484771729,
+      "learning_rate": 4.993664533106835e-05,
+      "loss": 1.3545,
+      "step": 1515
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4013103246688843,
+      "learning_rate": 4.993622663654156e-05,
+      "loss": 1.4475,
+      "step": 1520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9690436124801636,
+      "learning_rate": 4.993580656481288e-05,
+      "loss": 1.3327,
+      "step": 1525
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6049666404724121,
+      "learning_rate": 4.993538511590553e-05,
+      "loss": 1.1947,
+      "step": 1530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7417116165161133,
+      "learning_rate": 4.993496228984278e-05,
+      "loss": 1.5365,
+      "step": 1535
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1022894382476807,
+      "learning_rate": 4.993453808664797e-05,
+      "loss": 1.4447,
+      "step": 1540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.834717333316803,
+      "learning_rate": 4.993411250634455e-05,
+      "loss": 1.4464,
+      "step": 1545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6041282415390015,
+      "learning_rate": 4.9933685548956014e-05,
+      "loss": 1.388,
+      "step": 1550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8697533011436462,
+      "learning_rate": 4.993325721450594e-05,
+      "loss": 1.425,
+      "step": 1555
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2372421026229858,
+      "learning_rate": 4.993282750301799e-05,
+      "loss": 1.3586,
+      "step": 1560
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.791959285736084,
+      "learning_rate": 4.993239641451588e-05,
+      "loss": 1.7594,
+      "step": 1565
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.356498956680298,
+      "learning_rate": 4.993196394902344e-05,
+      "loss": 1.4429,
+      "step": 1570
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8572433590888977,
+      "learning_rate": 4.993153010656455e-05,
+      "loss": 1.3813,
+      "step": 1575
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.199879765510559,
+      "learning_rate": 4.993109488716316e-05,
+      "loss": 1.2936,
+      "step": 1580
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6142605543136597,
+      "learning_rate": 4.993065829084332e-05,
+      "loss": 1.3402,
+      "step": 1585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4303763806819916,
+      "learning_rate": 4.993022031762914e-05,
+      "loss": 1.3431,
+      "step": 1590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7792863249778748,
+      "learning_rate": 4.99297809675448e-05,
+      "loss": 1.5333,
+      "step": 1595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.015516996383667,
+      "learning_rate": 4.992934024061456e-05,
+      "loss": 1.5036,
+      "step": 1600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5127567648887634,
+      "learning_rate": 4.992889813686279e-05,
+      "loss": 1.3214,
+      "step": 1605
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4127864837646484,
+      "learning_rate": 4.992845465631388e-05,
+      "loss": 1.4556,
+      "step": 1610
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0938849449157715,
+      "learning_rate": 4.992800979899233e-05,
+      "loss": 1.372,
+      "step": 1615
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7442615032196045,
+      "learning_rate": 4.992756356492271e-05,
+      "loss": 1.293,
+      "step": 1620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.879451334476471,
+      "learning_rate": 4.9927115954129665e-05,
+      "loss": 1.4263,
+      "step": 1625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.867012619972229,
+      "learning_rate": 4.9926666966637914e-05,
+      "loss": 1.3219,
+      "step": 1630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4864187240600586,
+      "learning_rate": 4.992621660247226e-05,
+      "loss": 1.5152,
+      "step": 1635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.02019202709198,
+      "learning_rate": 4.992576486165758e-05,
+      "loss": 1.3683,
+      "step": 1640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8681535720825195,
+      "learning_rate": 4.99253117442188e-05,
+      "loss": 1.4345,
+      "step": 1645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6605016589164734,
+      "learning_rate": 4.992485725018097e-05,
+      "loss": 1.3767,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7456071376800537,
+      "learning_rate": 4.9924401379569174e-05,
+      "loss": 1.17,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.299971342086792,
+      "learning_rate": 4.992394413240861e-05,
+      "loss": 1.545,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6303769946098328,
+      "learning_rate": 4.992348550872451e-05,
+      "loss": 1.3899,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5625646710395813,
+      "learning_rate": 4.992302550854221e-05,
+      "loss": 1.3778,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0175740718841553,
+      "learning_rate": 4.992256413188712e-05,
+      "loss": 1.3702,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6424248814582825,
+      "learning_rate": 4.992210137878472e-05,
+      "loss": 1.3868,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0193471908569336,
+      "learning_rate": 4.992163724926057e-05,
+      "loss": 1.3882,
+      "step": 1685
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6782774925231934,
+      "learning_rate": 4.992117174334029e-05,
+      "loss": 1.4042,
+      "step": 1690
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7803168892860413,
+      "learning_rate": 4.99207048610496e-05,
+      "loss": 1.2801,
+      "step": 1695
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7628136873245239,
+      "learning_rate": 4.992023660241429e-05,
+      "loss": 1.4438,
+      "step": 1700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7955535054206848,
+      "learning_rate": 4.991976696746021e-05,
+      "loss": 1.2711,
+      "step": 1705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0430067777633667,
+      "learning_rate": 4.991929595621331e-05,
+      "loss": 1.3843,
+      "step": 1710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0455753803253174,
+      "learning_rate": 4.991882356869959e-05,
+      "loss": 1.3456,
+      "step": 1715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5499788522720337,
+      "learning_rate": 4.9918349804945154e-05,
+      "loss": 1.4326,
+      "step": 1720
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6572498083114624,
+      "learning_rate": 4.991787466497615e-05,
+      "loss": 1.5624,
+      "step": 1725
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2623273134231567,
+      "learning_rate": 4.9917398148818836e-05,
+      "loss": 1.5035,
+      "step": 1730
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5261033773422241,
+      "learning_rate": 4.991692025649952e-05,
+      "loss": 1.492,
+      "step": 1735
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8616695404052734,
+      "learning_rate": 4.991644098804459e-05,
+      "loss": 1.3531,
+      "step": 1740
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3723504543304443,
+      "learning_rate": 4.991596034348053e-05,
+      "loss": 1.5159,
+      "step": 1745
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5867530703544617,
+      "learning_rate": 4.991547832283389e-05,
+      "loss": 1.5382,
+      "step": 1750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0854787826538086,
+      "learning_rate": 4.9914994926131265e-05,
+      "loss": 1.4862,
+      "step": 1755
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.023762822151184,
+      "learning_rate": 4.991451015339937e-05,
+      "loss": 1.4294,
+      "step": 1760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.219565749168396,
+      "learning_rate": 4.9914024004664986e-05,
+      "loss": 1.3301,
+      "step": 1765
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8434814214706421,
+      "learning_rate": 4.991353647995494e-05,
+      "loss": 1.5169,
+      "step": 1770
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8482433557510376,
+      "learning_rate": 4.9913047579296177e-05,
+      "loss": 1.4846,
+      "step": 1775
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.639321506023407,
+      "learning_rate": 4.991255730271569e-05,
+      "loss": 1.3135,
+      "step": 1780
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1192939281463623,
+      "learning_rate": 4.991206565024056e-05,
+      "loss": 1.3945,
+      "step": 1785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9012823700904846,
+      "learning_rate": 4.991157262189794e-05,
+      "loss": 1.2153,
+      "step": 1790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0617835521697998,
+      "learning_rate": 4.991107821771506e-05,
+      "loss": 1.3468,
+      "step": 1795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.25299870967865,
+      "learning_rate": 4.991058243771922e-05,
+      "loss": 1.436,
+      "step": 1800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.779280960559845,
+      "learning_rate": 4.9910085281937804e-05,
+      "loss": 1.3551,
+      "step": 1805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6145398020744324,
+      "learning_rate": 4.9909586750398274e-05,
+      "loss": 1.2627,
+      "step": 1810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7118735313415527,
+      "learning_rate": 4.9909086843128154e-05,
+      "loss": 1.5455,
+      "step": 1815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2611353397369385,
+      "learning_rate": 4.990858556015506e-05,
+      "loss": 1.4429,
+      "step": 1820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1818840503692627,
+      "learning_rate": 4.990808290150668e-05,
+      "loss": 1.586,
+      "step": 1825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6355205774307251,
+      "learning_rate": 4.990757886721077e-05,
+      "loss": 1.3538,
+      "step": 1830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7018294334411621,
+      "learning_rate": 4.990707345729517e-05,
+      "loss": 1.5147,
+      "step": 1835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5799555778503418,
+      "learning_rate": 4.99065666717878e-05,
+      "loss": 1.4748,
+      "step": 1840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1363238096237183,
+      "learning_rate": 4.990605851071664e-05,
+      "loss": 1.3812,
+      "step": 1845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8724756836891174,
+      "learning_rate": 4.9905548974109746e-05,
+      "loss": 1.4634,
+      "step": 1850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3491142988204956,
+      "learning_rate": 4.990503806199527e-05,
+      "loss": 1.4406,
+      "step": 1855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8592522740364075,
+      "learning_rate": 4.990452577440144e-05,
+      "loss": 1.55,
+      "step": 1860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8113953471183777,
+      "learning_rate": 4.9904012111356536e-05,
+      "loss": 1.5138,
+      "step": 1865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8070884943008423,
+      "learning_rate": 4.990349707288892e-05,
+      "loss": 1.5115,
+      "step": 1870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.018451452255249,
+      "learning_rate": 4.990298065902706e-05,
+      "loss": 1.4229,
+      "step": 1875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5133267641067505,
+      "learning_rate": 4.9902462869799446e-05,
+      "loss": 1.1838,
+      "step": 1880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.795879065990448,
+      "learning_rate": 4.990194370523471e-05,
+      "loss": 1.5473,
+      "step": 1885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8955612182617188,
+      "learning_rate": 4.990142316536149e-05,
+      "loss": 1.4088,
+      "step": 1890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3825359344482422,
+      "learning_rate": 4.990090125020857e-05,
+      "loss": 1.3161,
+      "step": 1895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9346737265586853,
+      "learning_rate": 4.990037795980474e-05,
+      "loss": 1.3944,
+      "step": 1900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5142995119094849,
+      "learning_rate": 4.989985329417893e-05,
+      "loss": 1.3262,
+      "step": 1905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0340321063995361,
+      "learning_rate": 4.98993272533601e-05,
+      "loss": 1.4552,
+      "step": 1910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5901709794998169,
+      "learning_rate": 4.989879983737732e-05,
+      "loss": 1.2281,
+      "step": 1915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0993763208389282,
+      "learning_rate": 4.989827104625969e-05,
+      "loss": 1.4189,
+      "step": 1920
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9149566292762756,
+      "learning_rate": 4.989774088003644e-05,
+      "loss": 1.5347,
+      "step": 1925
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6457690596580505,
+      "learning_rate": 4.989720933873683e-05,
+      "loss": 1.2577,
+      "step": 1930
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5968732237815857,
+      "learning_rate": 4.989667642239023e-05,
+      "loss": 1.3495,
+      "step": 1935
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9515772461891174,
+      "learning_rate": 4.989614213102608e-05,
+      "loss": 1.4536,
+      "step": 1940
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9571760892868042,
+      "learning_rate": 4.989560646467387e-05,
+      "loss": 1.4527,
+      "step": 1945
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5556336045265198,
+      "learning_rate": 4.989506942336319e-05,
+      "loss": 1.414,
+      "step": 1950
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.146568775177002,
+      "learning_rate": 4.989453100712371e-05,
+      "loss": 1.3903,
+      "step": 1955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.47517749667167664,
+      "learning_rate": 4.989399121598515e-05,
+      "loss": 1.2966,
+      "step": 1960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3402255773544312,
+      "learning_rate": 4.989345004997734e-05,
+      "loss": 1.4773,
+      "step": 1965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.027206540107727,
+      "learning_rate": 4.9892907509130156e-05,
+      "loss": 1.5167,
+      "step": 1970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7391974925994873,
+      "learning_rate": 4.989236359347356e-05,
+      "loss": 1.492,
+      "step": 1975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5391299724578857,
+      "learning_rate": 4.989181830303761e-05,
+      "loss": 1.3457,
+      "step": 1980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5835283398628235,
+      "learning_rate": 4.9891271637852396e-05,
+      "loss": 1.3803,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2071964740753174,
+      "learning_rate": 4.9890723597948126e-05,
+      "loss": 1.4644,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8295852541923523,
+      "learning_rate": 4.989017418335507e-05,
+      "loss": 1.3478,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2899019718170166,
+      "learning_rate": 4.988962339410356e-05,
+      "loss": 1.5883,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8376094698905945,
+      "learning_rate": 4.988907123022401e-05,
+      "loss": 1.6507,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5938405990600586,
+      "learning_rate": 4.988851769174695e-05,
+      "loss": 1.5191,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7297775149345398,
+      "learning_rate": 4.988796277870291e-05,
+      "loss": 1.4755,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8325863480567932,
+      "learning_rate": 4.988740649112256e-05,
+      "loss": 1.455,
+      "step": 2020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6817718148231506,
+      "learning_rate": 4.9886848829036624e-05,
+      "loss": 1.3709,
+      "step": 2025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8563414812088013,
+      "learning_rate": 4.9886289792475894e-05,
+      "loss": 1.3933,
+      "step": 2030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8419439792633057,
+      "learning_rate": 4.9885729381471244e-05,
+      "loss": 1.2793,
+      "step": 2035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1437691450119019,
+      "learning_rate": 4.988516759605363e-05,
+      "loss": 1.4561,
+      "step": 2040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0974056720733643,
+      "learning_rate": 4.9884604436254065e-05,
+      "loss": 1.3317,
+      "step": 2045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7327514886856079,
+      "learning_rate": 4.9884039902103674e-05,
+      "loss": 1.5174,
+      "step": 2050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5889623165130615,
+      "learning_rate": 4.9883473993633626e-05,
+      "loss": 1.4063,
+      "step": 2055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.461192011833191,
+      "learning_rate": 4.988290671087517e-05,
+      "loss": 1.4681,
+      "step": 2060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7430127859115601,
+      "learning_rate": 4.9882338053859646e-05,
+      "loss": 1.3817,
+      "step": 2065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4986989498138428,
+      "learning_rate": 4.988176802261845e-05,
+      "loss": 1.3504,
+      "step": 2070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7772125601768494,
+      "learning_rate": 4.988119661718307e-05,
+      "loss": 1.36,
+      "step": 2075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2904176712036133,
+      "learning_rate": 4.988062383758506e-05,
+      "loss": 1.4057,
+      "step": 2080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2020084857940674,
+      "learning_rate": 4.9880049683856066e-05,
+      "loss": 1.5082,
+      "step": 2085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7529141306877136,
+      "learning_rate": 4.987947415602778e-05,
+      "loss": 1.3725,
+      "step": 2090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8067646622657776,
+      "learning_rate": 4.987889725413201e-05,
+      "loss": 1.3523,
+      "step": 2095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7970091104507446,
+      "learning_rate": 4.987831897820059e-05,
+      "loss": 1.6039,
+      "step": 2100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6938903331756592,
+      "learning_rate": 4.987773932826548e-05,
+      "loss": 1.4473,
+      "step": 2105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8393217921257019,
+      "learning_rate": 4.9877158304358687e-05,
+      "loss": 1.4723,
+      "step": 2110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2300384044647217,
+      "learning_rate": 4.987657590651229e-05,
+      "loss": 1.2791,
+      "step": 2115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7245568633079529,
+      "learning_rate": 4.9875992134758476e-05,
+      "loss": 1.4415,
+      "step": 2120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7715205550193787,
+      "learning_rate": 4.987540698912947e-05,
+      "loss": 1.4555,
+      "step": 2125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9755416512489319,
+      "learning_rate": 4.987482046965759e-05,
+      "loss": 1.4118,
+      "step": 2130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0845292806625366,
+      "learning_rate": 4.987423257637523e-05,
+      "loss": 1.4648,
+      "step": 2135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7946663498878479,
+      "learning_rate": 4.9873643309314864e-05,
+      "loss": 1.4335,
+      "step": 2140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.767040491104126,
+      "learning_rate": 4.987305266850903e-05,
+      "loss": 1.3784,
+      "step": 2145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0675404071807861,
+      "learning_rate": 4.987246065399035e-05,
+      "loss": 1.573,
+      "step": 2150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.144506573677063,
+      "learning_rate": 4.987186726579153e-05,
+      "loss": 1.3506,
+      "step": 2155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6684775948524475,
+      "learning_rate": 4.987127250394532e-05,
+      "loss": 1.2895,
+      "step": 2160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9211767911911011,
+      "learning_rate": 4.987067636848459e-05,
+      "loss": 1.2889,
+      "step": 2165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1036546230316162,
+      "learning_rate": 4.987007885944226e-05,
+      "loss": 1.39,
+      "step": 2170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8043128252029419,
+      "learning_rate": 4.986947997685132e-05,
+      "loss": 1.4297,
+      "step": 2175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3548777103424072,
+      "learning_rate": 4.986887972074485e-05,
+      "loss": 1.4425,
+      "step": 2180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4966070652008057,
+      "learning_rate": 4.9868278091156e-05,
+      "loss": 1.4042,
+      "step": 2185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8505388498306274,
+      "learning_rate": 4.986767508811801e-05,
+      "loss": 1.4276,
+      "step": 2190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0283442735671997,
+      "learning_rate": 4.986707071166417e-05,
+      "loss": 1.4598,
+      "step": 2195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.680404007434845,
+      "learning_rate": 4.9866464961827856e-05,
+      "loss": 1.4773,
+      "step": 2200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.575295090675354,
+      "learning_rate": 4.986585783864254e-05,
+      "loss": 1.3162,
+      "step": 2205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7415833473205566,
+      "learning_rate": 4.9865249342141726e-05,
+      "loss": 1.4289,
+      "step": 2210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8124527335166931,
+      "learning_rate": 4.986463947235905e-05,
+      "loss": 1.4606,
+      "step": 2215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7351284623146057,
+      "learning_rate": 4.9864028229328186e-05,
+      "loss": 1.5149,
+      "step": 2220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9113266468048096,
+      "learning_rate": 4.9863415613082876e-05,
+      "loss": 1.5249,
+      "step": 2225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.276861310005188,
+      "learning_rate": 4.986280162365697e-05,
+      "loss": 1.4372,
+      "step": 2230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6544708013534546,
+      "learning_rate": 4.9862186261084374e-05,
+      "loss": 1.3714,
+      "step": 2235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9754915833473206,
+      "learning_rate": 4.986156952539908e-05,
+      "loss": 1.3048,
+      "step": 2240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7614801526069641,
+      "learning_rate": 4.9860951416635126e-05,
+      "loss": 1.4061,
+      "step": 2245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1788151264190674,
+      "learning_rate": 4.986033193482668e-05,
+      "loss": 1.2955,
+      "step": 2250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.917193591594696,
+      "learning_rate": 4.9859711080007944e-05,
+      "loss": 1.351,
+      "step": 2255
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7633816003799438,
+      "learning_rate": 4.9859088852213196e-05,
+      "loss": 1.3889,
+      "step": 2260
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4886833429336548,
+      "learning_rate": 4.985846525147681e-05,
+      "loss": 1.4501,
+      "step": 2265
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9673319458961487,
+      "learning_rate": 4.9857840277833236e-05,
+      "loss": 1.4624,
+      "step": 2270
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7718215584754944,
+      "learning_rate": 4.9857213931316984e-05,
+      "loss": 1.4116,
+      "step": 2275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6662101745605469,
+      "learning_rate": 4.985658621196263e-05,
+      "loss": 1.425,
+      "step": 2280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7783111333847046,
+      "learning_rate": 4.985595711980486e-05,
+      "loss": 1.5157,
+      "step": 2285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4051775634288788,
+      "learning_rate": 4.985532665487843e-05,
+      "loss": 1.3326,
+      "step": 2290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8741587996482849,
+      "learning_rate": 4.9854694817218125e-05,
+      "loss": 1.3785,
+      "step": 2295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6999533772468567,
+      "learning_rate": 4.985406160685887e-05,
+      "loss": 1.3553,
+      "step": 2300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7438477277755737,
+      "learning_rate": 4.985342702383563e-05,
+      "loss": 1.4612,
+      "step": 2305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4302748441696167,
+      "learning_rate": 4.985279106818345e-05,
+      "loss": 1.4045,
+      "step": 2310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8926997780799866,
+      "learning_rate": 4.9852153739937444e-05,
+      "loss": 1.448,
+      "step": 2315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7506667375564575,
+      "learning_rate": 4.9851515039132824e-05,
+      "loss": 1.3386,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7566189169883728,
+      "learning_rate": 4.985087496580485e-05,
+      "loss": 1.5076,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2634682655334473,
+      "learning_rate": 4.98502335199889e-05,
+      "loss": 1.5423,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0470796823501587,
+      "learning_rate": 4.984959070172037e-05,
+      "loss": 1.3496,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8911975026130676,
+      "learning_rate": 4.984894651103478e-05,
+      "loss": 1.4615,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8113840222358704,
+      "learning_rate": 4.98483009479677e-05,
+      "loss": 1.4067,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6011930704116821,
+      "learning_rate": 4.984765401255479e-05,
+      "loss": 1.3069,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4766760468482971,
+      "learning_rate": 4.984700570483178e-05,
+      "loss": 1.552,
+      "step": 2355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0068391561508179,
+      "learning_rate": 4.984635602483447e-05,
+      "loss": 1.3886,
+      "step": 2360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.179116129875183,
+      "learning_rate": 4.984570497259874e-05,
+      "loss": 1.4761,
+      "step": 2365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.623014509677887,
+      "learning_rate": 4.9845052548160554e-05,
+      "loss": 1.627,
+      "step": 2370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6420192718505859,
+      "learning_rate": 4.984439875155593e-05,
+      "loss": 1.2861,
+      "step": 2375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6435323357582092,
+      "learning_rate": 4.9843743582821e-05,
+      "loss": 1.3463,
+      "step": 2380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8503981828689575,
+      "learning_rate": 4.984308704199193e-05,
+      "loss": 1.4085,
+      "step": 2385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8631937503814697,
+      "learning_rate": 4.984242912910499e-05,
+      "loss": 1.3638,
+      "step": 2390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8973201513290405,
+      "learning_rate": 4.984176984419651e-05,
+      "loss": 1.3254,
+      "step": 2395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8003751635551453,
+      "learning_rate": 4.9841109187302896e-05,
+      "loss": 1.3815,
+      "step": 2400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4101213216781616,
+      "learning_rate": 4.984044715846065e-05,
+      "loss": 1.5289,
+      "step": 2405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5882824063301086,
+      "learning_rate": 4.983978375770633e-05,
+      "loss": 1.369,
+      "step": 2410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5322262644767761,
+      "learning_rate": 4.983911898507656e-05,
+      "loss": 1.6157,
+      "step": 2415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2763606309890747,
+      "learning_rate": 4.9838452840608076e-05,
+      "loss": 1.4812,
+      "step": 2420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9247528910636902,
+      "learning_rate": 4.983778532433766e-05,
+      "loss": 1.4012,
+      "step": 2425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0083926916122437,
+      "learning_rate": 4.983711643630218e-05,
+      "loss": 1.4391,
+      "step": 2430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7945544719696045,
+      "learning_rate": 4.983644617653857e-05,
+      "loss": 1.3384,
+      "step": 2435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8284363150596619,
+      "learning_rate": 4.983577454508386e-05,
+      "loss": 1.497,
+      "step": 2440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9625481367111206,
+      "learning_rate": 4.9835101541975125e-05,
+      "loss": 1.5237,
+      "step": 2445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6448972821235657,
+      "learning_rate": 4.983442716724956e-05,
+      "loss": 1.3189,
+      "step": 2450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6880027651786804,
+      "learning_rate": 4.983375142094439e-05,
+      "loss": 1.4792,
+      "step": 2455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7837314605712891,
+      "learning_rate": 4.983307430309695e-05,
+      "loss": 1.4544,
+      "step": 2460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7839219570159912,
+      "learning_rate": 4.9832395813744614e-05,
+      "loss": 1.4142,
+      "step": 2465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2939424514770508,
+      "learning_rate": 4.983171595292489e-05,
+      "loss": 1.3861,
+      "step": 2470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.47123095393180847,
+      "learning_rate": 4.983103472067529e-05,
+      "loss": 1.4055,
+      "step": 2475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9300950169563293,
+      "learning_rate": 4.983035211703345e-05,
+      "loss": 1.3981,
+      "step": 2480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5792704820632935,
+      "learning_rate": 4.982966814203708e-05,
+      "loss": 1.2605,
+      "step": 2485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7978287935256958,
+      "learning_rate": 4.9828982795723944e-05,
+      "loss": 1.573,
+      "step": 2490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7604524493217468,
+      "learning_rate": 4.98282960781319e-05,
+      "loss": 1.4199,
+      "step": 2495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9219200015068054,
+      "learning_rate": 4.982760798929887e-05,
+      "loss": 1.3895,
+      "step": 2500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7405526638031006,
+      "learning_rate": 4.982691852926286e-05,
+      "loss": 1.3229,
+      "step": 2505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8719181418418884,
+      "learning_rate": 4.982622769806193e-05,
+      "loss": 1.3892,
+      "step": 2510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5789086818695068,
+      "learning_rate": 4.982553549573427e-05,
+      "loss": 1.3667,
+      "step": 2515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6426258683204651,
+      "learning_rate": 4.982484192231808e-05,
+      "loss": 1.463,
+      "step": 2520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7936124205589294,
+      "learning_rate": 4.982414697785168e-05,
+      "loss": 1.4307,
+      "step": 2525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7490212917327881,
+      "learning_rate": 4.982345066237344e-05,
+      "loss": 1.3233,
+      "step": 2530
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6938835978507996,
+      "learning_rate": 4.9822752975921826e-05,
+      "loss": 1.4814,
+      "step": 2535
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6656054854393005,
+      "learning_rate": 4.982205391853536e-05,
+      "loss": 1.3243,
+      "step": 2540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8422638773918152,
+      "learning_rate": 4.982135349025266e-05,
+      "loss": 1.4865,
+      "step": 2545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7025824785232544,
+      "learning_rate": 4.982065169111241e-05,
+      "loss": 1.3322,
+      "step": 2550
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2645024061203003,
+      "learning_rate": 4.981994852115337e-05,
+      "loss": 1.405,
+      "step": 2555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8437483310699463,
+      "learning_rate": 4.981924398041437e-05,
+      "loss": 1.49,
+      "step": 2560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1599018573760986,
+      "learning_rate": 4.9818538068934314e-05,
+      "loss": 1.3704,
+      "step": 2565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4388209581375122,
+      "learning_rate": 4.981783078675221e-05,
+      "loss": 1.3873,
+      "step": 2570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.019699215888977,
+      "learning_rate": 4.98171221339071e-05,
+      "loss": 1.2774,
+      "step": 2575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7524467706680298,
+      "learning_rate": 4.981641211043814e-05,
+      "loss": 1.3507,
+      "step": 2580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7237170934677124,
+      "learning_rate": 4.981570071638453e-05,
+      "loss": 1.4106,
+      "step": 2585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.224061965942383,
+      "learning_rate": 4.981498795178556e-05,
+      "loss": 1.4087,
+      "step": 2590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.733247697353363,
+      "learning_rate": 4.98142738166806e-05,
+      "loss": 1.4756,
+      "step": 2595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9564744830131531,
+      "learning_rate": 4.9813558311109095e-05,
+      "loss": 1.4537,
+      "step": 2600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5492211580276489,
+      "learning_rate": 4.981284143511055e-05,
+      "loss": 1.1261,
+      "step": 2605
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8463557362556458,
+      "learning_rate": 4.981212318872457e-05,
+      "loss": 1.4015,
+      "step": 2610
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0302629470825195,
+      "learning_rate": 4.981140357199081e-05,
+      "loss": 1.4069,
+      "step": 2615
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7704905867576599,
+      "learning_rate": 4.981068258494903e-05,
+      "loss": 1.3306,
+      "step": 2620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.499060034751892,
+      "learning_rate": 4.980996022763904e-05,
+      "loss": 1.5796,
+      "step": 2625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.660788357257843,
+      "learning_rate": 4.980923650010072e-05,
+      "loss": 1.401,
+      "step": 2630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.033163547515869,
+      "learning_rate": 4.980851140237407e-05,
+      "loss": 1.442,
+      "step": 2635
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5330935716629028,
+      "learning_rate": 4.9807784934499125e-05,
+      "loss": 1.5339,
+      "step": 2640
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3255956172943115,
+      "learning_rate": 4.9807057096516e-05,
+      "loss": 1.5884,
+      "step": 2645
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7848919630050659,
+      "learning_rate": 4.9806327888464885e-05,
+      "loss": 1.4167,
+      "step": 2650
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9711475968360901,
+      "learning_rate": 4.980559731038608e-05,
+      "loss": 1.4576,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9157351851463318,
+      "learning_rate": 4.9804865362319914e-05,
+      "loss": 1.3899,
+      "step": 2660
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0029135942459106,
+      "learning_rate": 4.980413204430682e-05,
+      "loss": 1.4998,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5365571975708008,
+      "learning_rate": 4.980339735638729e-05,
+      "loss": 1.3776,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6640515327453613,
+      "learning_rate": 4.980266129860191e-05,
+      "loss": 1.3586,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6773358583450317,
+      "learning_rate": 4.9801923870991326e-05,
+      "loss": 1.3586,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.62249755859375,
+      "learning_rate": 4.9801185073596266e-05,
+      "loss": 1.5044,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2861700057983398,
+      "learning_rate": 4.980044490645754e-05,
+      "loss": 1.5279,
+      "step": 2690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6400649547576904,
+      "learning_rate": 4.979970336961601e-05,
+      "loss": 1.4849,
+      "step": 2695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2481294870376587,
+      "learning_rate": 4.9798960463112654e-05,
+      "loss": 1.3853,
+      "step": 2700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.038743019104004,
+      "learning_rate": 4.979821618698848e-05,
+      "loss": 1.3042,
+      "step": 2705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1778807640075684,
+      "learning_rate": 4.979747054128461e-05,
+      "loss": 1.4474,
+      "step": 2710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8384398818016052,
+      "learning_rate": 4.97967235260422e-05,
+      "loss": 1.3712,
+      "step": 2715
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2256791591644287,
+      "learning_rate": 4.9795975141302545e-05,
+      "loss": 1.6502,
+      "step": 2720
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5445688962936401,
+      "learning_rate": 4.979522538710695e-05,
+      "loss": 1.414,
+      "step": 2725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1054037809371948,
+      "learning_rate": 4.979447426349682e-05,
+      "loss": 1.3975,
+      "step": 2730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6710256934165955,
+      "learning_rate": 4.979372177051366e-05,
+      "loss": 1.3354,
+      "step": 2735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0434719324111938,
+      "learning_rate": 4.979296790819901e-05,
+      "loss": 1.4389,
+      "step": 2740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8868400454521179,
+      "learning_rate": 4.9792212676594516e-05,
+      "loss": 1.3882,
+      "step": 2745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7899145483970642,
+      "learning_rate": 4.9791456075741895e-05,
+      "loss": 1.2246,
+      "step": 2750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6027569770812988,
+      "learning_rate": 4.979069810568292e-05,
+      "loss": 1.4354,
+      "step": 2755
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1097197532653809,
+      "learning_rate": 4.9789938766459446e-05,
+      "loss": 1.4536,
+      "step": 2760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8584468960762024,
+      "learning_rate": 4.9789178058113434e-05,
+      "loss": 1.3242,
+      "step": 2765
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6569173336029053,
+      "learning_rate": 4.978841598068688e-05,
+      "loss": 1.4388,
+      "step": 2770
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6124112010002136,
+      "learning_rate": 4.978765253422188e-05,
+      "loss": 1.3181,
+      "step": 2775
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7764132022857666,
+      "learning_rate": 4.978688771876059e-05,
+      "loss": 1.4396,
+      "step": 2780
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.035800576210022,
+      "learning_rate": 4.9786121534345265e-05,
+      "loss": 1.3593,
+      "step": 2785
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6600098609924316,
+      "learning_rate": 4.97853539810182e-05,
+      "loss": 1.4073,
+      "step": 2790
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6001609563827515,
+      "learning_rate": 4.9784585058821807e-05,
+      "loss": 1.3929,
+      "step": 2795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7586673498153687,
+      "learning_rate": 4.9783814767798545e-05,
+      "loss": 1.1887,
+      "step": 2800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.542594313621521,
+      "learning_rate": 4.9783043107990946e-05,
+      "loss": 1.4094,
+      "step": 2805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1905423402786255,
+      "learning_rate": 4.978227007944164e-05,
+      "loss": 1.4199,
+      "step": 2810
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1273014545440674,
+      "learning_rate": 4.978149568219332e-05,
+      "loss": 1.5138,
+      "step": 2815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8764979839324951,
+      "learning_rate": 4.978071991628875e-05,
+      "loss": 1.5155,
+      "step": 2820
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8033466339111328,
+      "learning_rate": 4.9779942781770776e-05,
+      "loss": 1.3844,
+      "step": 2825
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9946221113204956,
+      "learning_rate": 4.9779164278682324e-05,
+      "loss": 1.3978,
+      "step": 2830
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.596258282661438,
+      "learning_rate": 4.977838440706638e-05,
+      "loss": 1.4013,
+      "step": 2835
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.155501365661621,
+      "learning_rate": 4.977760316696603e-05,
+      "loss": 1.4486,
+      "step": 2840
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8394220471382141,
+      "learning_rate": 4.9776820558424396e-05,
+      "loss": 1.4483,
+      "step": 2845
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8399643301963806,
+      "learning_rate": 4.977603658148473e-05,
+      "loss": 1.4999,
+      "step": 2850
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7223281860351562,
+      "learning_rate": 4.977525123619031e-05,
+      "loss": 1.4209,
+      "step": 2855
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8585662245750427,
+      "learning_rate": 4.9774464522584516e-05,
+      "loss": 1.2834,
+      "step": 2860
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.903371274471283,
+      "learning_rate": 4.9773676440710804e-05,
+      "loss": 1.514,
+      "step": 2865
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6225999593734741,
+      "learning_rate": 4.977288699061269e-05,
+      "loss": 1.4325,
+      "step": 2870
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8646724224090576,
+      "learning_rate": 4.977209617233378e-05,
+      "loss": 1.2028,
+      "step": 2875
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7862356305122375,
+      "learning_rate": 4.9771303985917745e-05,
+      "loss": 1.4069,
+      "step": 2880
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.635420024394989,
+      "learning_rate": 4.977051043140834e-05,
+      "loss": 1.4109,
+      "step": 2885
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7677288055419922,
+      "learning_rate": 4.976971550884939e-05,
+      "loss": 1.1608,
+      "step": 2890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.845956027507782,
+      "learning_rate": 4.9768919218284804e-05,
+      "loss": 1.4072,
+      "step": 2895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9149651527404785,
+      "learning_rate": 4.976812155975855e-05,
+      "loss": 1.3957,
+      "step": 2900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9289765954017639,
+      "learning_rate": 4.9767322533314685e-05,
+      "loss": 1.4147,
+      "step": 2905
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1953096389770508,
+      "learning_rate": 4.9766522138997347e-05,
+      "loss": 1.4155,
+      "step": 2910
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9393072128295898,
+      "learning_rate": 4.976572037685073e-05,
+      "loss": 1.4166,
+      "step": 2915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.012731909751892,
+      "learning_rate": 4.9764917246919125e-05,
+      "loss": 1.4684,
+      "step": 2920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9099532961845398,
+      "learning_rate": 4.9764112749246876e-05,
+      "loss": 1.4124,
+      "step": 2925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.69303959608078,
+      "learning_rate": 4.976330688387842e-05,
+      "loss": 1.4344,
+      "step": 2930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6986569762229919,
+      "learning_rate": 4.9762499650858274e-05,
+      "loss": 1.3507,
+      "step": 2935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5651214718818665,
+      "learning_rate": 4.976169105023101e-05,
+      "loss": 1.2479,
+      "step": 2940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7775252461433411,
+      "learning_rate": 4.9760881082041275e-05,
+      "loss": 1.3942,
+      "step": 2945
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.379727840423584,
+      "learning_rate": 4.976006974633383e-05,
+      "loss": 1.4428,
+      "step": 2950
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.850517749786377,
+      "learning_rate": 4.9759257043153454e-05,
+      "loss": 1.4216,
+      "step": 2955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9250400066375732,
+      "learning_rate": 4.975844297254506e-05,
+      "loss": 1.4909,
+      "step": 2960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5304796695709229,
+      "learning_rate": 4.975762753455359e-05,
+      "loss": 1.4814,
+      "step": 2965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1414655447006226,
+      "learning_rate": 4.975681072922409e-05,
+      "loss": 1.408,
+      "step": 2970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5515819191932678,
+      "learning_rate": 4.975599255660166e-05,
+      "loss": 1.3775,
+      "step": 2975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8293300867080688,
+      "learning_rate": 4.97551730167315e-05,
+      "loss": 1.3132,
+      "step": 2980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0170328617095947,
+      "learning_rate": 4.9754352109658865e-05,
+      "loss": 1.4993,
+      "step": 2985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.937022864818573,
+      "learning_rate": 4.9753529835429094e-05,
+      "loss": 1.3501,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9053025245666504,
+      "learning_rate": 4.97527061940876e-05,
+      "loss": 1.4049,
+      "step": 2995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.49352753162384033,
+      "learning_rate": 4.975188118567986e-05,
+      "loss": 1.4339,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.048647165298462,
+      "learning_rate": 4.975105481025146e-05,
+      "loss": 1.3139,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6800316572189331,
+      "learning_rate": 4.9750227067848034e-05,
+      "loss": 1.2836,
+      "step": 3010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1411221027374268,
+      "learning_rate": 4.974939795851529e-05,
+      "loss": 1.4515,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7004803419113159,
+      "learning_rate": 4.9748567482299025e-05,
+      "loss": 1.551,
+      "step": 3020
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4398937225341797,
+      "learning_rate": 4.97477356392451e-05,
+      "loss": 1.4192,
+      "step": 3025
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5903295278549194,
+      "learning_rate": 4.974690242939946e-05,
+      "loss": 1.5211,
+      "step": 3030
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6511791348457336,
+      "learning_rate": 4.974606785280812e-05,
+      "loss": 1.5211,
+      "step": 3035
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2127279043197632,
+      "learning_rate": 4.9745231909517176e-05,
+      "loss": 1.3477,
+      "step": 3040
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.202561855316162,
+      "learning_rate": 4.9744394599572795e-05,
+      "loss": 1.5234,
+      "step": 3045
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7070451378822327,
+      "learning_rate": 4.974355592302122e-05,
+      "loss": 1.4987,
+      "step": 3050
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9108610153198242,
+      "learning_rate": 4.974271587990877e-05,
+      "loss": 1.2861,
+      "step": 3055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9873246550559998,
+      "learning_rate": 4.974187447028184e-05,
+      "loss": 1.4222,
+      "step": 3060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2977701425552368,
+      "learning_rate": 4.9741031694186904e-05,
+      "loss": 1.4062,
+      "step": 3065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.025950312614441,
+      "learning_rate": 4.9740187551670505e-05,
+      "loss": 1.5189,
+      "step": 3070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1865170001983643,
+      "learning_rate": 4.973934204277926e-05,
+      "loss": 1.3828,
+      "step": 3075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.818302571773529,
+      "learning_rate": 4.973849516755987e-05,
+      "loss": 1.467,
+      "step": 3080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8224700689315796,
+      "learning_rate": 4.9737646926059104e-05,
+      "loss": 1.7332,
+      "step": 3085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3279811143875122,
+      "learning_rate": 4.973679731832381e-05,
+      "loss": 1.5156,
+      "step": 3090
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5964494347572327,
+      "learning_rate": 4.973594634440092e-05,
+      "loss": 1.3524,
+      "step": 3095
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.196366786956787,
+      "learning_rate": 4.9735094004337427e-05,
+      "loss": 1.5278,
+      "step": 3100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7338801622390747,
+      "learning_rate": 4.97342402981804e-05,
+      "loss": 1.4148,
+      "step": 3105
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7936698794364929,
+      "learning_rate": 4.973338522597698e-05,
+      "loss": 1.4142,
+      "step": 3110
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5636515021324158,
+      "learning_rate": 4.9732528787774416e-05,
+      "loss": 1.5779,
+      "step": 3115
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5962295532226562,
+      "learning_rate": 4.973167098361999e-05,
+      "loss": 1.6513,
+      "step": 3120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.47786709666252136,
+      "learning_rate": 4.9730811813561083e-05,
+      "loss": 1.3843,
+      "step": 3125
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7690017223358154,
+      "learning_rate": 4.972995127764515e-05,
+      "loss": 1.2518,
+      "step": 3130
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7740575671195984,
+      "learning_rate": 4.97290893759197e-05,
+      "loss": 1.487,
+      "step": 3135
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7592212557792664,
+      "learning_rate": 4.972822610843236e-05,
+      "loss": 1.4385,
+      "step": 3140
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5696493983268738,
+      "learning_rate": 4.972736147523079e-05,
+      "loss": 1.4802,
+      "step": 3145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5601330995559692,
+      "learning_rate": 4.9726495476362756e-05,
+      "loss": 1.4203,
+      "step": 3150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0239603519439697,
+      "learning_rate": 4.972562811187608e-05,
+      "loss": 1.4046,
+      "step": 3155
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7492188811302185,
+      "learning_rate": 4.972475938181865e-05,
+      "loss": 1.3938,
+      "step": 3160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0243171453475952,
+      "learning_rate": 4.972388928623847e-05,
+      "loss": 1.2638,
+      "step": 3165
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4914931058883667,
+      "learning_rate": 4.972301782518358e-05,
+      "loss": 1.4414,
+      "step": 3170
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6274807453155518,
+      "learning_rate": 4.972214499870212e-05,
+      "loss": 1.4207,
+      "step": 3175
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9303854703903198,
+      "learning_rate": 4.9721270806842277e-05,
+      "loss": 1.4353,
+      "step": 3180
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8997390866279602,
+      "learning_rate": 4.9720395249652355e-05,
+      "loss": 1.3493,
+      "step": 3185
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1290831565856934,
+      "learning_rate": 4.97195183271807e-05,
+      "loss": 1.4754,
+      "step": 3190
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7762711644172668,
+      "learning_rate": 4.971864003947573e-05,
+      "loss": 1.45,
+      "step": 3195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6150183081626892,
+      "learning_rate": 4.971776038658598e-05,
+      "loss": 1.2738,
+      "step": 3200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6307562589645386,
+      "learning_rate": 4.971687936856e-05,
+      "loss": 1.4348,
+      "step": 3205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7655115127563477,
+      "learning_rate": 4.971599698544648e-05,
+      "loss": 1.2861,
+      "step": 3210
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8392760157585144,
+      "learning_rate": 4.971511323729412e-05,
+      "loss": 1.4351,
+      "step": 3215
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.204959511756897,
+      "learning_rate": 4.9714228124151756e-05,
+      "loss": 1.4165,
+      "step": 3220
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6539895534515381,
+      "learning_rate": 4.9713341646068264e-05,
+      "loss": 1.3264,
+      "step": 3225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6086089611053467,
+      "learning_rate": 4.97124538030926e-05,
+      "loss": 1.3271,
+      "step": 3230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9812304377555847,
+      "learning_rate": 4.97115645952738e-05,
+      "loss": 1.4253,
+      "step": 3235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5627865791320801,
+      "learning_rate": 4.9710674022660964e-05,
+      "loss": 1.4749,
+      "step": 3240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.062355399131775,
+      "learning_rate": 4.970978208530329e-05,
+      "loss": 1.3088,
+      "step": 3245
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5620614290237427,
+      "learning_rate": 4.9708888783250047e-05,
+      "loss": 1.3848,
+      "step": 3250
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4596127271652222,
+      "learning_rate": 4.970799411655055e-05,
+      "loss": 1.5558,
+      "step": 3255
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5726821422576904,
+      "learning_rate": 4.9707098085254224e-05,
+      "loss": 1.4792,
+      "step": 3260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5486913919448853,
+      "learning_rate": 4.970620068941055e-05,
+      "loss": 1.3452,
+      "step": 3265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5982449650764465,
+      "learning_rate": 4.9705301929069094e-05,
+      "loss": 1.3137,
+      "step": 3270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.129350185394287,
+      "learning_rate": 4.9704401804279495e-05,
+      "loss": 1.3673,
+      "step": 3275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7439037561416626,
+      "learning_rate": 4.9703500315091455e-05,
+      "loss": 1.3652,
+      "step": 3280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6493015885353088,
+      "learning_rate": 4.970259746155478e-05,
+      "loss": 1.3959,
+      "step": 3285
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.614264726638794,
+      "learning_rate": 4.9701693243719324e-05,
+      "loss": 1.5581,
+      "step": 3290
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9825220108032227,
+      "learning_rate": 4.970078766163502e-05,
+      "loss": 1.3638,
+      "step": 3295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.509028971195221,
+      "learning_rate": 4.9699880715351884e-05,
+      "loss": 1.4559,
+      "step": 3300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7303828001022339,
+      "learning_rate": 4.969897240492002e-05,
+      "loss": 1.2282,
+      "step": 3305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7061027884483337,
+      "learning_rate": 4.9698062730389586e-05,
+      "loss": 1.4301,
+      "step": 3310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5220503211021423,
+      "learning_rate": 4.9697151691810814e-05,
+      "loss": 1.5012,
+      "step": 3315
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2105666399002075,
+      "learning_rate": 4.969623928923402e-05,
+      "loss": 1.5438,
+      "step": 3320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6172266006469727,
+      "learning_rate": 4.969532552270961e-05,
+      "loss": 1.3556,
+      "step": 3325
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.089318037033081,
+      "learning_rate": 4.969441039228803e-05,
+      "loss": 1.3997,
+      "step": 3330
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8561093211174011,
+      "learning_rate": 4.969349389801984e-05,
+      "loss": 1.3716,
+      "step": 3335
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4685690701007843,
+      "learning_rate": 4.969257603995565e-05,
+      "loss": 1.4295,
+      "step": 3340
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1991277933120728,
+      "learning_rate": 4.9691656818146146e-05,
+      "loss": 1.3521,
+      "step": 3345
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0424867868423462,
+      "learning_rate": 4.969073623264211e-05,
+      "loss": 1.4858,
+      "step": 3350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1299071311950684,
+      "learning_rate": 4.968981428349438e-05,
+      "loss": 1.3135,
+      "step": 3355
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6989724636077881,
+      "learning_rate": 4.9688890970753856e-05,
+      "loss": 1.4063,
+      "step": 3360
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4703225791454315,
+      "learning_rate": 4.968796629447155e-05,
+      "loss": 1.3124,
+      "step": 3365
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.0257688760757446,
+      "learning_rate": 4.968704025469853e-05,
+      "loss": 1.3443,
+      "step": 3370
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8301354050636292,
+      "learning_rate": 4.968611285148594e-05,
+      "loss": 1.2692,
+      "step": 3375
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.3049982786178589,
+      "learning_rate": 4.968518408488499e-05,
+      "loss": 1.4013,
+      "step": 3380
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.38068425655365,
+      "learning_rate": 4.968425395494699e-05,
+      "loss": 1.3014,
+      "step": 3385
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8539844751358032,
+      "learning_rate": 4.96833224617233e-05,
+      "loss": 1.3768,
+      "step": 3390
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6281200647354126,
+      "learning_rate": 4.968238960526537e-05,
+      "loss": 1.3423,
+      "step": 3395
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.638692855834961,
+      "learning_rate": 4.968145538562471e-05,
+      "loss": 1.4069,
+      "step": 3400
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.0412074327468872,
+      "learning_rate": 4.968051980285293e-05,
+      "loss": 1.3327,
+      "step": 3405
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.5191795825958252,
+      "learning_rate": 4.967958285700169e-05,
+      "loss": 1.4662,
+      "step": 3410
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.035595417022705,
+      "learning_rate": 4.967864454812274e-05,
+      "loss": 1.4095,
+      "step": 3415
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.3538129329681396,
+      "learning_rate": 4.967770487626791e-05,
+      "loss": 1.4562,
+      "step": 3420
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.785943865776062,
+      "learning_rate": 4.9676763841489093e-05,
+      "loss": 1.2219,
+      "step": 3425
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9118176698684692,
+      "learning_rate": 4.967582144383826e-05,
+      "loss": 1.3501,
+      "step": 3430
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7986450791358948,
+      "learning_rate": 4.967487768336745e-05,
+      "loss": 1.5203,
+      "step": 3435
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6016656160354614,
+      "learning_rate": 4.967393256012879e-05,
+      "loss": 1.3272,
+      "step": 3440
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.1561404466629028,
+      "learning_rate": 4.967298607417449e-05,
+      "loss": 1.3276,
+      "step": 3445
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.0883400440216064,
+      "learning_rate": 4.9672038225556816e-05,
+      "loss": 1.4162,
+      "step": 3450
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.2652868032455444,
+      "learning_rate": 4.967108901432811e-05,
+      "loss": 1.4396,
+      "step": 3455
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.0570058822631836,
+      "learning_rate": 4.967013844054081e-05,
+      "loss": 1.2359,
+      "step": 3460
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9641575217247009,
+      "learning_rate": 4.96691865042474e-05,
+      "loss": 1.3216,
+      "step": 3465
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.511062502861023,
+      "learning_rate": 4.966823320550047e-05,
+      "loss": 1.4389,
+      "step": 3470
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7163792252540588,
+      "learning_rate": 4.9667278544352653e-05,
+      "loss": 1.4032,
+      "step": 3475
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.194694995880127,
+      "learning_rate": 4.966632252085668e-05,
+      "loss": 1.3817,
+      "step": 3480
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.0525634288787842,
+      "learning_rate": 4.9665365135065365e-05,
+      "loss": 1.3344,
+      "step": 3485
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.383202075958252,
+      "learning_rate": 4.966440638703156e-05,
+      "loss": 1.3635,
+      "step": 3490
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7507001757621765,
+      "learning_rate": 4.9663446276808235e-05,
+      "loss": 1.1945,
+      "step": 3495
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.1535072326660156,
+      "learning_rate": 4.9662484804448404e-05,
+      "loss": 1.5449,
+      "step": 3500
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.666061282157898,
+      "learning_rate": 4.966152197000517e-05,
+      "loss": 1.4527,
+      "step": 3505
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6894571781158447,
+      "learning_rate": 4.9660557773531723e-05,
+      "loss": 1.4377,
+      "step": 3510
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7343246936798096,
+      "learning_rate": 4.9659592215081296e-05,
+      "loss": 1.6301,
+      "step": 3515
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.4154441356658936,
+      "learning_rate": 4.9658625294707226e-05,
+      "loss": 1.4235,
+      "step": 3520
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.732731580734253,
+      "learning_rate": 4.9657657012462904e-05,
+      "loss": 1.491,
+      "step": 3525
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.0298057794570923,
+      "learning_rate": 4.965668736840182e-05,
+      "loss": 1.4066,
+      "step": 3530
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.9874246716499329,
+      "learning_rate": 4.9655716362577525e-05,
+      "loss": 1.4336,
+      "step": 3535
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7678321003913879,
+      "learning_rate": 4.965474399504364e-05,
+      "loss": 1.4643,
+      "step": 3540
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8920181393623352,
+      "learning_rate": 4.9653770265853874e-05,
+      "loss": 1.514,
+      "step": 3545
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7572281956672668,
+      "learning_rate": 4.9652795175062005e-05,
+      "loss": 1.3438,
+      "step": 3550
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7819041609764099,
+      "learning_rate": 4.965181872272188e-05,
+      "loss": 1.3153,
+      "step": 3555
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8489050269126892,
+      "learning_rate": 4.965084090888743e-05,
+      "loss": 1.3211,
+      "step": 3560
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.9903485774993896,
+      "learning_rate": 4.9649861733612654e-05,
+      "loss": 1.3094,
+      "step": 3565
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8128799796104431,
+      "learning_rate": 4.9648881196951647e-05,
+      "loss": 1.3364,
+      "step": 3570
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.2447357177734375,
+      "learning_rate": 4.964789929895855e-05,
+      "loss": 1.5058,
+      "step": 3575
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.014669418334961,
+      "learning_rate": 4.96469160396876e-05,
+      "loss": 1.2516,
+      "step": 3580
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8545610308647156,
+      "learning_rate": 4.964593141919308e-05,
+      "loss": 1.3384,
+      "step": 3585
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7975364923477173,
+      "learning_rate": 4.96449454375294e-05,
+      "loss": 1.4322,
+      "step": 3590
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7121825218200684,
+      "learning_rate": 4.9643958094750996e-05,
+      "loss": 1.3345,
+      "step": 3595
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.9037702083587646,
+      "learning_rate": 4.9642969390912394e-05,
+      "loss": 1.4319,
+      "step": 3600
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.2663486003875732,
+      "learning_rate": 4.964197932606822e-05,
+      "loss": 1.6255,
+      "step": 3605
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8707221150398254,
+      "learning_rate": 4.964098790027314e-05,
+      "loss": 1.4954,
+      "step": 3610
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.0182719230651855,
+      "learning_rate": 4.963999511358191e-05,
+      "loss": 1.2573,
+      "step": 3615
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5255160927772522,
+      "learning_rate": 4.963900096604936e-05,
+      "loss": 1.2647,
+      "step": 3620
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.4247102737426758,
+      "learning_rate": 4.96380054577304e-05,
+      "loss": 1.3231,
+      "step": 3625
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5628049969673157,
+      "learning_rate": 4.963700858868e-05,
+      "loss": 1.2791,
+      "step": 3630
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.755516529083252,
+      "learning_rate": 4.9636010358953235e-05,
+      "loss": 1.4192,
+      "step": 3635
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.752839982509613,
+      "learning_rate": 4.963501076860522e-05,
+      "loss": 1.3058,
+      "step": 3640
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.969702959060669,
+      "learning_rate": 4.963400981769117e-05,
+      "loss": 1.2481,
+      "step": 3645
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.9159872531890869,
+      "learning_rate": 4.963300750626636e-05,
+      "loss": 1.4535,
+      "step": 3650
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8840392827987671,
+      "learning_rate": 4.963200383438615e-05,
+      "loss": 1.4412,
+      "step": 3655
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.0793917179107666,
+      "learning_rate": 4.9630998802105975e-05,
+      "loss": 1.3668,
+      "step": 3660
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.078147053718567,
+      "learning_rate": 4.962999240948134e-05,
+      "loss": 1.3659,
+      "step": 3665
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8372591137886047,
+      "learning_rate": 4.962898465656782e-05,
+      "loss": 1.3811,
+      "step": 3670
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.0935574769973755,
+      "learning_rate": 4.962797554342108e-05,
+      "loss": 1.4259,
+      "step": 3675
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.2438262701034546,
+      "learning_rate": 4.962696507009685e-05,
+      "loss": 1.4393,
+      "step": 3680
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8623482584953308,
+      "learning_rate": 4.962595323665094e-05,
+      "loss": 1.5026,
+      "step": 3685
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.854569911956787,
+      "learning_rate": 4.962494004313923e-05,
+      "loss": 1.4825,
+      "step": 3690
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.069012999534607,
+      "learning_rate": 4.962392548961767e-05,
+      "loss": 1.428,
+      "step": 3695
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.0599278211593628,
+      "learning_rate": 4.9622909576142304e-05,
+      "loss": 1.5312,
+      "step": 3700
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7758433818817139,
+      "learning_rate": 4.962189230276925e-05,
+      "loss": 1.3304,
+      "step": 3705
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.406180500984192,
+      "learning_rate": 4.962087366955466e-05,
+      "loss": 1.4616,
+      "step": 3710
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6990206837654114,
+      "learning_rate": 4.9619853676554814e-05,
+      "loss": 1.3334,
+      "step": 3715
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6970654726028442,
+      "learning_rate": 4.9618832323826036e-05,
+      "loss": 1.3963,
+      "step": 3720
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8413794040679932,
+      "learning_rate": 4.9617809611424745e-05,
+      "loss": 1.3082,
+      "step": 3725
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9142487645149231,
+      "learning_rate": 4.9616785539407414e-05,
+      "loss": 1.3984,
+      "step": 3730
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7050864696502686,
+      "learning_rate": 4.961576010783061e-05,
+      "loss": 1.3641,
+      "step": 3735
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7158265709877014,
+      "learning_rate": 4.961473331675096e-05,
+      "loss": 1.3355,
+      "step": 3740
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8010573387145996,
+      "learning_rate": 4.961370516622518e-05,
+      "loss": 1.3238,
+      "step": 3745
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9509638547897339,
+      "learning_rate": 4.961267565631004e-05,
+      "loss": 1.4979,
+      "step": 3750
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8768844604492188,
+      "learning_rate": 4.96116447870624e-05,
+      "loss": 1.3485,
+      "step": 3755
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.029345989227295,
+      "learning_rate": 4.961061255853921e-05,
+      "loss": 1.4198,
+      "step": 3760
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0226683616638184,
+      "learning_rate": 4.960957897079748e-05,
+      "loss": 1.501,
+      "step": 3765
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.1053155660629272,
+      "learning_rate": 4.9608544023894276e-05,
+      "loss": 1.5134,
+      "step": 3770
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7653586864471436,
+      "learning_rate": 4.960750771788676e-05,
+      "loss": 1.3514,
+      "step": 3775
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.1900826692581177,
+      "learning_rate": 4.9606470052832174e-05,
+      "loss": 1.3914,
+      "step": 3780
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7287219762802124,
+      "learning_rate": 4.960543102878782e-05,
+      "loss": 1.2752,
+      "step": 3785
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8212308287620544,
+      "learning_rate": 4.96043906458111e-05,
+      "loss": 1.3529,
+      "step": 3790
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0006697177886963,
+      "learning_rate": 4.960334890395944e-05,
+      "loss": 1.2957,
+      "step": 3795
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5815529823303223,
+      "learning_rate": 4.96023058032904e-05,
+      "loss": 1.6203,
+      "step": 3800
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.1902085542678833,
+      "learning_rate": 4.9601261343861586e-05,
+      "loss": 1.5133,
+      "step": 3805
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7083724141120911,
+      "learning_rate": 4.960021552573068e-05,
+      "loss": 1.415,
+      "step": 3810
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.382954716682434,
+      "learning_rate": 4.959916834895544e-05,
+      "loss": 1.4189,
+      "step": 3815
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.2294280529022217,
+      "learning_rate": 4.95981198135937e-05,
+      "loss": 1.5006,
+      "step": 3820
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6995631456375122,
+      "learning_rate": 4.9597069919703375e-05,
+      "loss": 1.5182,
+      "step": 3825
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8354824185371399,
+      "learning_rate": 4.959601866734245e-05,
+      "loss": 1.4081,
+      "step": 3830
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.210512399673462,
+      "learning_rate": 4.959496605656897e-05,
+      "loss": 1.428,
+      "step": 3835
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7832702994346619,
+      "learning_rate": 4.959391208744108e-05,
+      "loss": 1.2983,
+      "step": 3840
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.9913911819458008,
+      "learning_rate": 4.959285676001699e-05,
+      "loss": 1.4373,
+      "step": 3845
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5717589855194092,
+      "learning_rate": 4.9591800074354987e-05,
+      "loss": 1.446,
+      "step": 3850
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.780748188495636,
+      "learning_rate": 4.959074203051343e-05,
+      "loss": 1.2086,
+      "step": 3855
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7067710161209106,
+      "learning_rate": 4.958968262855075e-05,
+      "loss": 1.4519,
+      "step": 3860
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.3409600257873535,
+      "learning_rate": 4.958862186852545e-05,
+      "loss": 1.3027,
+      "step": 3865
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7507779598236084,
+      "learning_rate": 4.9587559750496135e-05,
+      "loss": 1.415,
+      "step": 3870
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7299233078956604,
+      "learning_rate": 4.9586496274521446e-05,
+      "loss": 1.3096,
+      "step": 3875
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.76593017578125,
+      "learning_rate": 4.958543144066012e-05,
+      "loss": 1.4701,
+      "step": 3880
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9224169254302979,
+      "learning_rate": 4.958436524897098e-05,
+      "loss": 1.4677,
+      "step": 3885
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.777151882648468,
+      "learning_rate": 4.95832976995129e-05,
+      "loss": 1.4735,
+      "step": 3890
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.2696973085403442,
+      "learning_rate": 4.958222879234483e-05,
+      "loss": 1.4598,
+      "step": 3895
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.122017741203308,
+      "learning_rate": 4.958115852752583e-05,
+      "loss": 1.2936,
+      "step": 3900
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7346524000167847,
+      "learning_rate": 4.9580086905114984e-05,
+      "loss": 1.4084,
+      "step": 3905
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.49281907081604,
+      "learning_rate": 4.957901392517149e-05,
+      "loss": 1.2421,
+      "step": 3910
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9897387027740479,
+      "learning_rate": 4.957793958775461e-05,
+      "loss": 1.4164,
+      "step": 3915
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.001846432685852,
+      "learning_rate": 4.9576863892923676e-05,
+      "loss": 1.1946,
+      "step": 3920
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9374973773956299,
+      "learning_rate": 4.9575786840738085e-05,
+      "loss": 1.4949,
+      "step": 3925
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.216041088104248,
+      "learning_rate": 4.957470843125734e-05,
+      "loss": 1.4853,
+      "step": 3930
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8419601917266846,
+      "learning_rate": 4.9573628664540985e-05,
+      "loss": 1.2595,
+      "step": 3935
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.0801103115081787,
+      "learning_rate": 4.957254754064867e-05,
+      "loss": 1.2968,
+      "step": 3940
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.5771223306655884,
+      "learning_rate": 4.9571465059640094e-05,
+      "loss": 1.3501,
+      "step": 3945
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.0766230821609497,
+      "learning_rate": 4.957038122157504e-05,
+      "loss": 1.4484,
+      "step": 3950
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.5095306634902954,
+      "learning_rate": 4.9569296026513374e-05,
+      "loss": 1.3842,
+      "step": 3955
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.0730971097946167,
+      "learning_rate": 4.956820947451503e-05,
+      "loss": 1.3855,
+      "step": 3960
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.0428813695907593,
+      "learning_rate": 4.956712156564001e-05,
+      "loss": 1.4501,
+      "step": 3965
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.020739197731018,
+      "learning_rate": 4.9566032299948394e-05,
+      "loss": 1.2739,
+      "step": 3970
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.895963191986084,
+      "learning_rate": 4.956494167750036e-05,
+      "loss": 1.2692,
+      "step": 3975
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5474923849105835,
+      "learning_rate": 4.956384969835612e-05,
+      "loss": 1.348,
+      "step": 3980
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.9486867785453796,
+      "learning_rate": 4.956275636257601e-05,
+      "loss": 1.3352,
+      "step": 3985
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.9013283848762512,
+      "learning_rate": 4.9561661670220386e-05,
+      "loss": 1.4037,
+      "step": 3990
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8151796460151672,
+      "learning_rate": 4.956056562134972e-05,
+      "loss": 1.383,
+      "step": 3995
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0700517892837524,
+      "learning_rate": 4.955946821602455e-05,
+      "loss": 1.4735,
+      "step": 4000
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7493061423301697,
+      "learning_rate": 4.9558369454305476e-05,
+      "loss": 1.4365,
+      "step": 4005
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9456654191017151,
+      "learning_rate": 4.955726933625318e-05,
+      "loss": 1.2838,
+      "step": 4010
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0961682796478271,
+      "learning_rate": 4.955616786192843e-05,
+      "loss": 1.3222,
+      "step": 4015
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0026607513427734,
+      "learning_rate": 4.9555065031392044e-05,
+      "loss": 1.3279,
+      "step": 4020
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.1015264987945557,
+      "learning_rate": 4.955396084470495e-05,
+      "loss": 1.2958,
+      "step": 4025
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.213651180267334,
+      "learning_rate": 4.9552855301928114e-05,
+      "loss": 1.3827,
+      "step": 4030
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8415979743003845,
+      "learning_rate": 4.9551748403122604e-05,
+      "loss": 1.3701,
+      "step": 4035
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.27717924118042,
+      "learning_rate": 4.955064014834955e-05,
+      "loss": 1.2613,
+      "step": 4040
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7791318297386169,
+      "learning_rate": 4.954953053767016e-05,
+      "loss": 1.2606,
+      "step": 4045
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8190057277679443,
+      "learning_rate": 4.954841957114572e-05,
+      "loss": 1.3877,
+      "step": 4050
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.4794509410858154,
+      "learning_rate": 4.954730724883757e-05,
+      "loss": 1.3563,
+      "step": 4055
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.701057493686676,
+      "learning_rate": 4.954619357080717e-05,
+      "loss": 1.3641,
+      "step": 4060
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7502091526985168,
+      "learning_rate": 4.954507853711601e-05,
+      "loss": 1.3876,
+      "step": 4065
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.5233298540115356,
+      "learning_rate": 4.9543962147825675e-05,
+      "loss": 1.3748,
+      "step": 4070
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7826765775680542,
+      "learning_rate": 4.954284440299782e-05,
+      "loss": 1.3566,
+      "step": 4075
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7345137596130371,
+      "learning_rate": 4.9541725302694185e-05,
+      "loss": 1.3793,
+      "step": 4080
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.9479917883872986,
+      "learning_rate": 4.954060484697657e-05,
+      "loss": 1.4317,
+      "step": 4085
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6998075246810913,
+      "learning_rate": 4.9539483035906854e-05,
+      "loss": 1.5301,
+      "step": 4090
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9664790034294128,
+      "learning_rate": 4.9538359869546996e-05,
+      "loss": 1.4804,
+      "step": 4095
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9688673615455627,
+      "learning_rate": 4.9537235347959034e-05,
+      "loss": 1.3919,
+      "step": 4100
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8500693440437317,
+      "learning_rate": 4.953610947120506e-05,
+      "loss": 1.4584,
+      "step": 4105
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.2263480424880981,
+      "learning_rate": 4.953498223934727e-05,
+      "loss": 1.3839,
+      "step": 4110
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9976312518119812,
+      "learning_rate": 4.953385365244791e-05,
+      "loss": 1.4062,
+      "step": 4115
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.165010690689087,
+      "learning_rate": 4.953272371056933e-05,
+      "loss": 1.3283,
+      "step": 4120
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": NaN,
+      "learning_rate": 4.953181878152334e-05,
+      "loss": 1.3901,
+      "step": 4125
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8692017793655396,
+      "learning_rate": 4.9530686400839435e-05,
+      "loss": 1.4443,
+      "step": 4130
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.5308072566986084,
+      "learning_rate": 4.952955266535122e-05,
+      "loss": 1.3798,
+      "step": 4135
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.0614334344863892,
+      "learning_rate": 4.952841757512131e-05,
+      "loss": 1.2199,
+      "step": 4140
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5995301604270935,
+      "learning_rate": 4.952728113021239e-05,
+      "loss": 1.3149,
+      "step": 4145
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9404197335243225,
+      "learning_rate": 4.9526143330687224e-05,
+      "loss": 1.4628,
+      "step": 4150
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.0483546257019043,
+      "learning_rate": 4.9525004176608655e-05,
+      "loss": 1.4337,
+      "step": 4155
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.1325901746749878,
+      "learning_rate": 4.9523863668039606e-05,
+      "loss": 1.3896,
+      "step": 4160
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.8560702800750732,
+      "learning_rate": 4.952272180504306e-05,
+      "loss": 1.5117,
+      "step": 4165
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0659074783325195,
+      "learning_rate": 4.9521578587682074e-05,
+      "loss": 1.4522,
+      "step": 4170
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6918451189994812,
+      "learning_rate": 4.952043401601979e-05,
+      "loss": 1.2501,
+      "step": 4175
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.1893597841262817,
+      "learning_rate": 4.951928809011942e-05,
+      "loss": 1.507,
+      "step": 4180
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9000897407531738,
+      "learning_rate": 4.951814081004426e-05,
+      "loss": 1.3378,
+      "step": 4185
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.5121021270751953,
+      "learning_rate": 4.9516992175857665e-05,
+      "loss": 1.2653,
+      "step": 4190
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.9484086632728577,
+      "learning_rate": 4.9515842187623076e-05,
+      "loss": 1.304,
+      "step": 4195
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6229005455970764,
+      "learning_rate": 4.951469084540401e-05,
+      "loss": 1.3531,
+      "step": 4200
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.877293050289154,
+      "learning_rate": 4.951353814926405e-05,
+      "loss": 1.2548,
+      "step": 4205
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8774591088294983,
+      "learning_rate": 4.9512384099266854e-05,
+      "loss": 1.3658,
+      "step": 4210
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7639179229736328,
+      "learning_rate": 4.9511228695476165e-05,
+      "loss": 1.4088,
+      "step": 4215
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.9273248314857483,
+      "learning_rate": 4.9510071937955794e-05,
+      "loss": 1.3634,
+      "step": 4220
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.5506839752197266,
+      "learning_rate": 4.950891382676963e-05,
+      "loss": 1.4951,
+      "step": 4225
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.1920723915100098,
+      "learning_rate": 4.9507754361981625e-05,
+      "loss": 1.3341,
+      "step": 4230
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6930038332939148,
+      "learning_rate": 4.9506593543655824e-05,
+      "loss": 1.4112,
+      "step": 4235
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8218011260032654,
+      "learning_rate": 4.9505431371856334e-05,
+      "loss": 1.2737,
+      "step": 4240
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.1834139823913574,
+      "learning_rate": 4.950426784664734e-05,
+      "loss": 1.4452,
+      "step": 4245
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9107833504676819,
+      "learning_rate": 4.950310296809311e-05,
+      "loss": 1.3029,
+      "step": 4250
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9917646646499634,
+      "learning_rate": 4.950193673625796e-05,
+      "loss": 1.181,
+      "step": 4255
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8065804243087769,
+      "learning_rate": 4.9500769151206325e-05,
+      "loss": 1.3394,
+      "step": 4260
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.371374249458313,
+      "learning_rate": 4.9499600213002673e-05,
+      "loss": 1.374,
+      "step": 4265
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.1488021612167358,
+      "learning_rate": 4.9498429921711566e-05,
+      "loss": 1.3386,
+      "step": 4270
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6660547256469727,
+      "learning_rate": 4.9497258277397635e-05,
+      "loss": 1.4149,
+      "step": 4275
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.1452709436416626,
+      "learning_rate": 4.94960852801256e-05,
+      "loss": 1.5156,
+      "step": 4280
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7714586853981018,
+      "learning_rate": 4.949491092996023e-05,
+      "loss": 1.3724,
+      "step": 4285
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.2154569625854492,
+      "learning_rate": 4.94937352269664e-05,
+      "loss": 1.3114,
+      "step": 4290
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.180679440498352,
+      "learning_rate": 4.949255817120903e-05,
+      "loss": 1.4339,
+      "step": 4295
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.1234863996505737,
+      "learning_rate": 4.949137976275312e-05,
+      "loss": 1.238,
+      "step": 4300
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8907201290130615,
+      "learning_rate": 4.949020000166378e-05,
+      "loss": 1.1503,
+      "step": 4305
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9505663514137268,
+      "learning_rate": 4.9489018888006136e-05,
+      "loss": 1.3278,
+      "step": 4310
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0911295413970947,
+      "learning_rate": 4.948783642184544e-05,
+      "loss": 1.3792,
+      "step": 4315
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.295822262763977,
+      "learning_rate": 4.948665260324699e-05,
+      "loss": 1.4859,
+      "step": 4320
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6218513250350952,
+      "learning_rate": 4.948546743227617e-05,
+      "loss": 1.2983,
+      "step": 4325
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.901931881904602,
+      "learning_rate": 4.948428090899844e-05,
+      "loss": 1.4346,
+      "step": 4330
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.49190616607666,
+      "learning_rate": 4.9483093033479324e-05,
+      "loss": 1.245,
+      "step": 4335
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7503476738929749,
+      "learning_rate": 4.948190380578442e-05,
+      "loss": 1.3196,
+      "step": 4340
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7911471724510193,
+      "learning_rate": 4.948071322597942e-05,
+      "loss": 1.3248,
+      "step": 4345
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.4888834953308105,
+      "learning_rate": 4.947952129413008e-05,
+      "loss": 1.4045,
+      "step": 4350
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8818658590316772,
+      "learning_rate": 4.9478328010302225e-05,
+      "loss": 1.4385,
+      "step": 4355
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.1271659135818481,
+      "learning_rate": 4.947713337456175e-05,
+      "loss": 1.3797,
+      "step": 4360
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.77896785736084,
+      "learning_rate": 4.9475937386974645e-05,
+      "loss": 1.4799,
+      "step": 4365
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8201467394828796,
+      "learning_rate": 4.947474004760696e-05,
+      "loss": 1.4012,
+      "step": 4370
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.4401566982269287,
+      "learning_rate": 4.947354135652482e-05,
+      "loss": 1.3238,
+      "step": 4375
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7641358971595764,
+      "learning_rate": 4.947234131379444e-05,
+      "loss": 1.4674,
+      "step": 4380
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.366176962852478,
+      "learning_rate": 4.947113991948207e-05,
+      "loss": 1.3375,
+      "step": 4385
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.671121597290039,
+      "learning_rate": 4.9469937173654094e-05,
+      "loss": 1.251,
+      "step": 4390
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6068382859230042,
+      "learning_rate": 4.9468733076376906e-05,
+      "loss": 1.4187,
+      "step": 4395
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8216054439544678,
+      "learning_rate": 4.9467527627717036e-05,
+      "loss": 1.3438,
+      "step": 4400
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8097289800643921,
+      "learning_rate": 4.946632082774104e-05,
+      "loss": 1.5979,
+      "step": 4405
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.1140400171279907,
+      "learning_rate": 4.946511267651559e-05,
+      "loss": 1.2796,
+      "step": 4410
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.9381734728813171,
+      "learning_rate": 4.9463903174107386e-05,
+      "loss": 1.3549,
+      "step": 4415
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.5891177654266357,
+      "learning_rate": 4.9462692320583236e-05,
+      "loss": 1.4088,
+      "step": 4420
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.4111642837524414,
+      "learning_rate": 4.946148011601003e-05,
+      "loss": 1.32,
+      "step": 4425
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.5155541896820068,
+      "learning_rate": 4.9460266560454696e-05,
+      "loss": 1.2985,
+      "step": 4430
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.3506906032562256,
+      "learning_rate": 4.9459051653984253e-05,
+      "loss": 1.2462,
+      "step": 4435
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.700868844985962,
+      "learning_rate": 4.945783539666583e-05,
+      "loss": 1.5184,
+      "step": 4440
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9799214601516724,
+      "learning_rate": 4.9456617788566576e-05,
+      "loss": 1.4099,
+      "step": 4445
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.3236010074615479,
+      "learning_rate": 4.945539882975373e-05,
+      "loss": 1.3994,
+      "step": 4450
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.069741725921631,
+      "learning_rate": 4.9454178520294634e-05,
+      "loss": 1.5627,
+      "step": 4455
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.1382434368133545,
+      "learning_rate": 4.9452956860256685e-05,
+      "loss": 1.415,
+      "step": 4460
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.9453111886978149,
+      "learning_rate": 4.945173384970734e-05,
+      "loss": 1.3312,
+      "step": 4465
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.3048676252365112,
+      "learning_rate": 4.9450509488714146e-05,
+      "loss": 1.3344,
+      "step": 4470
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.9417716860771179,
+      "learning_rate": 4.9449283777344736e-05,
+      "loss": 1.4794,
+      "step": 4475
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.0928821563720703,
+      "learning_rate": 4.944805671566679e-05,
+      "loss": 1.5453,
+      "step": 4480
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.025001049041748,
+      "learning_rate": 4.944682830374809e-05,
+      "loss": 1.3634,
+      "step": 4485
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.4281930923461914,
+      "learning_rate": 4.944559854165647e-05,
+      "loss": 1.2763,
+      "step": 4490
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.3360381126403809,
+      "learning_rate": 4.944436742945985e-05,
+      "loss": 1.4286,
+      "step": 4495
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.891915202140808,
+      "learning_rate": 4.944313496722623e-05,
+      "loss": 1.2857,
+      "step": 4500
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.244935154914856,
+      "learning_rate": 4.944190115502367e-05,
+      "loss": 1.3384,
+      "step": 4505
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.4173117876052856,
+      "learning_rate": 4.944066599292032e-05,
+      "loss": 1.4777,
+      "step": 4510
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.03322434425354,
+      "learning_rate": 4.943942948098439e-05,
+      "loss": 1.4247,
+      "step": 4515
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9695542454719543,
+      "learning_rate": 4.943819161928417e-05,
+      "loss": 1.3148,
+      "step": 4520
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.5252701044082642,
+      "learning_rate": 4.943695240788803e-05,
+      "loss": 1.4396,
+      "step": 4525
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.4560002088546753,
+      "learning_rate": 4.943571184686441e-05,
+      "loss": 1.5466,
+      "step": 4530
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7746248245239258,
+      "learning_rate": 4.9434469936281825e-05,
+      "loss": 1.4468,
+      "step": 4535
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7101891040802002,
+      "learning_rate": 4.943322667620886e-05,
+      "loss": 1.3618,
+      "step": 4540
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8597649931907654,
+      "learning_rate": 4.9431982066714186e-05,
+      "loss": 1.3144,
+      "step": 4545
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.0961127281188965,
+      "learning_rate": 4.9430736107866535e-05,
+      "loss": 1.3769,
+      "step": 4550
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.1628168821334839,
+      "learning_rate": 4.942948879973472e-05,
+      "loss": 1.427,
+      "step": 4555
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.946785569190979,
+      "learning_rate": 4.9428240142387646e-05,
+      "loss": 1.4057,
+      "step": 4560
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9042614698410034,
+      "learning_rate": 4.9426990135894245e-05,
+      "loss": 1.3838,
+      "step": 4565
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.2647825479507446,
+      "learning_rate": 4.942573878032358e-05,
+      "loss": 1.3895,
+      "step": 4570
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.29331636428833,
+      "learning_rate": 4.9424486075744745e-05,
+      "loss": 1.4304,
+      "step": 4575
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.3520721197128296,
+      "learning_rate": 4.9423232022226936e-05,
+      "loss": 1.2225,
+      "step": 4580
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6012137532234192,
+      "learning_rate": 4.942197661983941e-05,
+      "loss": 1.3025,
+      "step": 4585
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9096260070800781,
+      "learning_rate": 4.9420719868651496e-05,
+      "loss": 1.3349,
+      "step": 4590
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.007503628730774,
+      "learning_rate": 4.941946176873261e-05,
+      "loss": 1.3936,
+      "step": 4595
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.1042200326919556,
+      "learning_rate": 4.941820232015223e-05,
+      "loss": 1.3694,
+      "step": 4600
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5444115400314331,
+      "learning_rate": 4.941694152297992e-05,
+      "loss": 1.3672,
+      "step": 4605
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9802212715148926,
+      "learning_rate": 4.9415679377285305e-05,
+      "loss": 1.4417,
+      "step": 4610
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9913745522499084,
+      "learning_rate": 4.9414415883138106e-05,
+      "loss": 1.3331,
+      "step": 4615
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.2390682697296143,
+      "learning_rate": 4.941315104060808e-05,
+      "loss": 1.4843,
+      "step": 4620
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7250832915306091,
+      "learning_rate": 4.941188484976512e-05,
+      "loss": 1.4415,
+      "step": 4625
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6558685898780823,
+      "learning_rate": 4.941061731067912e-05,
+      "loss": 1.3587,
+      "step": 4630
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1380136013031006,
+      "learning_rate": 4.94093484234201e-05,
+      "loss": 1.3663,
+      "step": 4635
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.294971227645874,
+      "learning_rate": 4.9408078188058145e-05,
+      "loss": 1.402,
+      "step": 4640
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2890818119049072,
+      "learning_rate": 4.940680660466339e-05,
+      "loss": 1.3872,
+      "step": 4645
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7086050510406494,
+      "learning_rate": 4.9405533673306094e-05,
+      "loss": 1.4863,
+      "step": 4650
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.047402024269104,
+      "learning_rate": 4.940425939405653e-05,
+      "loss": 1.4587,
+      "step": 4655
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.5536178350448608,
+      "learning_rate": 4.940298376698508e-05,
+      "loss": 1.333,
+      "step": 4660
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.2016557455062866,
+      "learning_rate": 4.9401706792162215e-05,
+      "loss": 1.3375,
+      "step": 4665
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0145758390426636,
+      "learning_rate": 4.940042846965844e-05,
+      "loss": 1.3985,
+      "step": 4670
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.1620700359344482,
+      "learning_rate": 4.939914879954437e-05,
+      "loss": 1.4384,
+      "step": 4675
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6000904440879822,
+      "learning_rate": 4.9397867781890665e-05,
+      "loss": 1.3831,
+      "step": 4680
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.2584675550460815,
+      "learning_rate": 4.939658541676809e-05,
+      "loss": 1.3704,
+      "step": 4685
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9861582517623901,
+      "learning_rate": 4.939530170424745e-05,
+      "loss": 1.4437,
+      "step": 4690
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.411476492881775,
+      "learning_rate": 4.9394016644399666e-05,
+      "loss": 1.3568,
+      "step": 4695
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8464270234107971,
+      "learning_rate": 4.93927302372957e-05,
+      "loss": 1.414,
+      "step": 4700
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.0788897275924683,
+      "learning_rate": 4.939144248300659e-05,
+      "loss": 1.4118,
+      "step": 4705
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8137201070785522,
+      "learning_rate": 4.9390153381603466e-05,
+      "loss": 1.4496,
+      "step": 4710
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9756670594215393,
+      "learning_rate": 4.938886293315753e-05,
+      "loss": 1.4993,
+      "step": 4715
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.05425763130188,
+      "learning_rate": 4.938757113774004e-05,
+      "loss": 1.2575,
+      "step": 4720
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.3005917072296143,
+      "learning_rate": 4.938627799542235e-05,
+      "loss": 1.3646,
+      "step": 4725
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8430652618408203,
+      "learning_rate": 4.9384983506275864e-05,
+      "loss": 1.4074,
+      "step": 4730
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.666009247303009,
+      "learning_rate": 4.9383687670372094e-05,
+      "loss": 1.2717,
+      "step": 4735
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5754490494728088,
+      "learning_rate": 4.9382390487782594e-05,
+      "loss": 1.3396,
+      "step": 4740
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9268550276756287,
+      "learning_rate": 4.938109195857902e-05,
+      "loss": 1.3328,
+      "step": 4745
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.1448147296905518,
+      "learning_rate": 4.9379792082833076e-05,
+      "loss": 1.4236,
+      "step": 4750
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8515195250511169,
+      "learning_rate": 4.937849086061656e-05,
+      "loss": 1.3495,
+      "step": 4755
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.5721395015716553,
+      "learning_rate": 4.937718829200132e-05,
+      "loss": 1.5374,
+      "step": 4760
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.147706389427185,
+      "learning_rate": 4.9375884377059324e-05,
+      "loss": 1.2307,
+      "step": 4765
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7554968595504761,
+      "learning_rate": 4.937457911586256e-05,
+      "loss": 1.3832,
+      "step": 4770
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.0911380052566528,
+      "learning_rate": 4.9373272508483135e-05,
+      "loss": 1.2698,
+      "step": 4775
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6409761905670166,
+      "learning_rate": 4.93719645549932e-05,
+      "loss": 1.2276,
+      "step": 4780
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.3163679838180542,
+      "learning_rate": 4.9370655255464996e-05,
+      "loss": 1.3913,
+      "step": 4785
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.9681575298309326,
+      "learning_rate": 4.9369344609970837e-05,
+      "loss": 1.3934,
+      "step": 4790
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6123712658882141,
+      "learning_rate": 4.9368032618583106e-05,
+      "loss": 1.3078,
+      "step": 4795
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.4885551929473877,
+      "learning_rate": 4.9366719281374264e-05,
+      "loss": 1.3818,
+      "step": 4800
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8683892488479614,
+      "learning_rate": 4.936540459841684e-05,
+      "loss": 1.2567,
+      "step": 4805
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.2614054679870605,
+      "learning_rate": 4.936408856978345e-05,
+      "loss": 1.3863,
+      "step": 4810
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7793049216270447,
+      "learning_rate": 4.9362771195546767e-05,
+      "loss": 1.3391,
+      "step": 4815
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7818357944488525,
+      "learning_rate": 4.936145247577956e-05,
+      "loss": 1.425,
+      "step": 4820
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.959354877471924,
+      "learning_rate": 4.936013241055465e-05,
+      "loss": 1.4967,
+      "step": 4825
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.2867703437805176,
+      "learning_rate": 4.935881099994495e-05,
+      "loss": 1.3091,
+      "step": 4830
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.163822889328003,
+      "learning_rate": 4.935748824402344e-05,
+      "loss": 1.3299,
+      "step": 4835
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.4018052816390991,
+      "learning_rate": 4.9356164142863174e-05,
+      "loss": 1.2329,
+      "step": 4840
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.0731418132781982,
+      "learning_rate": 4.935483869653728e-05,
+      "loss": 1.448,
+      "step": 4845
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6475387811660767,
+      "learning_rate": 4.9353511905118954e-05,
+      "loss": 1.3146,
+      "step": 4850
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7880772948265076,
+      "learning_rate": 4.935218376868149e-05,
+      "loss": 1.3605,
+      "step": 4855
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6786603927612305,
+      "learning_rate": 4.9350854287298224e-05,
+      "loss": 1.2491,
+      "step": 4860
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7267608642578125,
+      "learning_rate": 4.9349523461042576e-05,
+      "loss": 1.4106,
+      "step": 4865
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.2122886180877686,
+      "learning_rate": 4.9348191289988064e-05,
+      "loss": 1.4208,
+      "step": 4870
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.0894712209701538,
+      "learning_rate": 4.934685777420827e-05,
+      "loss": 1.3931,
+      "step": 4875
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.950336217880249,
+      "learning_rate": 4.934552291377681e-05,
+      "loss": 1.4107,
+      "step": 4880
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.30669367313385,
+      "learning_rate": 4.934418670876743e-05,
+      "loss": 1.3144,
+      "step": 4885
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.539128065109253,
+      "learning_rate": 4.934284915925392e-05,
+      "loss": 1.2911,
+      "step": 4890
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.0481332540512085,
+      "learning_rate": 4.934151026531016e-05,
+      "loss": 1.3932,
+      "step": 4895
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9544585943222046,
+      "learning_rate": 4.934017002701009e-05,
+      "loss": 1.4928,
+      "step": 4900
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.628352165222168,
+      "learning_rate": 4.933882844442773e-05,
+      "loss": 1.3738,
+      "step": 4905
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.2132518291473389,
+      "learning_rate": 4.9337485517637174e-05,
+      "loss": 1.3586,
+      "step": 4910
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.3063114881515503,
+      "learning_rate": 4.9336141246712585e-05,
+      "loss": 1.2829,
+      "step": 4915
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.5061261653900146,
+      "learning_rate": 4.933479563172822e-05,
+      "loss": 1.3731,
+      "step": 4920
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.1727710962295532,
+      "learning_rate": 4.933344867275837e-05,
+      "loss": 1.5842,
+      "step": 4925
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.4313628673553467,
+      "learning_rate": 4.9332100369877457e-05,
+      "loss": 1.3292,
+      "step": 4930
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8247719407081604,
+      "learning_rate": 4.9330750723159924e-05,
+      "loss": 1.2706,
+      "step": 4935
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8011229634284973,
+      "learning_rate": 4.932939973268033e-05,
+      "loss": 1.5348,
+      "step": 4940
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.4439839124679565,
+      "learning_rate": 4.932804739851327e-05,
+      "loss": 1.5019,
+      "step": 4945
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8708484172821045,
+      "learning_rate": 4.9326693720733434e-05,
+      "loss": 1.3857,
+      "step": 4950
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.144667387008667,
+      "learning_rate": 4.93253386994156e-05,
+      "loss": 1.3401,
+      "step": 4955
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.9807694554328918,
+      "learning_rate": 4.932398233463459e-05,
+      "loss": 1.3908,
+      "step": 4960
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.2701613903045654,
+      "learning_rate": 4.932262462646532e-05,
+      "loss": 1.3377,
+      "step": 4965
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.4072773456573486,
+      "learning_rate": 4.932126557498278e-05,
+      "loss": 1.4369,
+      "step": 4970
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.901105523109436,
+      "learning_rate": 4.931990518026202e-05,
+      "loss": 1.4648,
+      "step": 4975
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.1365903615951538,
+      "learning_rate": 4.931854344237816e-05,
+      "loss": 1.4489,
+      "step": 4980
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.307646632194519,
+      "learning_rate": 4.931718036140644e-05,
+      "loss": 1.3409,
+      "step": 4985
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.375880479812622,
+      "learning_rate": 4.9315815937422124e-05,
+      "loss": 1.3656,
+      "step": 4990
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6324434876441956,
+      "learning_rate": 4.931445017050056e-05,
+      "loss": 1.4595,
+      "step": 4995
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9427233934402466,
+      "learning_rate": 4.93130830607172e-05,
+      "loss": 1.3704,
+      "step": 5000
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.6412770748138428,
+      "learning_rate": 4.931171460814753e-05,
+      "loss": 1.4681,
+      "step": 5005
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.330474615097046,
+      "learning_rate": 4.931034481286713e-05,
+      "loss": 1.3628,
+      "step": 5010
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8582301735877991,
+      "learning_rate": 4.9308973674951656e-05,
+      "loss": 1.3646,
+      "step": 5015
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6373203992843628,
+      "learning_rate": 4.9307601194476825e-05,
+      "loss": 1.2416,
+      "step": 5020
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.919718623161316,
+      "learning_rate": 4.9306227371518455e-05,
+      "loss": 1.3791,
+      "step": 5025
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8354704976081848,
+      "learning_rate": 4.9304852206152415e-05,
+      "loss": 1.393,
+      "step": 5030
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.773903489112854,
+      "learning_rate": 4.9303475698454645e-05,
+      "loss": 1.283,
+      "step": 5035
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.9701683521270752,
+      "learning_rate": 4.9302097848501176e-05,
+      "loss": 1.41,
+      "step": 5040
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.3933402299880981,
+      "learning_rate": 4.9300718656368104e-05,
+      "loss": 1.3723,
+      "step": 5045
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.408264398574829,
+      "learning_rate": 4.92993381221316e-05,
+      "loss": 1.274,
+      "step": 5050
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.3229018449783325,
+      "learning_rate": 4.929795624586791e-05,
+      "loss": 1.3723,
+      "step": 5055
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.9255488514900208,
+      "learning_rate": 4.9296573027653353e-05,
+      "loss": 1.2437,
+      "step": 5060
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.3084803819656372,
+      "learning_rate": 4.9295188467564324e-05,
+      "loss": 1.2855,
+      "step": 5065
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.0244333744049072,
+      "learning_rate": 4.9293802565677284e-05,
+      "loss": 1.3545,
+      "step": 5070
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.627091646194458,
+      "learning_rate": 4.9292415322068785e-05,
+      "loss": 1.443,
+      "step": 5075
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5619713664054871,
+      "learning_rate": 4.929102673681544e-05,
+      "loss": 1.54,
+      "step": 5080
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.0287528038024902,
+      "learning_rate": 4.928963680999393e-05,
+      "loss": 1.3472,
+      "step": 5085
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9265909194946289,
+      "learning_rate": 4.9288245541681036e-05,
+      "loss": 1.3337,
+      "step": 5090
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.3812994956970215,
+      "learning_rate": 4.9286852931953576e-05,
+      "loss": 1.4148,
+      "step": 5095
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8753069043159485,
+      "learning_rate": 4.928545898088848e-05,
+      "loss": 1.4186,
+      "step": 5100
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.826627254486084,
+      "learning_rate": 4.928406368856273e-05,
+      "loss": 1.509,
+      "step": 5105
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.941596508026123,
+      "learning_rate": 4.928266705505338e-05,
+      "loss": 1.4212,
+      "step": 5110
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.5150552988052368,
+      "learning_rate": 4.928126908043757e-05,
+      "loss": 1.3308,
+      "step": 5115
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9515051245689392,
+      "learning_rate": 4.927986976479251e-05,
+      "loss": 1.3392,
+      "step": 5120
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.2691097259521484,
+      "learning_rate": 4.927846910819548e-05,
+      "loss": 1.2719,
+      "step": 5125
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8292602896690369,
+      "learning_rate": 4.927706711072383e-05,
+      "loss": 1.2852,
+      "step": 5130
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.4237391948699951,
+      "learning_rate": 4.927566377245501e-05,
+      "loss": 1.5021,
+      "step": 5135
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7673839926719666,
+      "learning_rate": 4.92742590934665e-05,
+      "loss": 1.5058,
+      "step": 5140
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.159803867340088,
+      "learning_rate": 4.92728530738359e-05,
+      "loss": 1.3966,
+      "step": 5145
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.0266951322555542,
+      "learning_rate": 4.927144571364085e-05,
+      "loss": 1.387,
+      "step": 5150
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.0855332612991333,
+      "learning_rate": 4.927003701295909e-05,
+      "loss": 1.5594,
+      "step": 5155
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.604235053062439,
+      "learning_rate": 4.92686269718684e-05,
+      "loss": 1.4042,
+      "step": 5160
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.2728831768035889,
+      "learning_rate": 4.926721559044668e-05,
+      "loss": 1.3788,
+      "step": 5165
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.2990281581878662,
+      "learning_rate": 4.926580286877187e-05,
+      "loss": 1.3904,
+      "step": 5170
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8917083144187927,
+      "learning_rate": 4.926438880692198e-05,
+      "loss": 1.3893,
+      "step": 5175
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.6582956314086914,
+      "learning_rate": 4.9262973404975124e-05,
+      "loss": 1.3848,
+      "step": 5180
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7999622225761414,
+      "learning_rate": 4.9261556663009465e-05,
+      "loss": 1.5204,
+      "step": 5185
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.0454286336898804,
+      "learning_rate": 4.926013858110326e-05,
+      "loss": 1.4529,
+      "step": 5190
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8179815411567688,
+      "learning_rate": 4.92587191593348e-05,
+      "loss": 1.3022,
+      "step": 5195
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9816410541534424,
+      "learning_rate": 4.92572983977825e-05,
+      "loss": 1.3634,
+      "step": 5200
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7995482087135315,
+      "learning_rate": 4.925587629652483e-05,
+      "loss": 1.464,
+      "step": 5205
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.1661982536315918,
+      "learning_rate": 4.925445285564032e-05,
+      "loss": 1.5452,
+      "step": 5210
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9501356482505798,
+      "learning_rate": 4.9253028075207595e-05,
+      "loss": 1.2182,
+      "step": 5215
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.2546149492263794,
+      "learning_rate": 4.925160195530534e-05,
+      "loss": 1.5899,
+      "step": 5220
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.617793619632721,
+      "learning_rate": 4.9250174496012316e-05,
+      "loss": 1.4278,
+      "step": 5225
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8397419452667236,
+      "learning_rate": 4.9248745697407353e-05,
+      "loss": 1.2532,
+      "step": 5230
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.4492918252944946,
+      "learning_rate": 4.924731555956938e-05,
+      "loss": 1.3777,
+      "step": 5235
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.5298033952713013,
+      "learning_rate": 4.924588408257736e-05,
+      "loss": 1.3783,
+      "step": 5240
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.9098453521728516,
+      "learning_rate": 4.9244451266510384e-05,
+      "loss": 1.3461,
+      "step": 5245
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8338463306427002,
+      "learning_rate": 4.9243017111447556e-05,
+      "loss": 1.4105,
+      "step": 5250
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.729804039001465,
+      "learning_rate": 4.924158161746809e-05,
+      "loss": 1.491,
+      "step": 5255
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.264284610748291,
+      "learning_rate": 4.9240144784651265e-05,
+      "loss": 1.3767,
+      "step": 5260
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7262938022613525,
+      "learning_rate": 4.923870661307645e-05,
+      "loss": 1.3735,
+      "step": 5265
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7457590103149414,
+      "learning_rate": 4.923726710282305e-05,
+      "loss": 1.485,
+      "step": 5270
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.9725492000579834,
+      "learning_rate": 4.923582625397059e-05,
+      "loss": 1.393,
+      "step": 5275
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.8428640365600586,
+      "learning_rate": 4.923438406659864e-05,
+      "loss": 1.3126,
+      "step": 5280
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.4272032976150513,
+      "learning_rate": 4.923294054078684e-05,
+      "loss": 1.3194,
+      "step": 5285
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.783457338809967,
+      "learning_rate": 4.923149567661492e-05,
+      "loss": 1.3534,
+      "step": 5290
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.0013182163238525,
+      "learning_rate": 4.9230049474162697e-05,
+      "loss": 1.3938,
+      "step": 5295
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.9497959017753601,
+      "learning_rate": 4.922860193351002e-05,
+      "loss": 1.4592,
+      "step": 5300
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.3006082773208618,
+      "learning_rate": 4.922715305473684e-05,
+      "loss": 1.3387,
+      "step": 5305
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.0174909830093384,
+      "learning_rate": 4.922570283792318e-05,
+      "loss": 1.3116,
+      "step": 5310
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.6044186353683472,
+      "learning_rate": 4.9224251283149136e-05,
+      "loss": 1.4901,
+      "step": 5315
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.9575896859169006,
+      "learning_rate": 4.9222798390494874e-05,
+      "loss": 1.549,
+      "step": 5320
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8215287923812866,
+      "learning_rate": 4.922134416004063e-05,
+      "loss": 1.2743,
+      "step": 5325
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.3678843975067139,
+      "learning_rate": 4.9219888591866725e-05,
+      "loss": 1.3831,
+      "step": 5330
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.1509759426116943,
+      "learning_rate": 4.921843168605355e-05,
+      "loss": 1.4874,
+      "step": 5335
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6248118281364441,
+      "learning_rate": 4.921697344268157e-05,
+      "loss": 1.4002,
+      "step": 5340
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9303049445152283,
+      "learning_rate": 4.9215513861831316e-05,
+      "loss": 1.4764,
+      "step": 5345
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.2478617429733276,
+      "learning_rate": 4.92140529435834e-05,
+      "loss": 1.3353,
+      "step": 5350
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.1409193277359009,
+      "learning_rate": 4.921259068801851e-05,
+      "loss": 1.4389,
+      "step": 5355
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.311739921569824,
+      "learning_rate": 4.921112709521741e-05,
+      "loss": 1.456,
+      "step": 5360
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.399682879447937,
+      "learning_rate": 4.920966216526092e-05,
+      "loss": 1.3554,
+      "step": 5365
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.2935621738433838,
+      "learning_rate": 4.920819589822995e-05,
+      "loss": 1.2121,
+      "step": 5370
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8797827363014221,
+      "learning_rate": 4.92067282942055e-05,
+      "loss": 1.2156,
+      "step": 5375
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6572903990745544,
+      "learning_rate": 4.9205259353268585e-05,
+      "loss": 1.4867,
+      "step": 5380
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.5460227727890015,
+      "learning_rate": 4.920378907550037e-05,
+      "loss": 1.4242,
+      "step": 5385
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.2212427854537964,
+      "learning_rate": 4.9202317460982037e-05,
+      "loss": 1.3154,
+      "step": 5390
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.2694352865219116,
+      "learning_rate": 4.9200844509794876e-05,
+      "loss": 1.3032,
+      "step": 5395
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9184812903404236,
+      "learning_rate": 4.919937022202022e-05,
+      "loss": 1.3117,
+      "step": 5400
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.309106469154358,
+      "learning_rate": 4.9197894597739505e-05,
+      "loss": 1.3399,
+      "step": 5405
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.238127589225769,
+      "learning_rate": 4.919641763703422e-05,
+      "loss": 1.3066,
+      "step": 5410
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8524130582809448,
+      "learning_rate": 4.919493933998594e-05,
+      "loss": 1.5195,
+      "step": 5415
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.081799030303955,
+      "learning_rate": 4.919345970667631e-05,
+      "loss": 1.2854,
+      "step": 5420
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8055438995361328,
+      "learning_rate": 4.919197873718705e-05,
+      "loss": 1.1592,
+      "step": 5425
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.1697745323181152,
+      "learning_rate": 4.919049643159995e-05,
+      "loss": 1.3942,
+      "step": 5430
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.0526283979415894,
+      "learning_rate": 4.918901278999687e-05,
+      "loss": 1.4919,
+      "step": 5435
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.9364911913871765,
+      "learning_rate": 4.918752781245976e-05,
+      "loss": 1.2354,
+      "step": 5440
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.036669135093689,
+      "learning_rate": 4.918604149907064e-05,
+      "loss": 1.4115,
+      "step": 5445
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.323757529258728,
+      "learning_rate": 4.918455384991159e-05,
+      "loss": 1.4734,
+      "step": 5450
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.043481707572937,
+      "learning_rate": 4.9183064865064756e-05,
+      "loss": 1.3592,
+      "step": 5455
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.0515973567962646,
+      "learning_rate": 4.918157454461238e-05,
+      "loss": 1.3578,
+      "step": 5460
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.5430241823196411,
+      "learning_rate": 4.918008288863679e-05,
+      "loss": 1.5445,
+      "step": 5465
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.392940640449524,
+      "learning_rate": 4.917858989722036e-05,
+      "loss": 1.3484,
+      "step": 5470
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8935360312461853,
+      "learning_rate": 4.917709557044553e-05,
+      "loss": 1.35,
+      "step": 5475
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7475391626358032,
+      "learning_rate": 4.9175599908394854e-05,
+      "loss": 1.331,
+      "step": 5480
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.0226927995681763,
+      "learning_rate": 4.917410291115092e-05,
+      "loss": 1.4033,
+      "step": 5485
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.5844924449920654,
+      "learning_rate": 4.917260457879641e-05,
+      "loss": 1.3624,
+      "step": 5490
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.3369362354278564,
+      "learning_rate": 4.917110491141407e-05,
+      "loss": 1.4102,
+      "step": 5495
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.9585330486297607,
+      "learning_rate": 4.9169603909086736e-05,
+      "loss": 1.3082,
+      "step": 5500
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.9265090227127075,
+      "learning_rate": 4.91681015718973e-05,
+      "loss": 1.3516,
+      "step": 5505
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8970497250556946,
+      "learning_rate": 4.9166597899928735e-05,
+      "loss": 1.2882,
+      "step": 5510
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.3045713901519775,
+      "learning_rate": 4.9165092893264086e-05,
+      "loss": 1.4484,
+      "step": 5515
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7835524082183838,
+      "learning_rate": 4.9163586551986475e-05,
+      "loss": 1.4134,
+      "step": 5520
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7851400971412659,
+      "learning_rate": 4.9162078876179095e-05,
+      "loss": 1.1785,
+      "step": 5525
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5514227747917175,
+      "learning_rate": 4.916056986592522e-05,
+      "loss": 1.2639,
+      "step": 5530
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.4473460912704468,
+      "learning_rate": 4.915905952130818e-05,
+      "loss": 1.3462,
+      "step": 5535
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.3507695198059082,
+      "learning_rate": 4.9157547842411387e-05,
+      "loss": 1.2437,
+      "step": 5540
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.2776063680648804,
+      "learning_rate": 4.9156034829318345e-05,
+      "loss": 1.3707,
+      "step": 5545
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.255897045135498,
+      "learning_rate": 4.915452048211261e-05,
+      "loss": 1.2916,
+      "step": 5550
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.0228066444396973,
+      "learning_rate": 4.915300480087781e-05,
+      "loss": 1.4058,
+      "step": 5555
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.0739355087280273,
+      "learning_rate": 4.915148778569767e-05,
+      "loss": 1.4693,
+      "step": 5560
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.723423719406128,
+      "learning_rate": 4.914996943665596e-05,
+      "loss": 1.0389,
+      "step": 5565
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.2610810995101929,
+      "learning_rate": 4.9148449753836534e-05,
+      "loss": 1.3547,
+      "step": 5570
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.9140579700469971,
+      "learning_rate": 4.9146928737323334e-05,
+      "loss": 1.3006,
+      "step": 5575
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7020224332809448,
+      "learning_rate": 4.914540638720035e-05,
+      "loss": 1.3814,
+      "step": 5580
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6752337217330933,
+      "learning_rate": 4.9143882703551685e-05,
+      "loss": 1.3788,
+      "step": 5585
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.1528271436691284,
+      "learning_rate": 4.914235768646147e-05,
+      "loss": 1.4718,
+      "step": 5590
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.9976640343666077,
+      "learning_rate": 4.9140831336013925e-05,
+      "loss": 1.4391,
+      "step": 5595
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7416092157363892,
+      "learning_rate": 4.9139303652293365e-05,
+      "loss": 1.2846,
+      "step": 5600
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8383505940437317,
+      "learning_rate": 4.913777463538416e-05,
+      "loss": 1.3933,
+      "step": 5605
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7380107045173645,
+      "learning_rate": 4.9136244285370746e-05,
+      "loss": 1.4613,
+      "step": 5610
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8241648077964783,
+      "learning_rate": 4.913471260233765e-05,
+      "loss": 1.3475,
+      "step": 5615
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.6132783889770508,
+      "learning_rate": 4.913317958636946e-05,
+      "loss": 1.4238,
+      "step": 5620
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8694246411323547,
+      "learning_rate": 4.913164523755085e-05,
+      "loss": 1.6346,
+      "step": 5625
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.2814140319824219,
+      "learning_rate": 4.9130109555966565e-05,
+      "loss": 1.3782,
+      "step": 5630
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.7262847423553467,
+      "learning_rate": 4.91285725417014e-05,
+      "loss": 1.3747,
+      "step": 5635
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8753448128700256,
+      "learning_rate": 4.912703419484026e-05,
+      "loss": 1.3918,
+      "step": 5640
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.002820611000061,
+      "learning_rate": 4.912549451546809e-05,
+      "loss": 1.2981,
+      "step": 5645
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.3369395732879639,
+      "learning_rate": 4.912395350366994e-05,
+      "loss": 1.3446,
+      "step": 5650
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.4645837545394897,
+      "learning_rate": 4.9122411159530916e-05,
+      "loss": 1.5831,
+      "step": 5655
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.1211307048797607,
+      "learning_rate": 4.91208674831362e-05,
+      "loss": 1.3518,
+      "step": 5660
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.3223671913146973,
+      "learning_rate": 4.911932247457104e-05,
+      "loss": 1.4612,
+      "step": 5665
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.5868595838546753,
+      "learning_rate": 4.911777613392077e-05,
+      "loss": 1.2637,
+      "step": 5670
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.2033578157424927,
+      "learning_rate": 4.91162284612708e-05,
+      "loss": 1.1865,
+      "step": 5675
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.1434866189956665,
+      "learning_rate": 4.9114679456706594e-05,
+      "loss": 1.5031,
+      "step": 5680
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.0561349391937256,
+      "learning_rate": 4.911312912031371e-05,
+      "loss": 1.3647,
+      "step": 5685
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.2689900398254395,
+      "learning_rate": 4.911157745217776e-05,
+      "loss": 1.4179,
+      "step": 5690
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9906610250473022,
+      "learning_rate": 4.911002445238446e-05,
+      "loss": 1.4319,
+      "step": 5695
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.0266451835632324,
+      "learning_rate": 4.9108470121019565e-05,
+      "loss": 1.2428,
+      "step": 5700
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.1348594427108765,
+      "learning_rate": 4.910691445816893e-05,
+      "loss": 1.3954,
+      "step": 5705
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.3033210039138794,
+      "learning_rate": 4.910535746391846e-05,
+      "loss": 1.3513,
+      "step": 5710
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.4596507549285889,
+      "learning_rate": 4.910379913835416e-05,
+      "loss": 1.4849,
+      "step": 5715
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.2325527667999268,
+      "learning_rate": 4.910223948156208e-05,
+      "loss": 1.231,
+      "step": 5720
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.0921927690505981,
+      "learning_rate": 4.9100678493628374e-05,
+      "loss": 1.4783,
+      "step": 5725
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5836876034736633,
+      "learning_rate": 4.909911617463925e-05,
+      "loss": 1.3841,
+      "step": 5730
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6414868831634521,
+      "learning_rate": 4.909755252468098e-05,
+      "loss": 1.414,
+      "step": 5735
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.1332942247390747,
+      "learning_rate": 4.909598754383994e-05,
+      "loss": 1.364,
+      "step": 5740
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.3707212209701538,
+      "learning_rate": 4.909442123220255e-05,
+      "loss": 1.3071,
+      "step": 5745
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9404252171516418,
+      "learning_rate": 4.909285358985532e-05,
+      "loss": 1.332,
+      "step": 5750
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9168292880058289,
+      "learning_rate": 4.9091284616884824e-05,
+      "loss": 1.313,
+      "step": 5755
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.628679633140564,
+      "learning_rate": 4.908971431337772e-05,
+      "loss": 1.3079,
+      "step": 5760
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.0647614002227783,
+      "learning_rate": 4.908814267942074e-05,
+      "loss": 1.2487,
+      "step": 5765
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.053024411201477,
+      "learning_rate": 4.908656971510068e-05,
+      "loss": 1.3222,
+      "step": 5770
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.9378986358642578,
+      "learning_rate": 4.908499542050441e-05,
+      "loss": 1.3492,
+      "step": 5775
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.9233798980712891,
+      "learning_rate": 4.9083419795718875e-05,
+      "loss": 1.4543,
+      "step": 5780
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.063833236694336,
+      "learning_rate": 4.9081842840831104e-05,
+      "loss": 1.3152,
+      "step": 5785
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.2852481603622437,
+      "learning_rate": 4.908026455592818e-05,
+      "loss": 1.2847,
+      "step": 5790
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.1695432662963867,
+      "learning_rate": 4.9078684941097275e-05,
+      "loss": 1.2973,
+      "step": 5795
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.908960223197937,
+      "learning_rate": 4.9077103996425625e-05,
+      "loss": 1.1404,
+      "step": 5800
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.2946605682373047,
+      "learning_rate": 4.9075521722000556e-05,
+      "loss": 1.3848,
+      "step": 5805
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6911953091621399,
+      "learning_rate": 4.907393811790945e-05,
+      "loss": 1.4204,
+      "step": 5810
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.9933083653450012,
+      "learning_rate": 4.907235318423975e-05,
+      "loss": 1.4721,
+      "step": 5815
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.4515044689178467,
+      "learning_rate": 4.9070766921079014e-05,
+      "loss": 1.3615,
+      "step": 5820
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.067295789718628,
+      "learning_rate": 4.906917932851484e-05,
+      "loss": 1.3254,
+      "step": 5825
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.1964728832244873,
+      "learning_rate": 4.9067590406634914e-05,
+      "loss": 1.2453,
+      "step": 5830
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9193742275238037,
+      "learning_rate": 4.906600015552698e-05,
+      "loss": 1.5016,
+      "step": 5835
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.4056816101074219,
+      "learning_rate": 4.906440857527888e-05,
+      "loss": 1.2901,
+      "step": 5840
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7364879846572876,
+      "learning_rate": 4.9062815665978504e-05,
+      "loss": 1.3488,
+      "step": 5845
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.1966763734817505,
+      "learning_rate": 4.9061221427713835e-05,
+      "loss": 1.3019,
+      "step": 5850
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.1406923532485962,
+      "learning_rate": 4.905962586057291e-05,
+      "loss": 1.27,
+      "step": 5855
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.5411876440048218,
+      "learning_rate": 4.9058028964643865e-05,
+      "loss": 1.2951,
+      "step": 5860
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9900470972061157,
+      "learning_rate": 4.9056430740014883e-05,
+      "loss": 1.3227,
+      "step": 5865
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.95987868309021,
+      "learning_rate": 4.905483118677423e-05,
+      "loss": 1.4196,
+      "step": 5870
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0949276685714722,
+      "learning_rate": 4.9053230305010264e-05,
+      "loss": 1.2241,
+      "step": 5875
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6950658559799194,
+      "learning_rate": 4.9051628094811386e-05,
+      "loss": 1.2839,
+      "step": 5880
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.1980772018432617,
+      "learning_rate": 4.905002455626609e-05,
+      "loss": 1.458,
+      "step": 5885
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.1213877201080322,
+      "learning_rate": 4.904841968946293e-05,
+      "loss": 1.4148,
+      "step": 5890
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.1307765245437622,
+      "learning_rate": 4.904681349449056e-05,
+      "loss": 1.4134,
+      "step": 5895
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.358079433441162,
+      "learning_rate": 4.904520597143767e-05,
+      "loss": 1.3841,
+      "step": 5900
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.1629695892333984,
+      "learning_rate": 4.904359712039304e-05,
+      "loss": 1.39,
+      "step": 5905
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.8049236536026,
+      "learning_rate": 4.904198694144554e-05,
+      "loss": 1.3112,
+      "step": 5910
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5202418565750122,
+      "learning_rate": 4.904037543468409e-05,
+      "loss": 1.2515,
+      "step": 5915
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.6032389402389526,
+      "learning_rate": 4.90387626001977e-05,
+      "loss": 1.4008,
+      "step": 5920
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.2656590938568115,
+      "learning_rate": 4.903714843807543e-05,
+      "loss": 1.4304,
+      "step": 5925
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.622446894645691,
+      "learning_rate": 4.9035532948406436e-05,
+      "loss": 1.4896,
+      "step": 5930
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.3047736883163452,
+      "learning_rate": 4.903391613127995e-05,
+      "loss": 1.3628,
+      "step": 5935
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8946179747581482,
+      "learning_rate": 4.903229798678525e-05,
+      "loss": 1.3129,
+      "step": 5940
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7396437525749207,
+      "learning_rate": 4.903067851501172e-05,
+      "loss": 1.4662,
+      "step": 5945
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7600730061531067,
+      "learning_rate": 4.9029057716048786e-05,
+      "loss": 1.3341,
+      "step": 5950
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.3639189004898071,
+      "learning_rate": 4.902743558998597e-05,
+      "loss": 1.4028,
+      "step": 5955
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2151553630828857,
+      "learning_rate": 4.9025812136912874e-05,
+      "loss": 1.2892,
+      "step": 5960
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.493870496749878,
+      "learning_rate": 4.902418735691914e-05,
+      "loss": 1.3913,
+      "step": 5965
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.329575538635254,
+      "learning_rate": 4.90225612500945e-05,
+      "loss": 1.3196,
+      "step": 5970
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.2628295421600342,
+      "learning_rate": 4.9020933816528784e-05,
+      "loss": 1.3211,
+      "step": 5975
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.0996872186660767,
+      "learning_rate": 4.901930505631186e-05,
+      "loss": 1.402,
+      "step": 5980
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.5428398251533508,
+      "learning_rate": 4.901767496953368e-05,
+      "loss": 1.5071,
+      "step": 5985
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6966472864151001,
+      "learning_rate": 4.9016043556284284e-05,
+      "loss": 1.4231,
+      "step": 5990
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.003456473350525,
+      "learning_rate": 4.901441081665376e-05,
+      "loss": 1.2952,
+      "step": 5995
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.4980223178863525,
+      "learning_rate": 4.9012776750732295e-05,
+      "loss": 1.2411,
+      "step": 6000
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0407806634902954,
+      "learning_rate": 4.901114135861012e-05,
+      "loss": 1.3536,
+      "step": 6005
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7962636947631836,
+      "learning_rate": 4.9009504640377565e-05,
+      "loss": 1.4762,
+      "step": 6010
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.4784387350082397,
+      "learning_rate": 4.900786659612504e-05,
+      "loss": 1.3125,
+      "step": 6015
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.9292671084403992,
+      "learning_rate": 4.900622722594299e-05,
+      "loss": 1.4477,
+      "step": 6020
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.320400357246399,
+      "learning_rate": 4.9004586529921955e-05,
+      "loss": 1.3478,
+      "step": 6025
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.593423068523407,
+      "learning_rate": 4.9002944508152567e-05,
+      "loss": 1.2153,
+      "step": 6030
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.3448033332824707,
+      "learning_rate": 4.90013011607255e-05,
+      "loss": 1.4262,
+      "step": 6035
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.65395987033844,
+      "learning_rate": 4.8999656487731516e-05,
+      "loss": 1.4308,
+      "step": 6040
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.5153121948242188,
+      "learning_rate": 4.899801048926146e-05,
+      "loss": 1.3404,
+      "step": 6045
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.942297637462616,
+      "learning_rate": 4.899636316540622e-05,
+      "loss": 1.4883,
+      "step": 6050
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.9569664597511292,
+      "learning_rate": 4.899471451625678e-05,
+      "loss": 1.3667,
+      "step": 6055
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.1966135501861572,
+      "learning_rate": 4.899306454190421e-05,
+      "loss": 1.1763,
+      "step": 6060
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.0947861671447754,
+      "learning_rate": 4.899141324243962e-05,
+      "loss": 1.2947,
+      "step": 6065
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9983508586883545,
+      "learning_rate": 4.8989760617954215e-05,
+      "loss": 1.7462,
+      "step": 6070
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8864436745643616,
+      "learning_rate": 4.898810666853927e-05,
+      "loss": 1.526,
+      "step": 6075
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.3127366304397583,
+      "learning_rate": 4.898645139428613e-05,
+      "loss": 1.3631,
+      "step": 6080
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9836210608482361,
+      "learning_rate": 4.8984794795286196e-05,
+      "loss": 1.2838,
+      "step": 6085
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.2298458814620972,
+      "learning_rate": 4.8983136871630995e-05,
+      "loss": 1.3527,
+      "step": 6090
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.1192662715911865,
+      "learning_rate": 4.8981477623412064e-05,
+      "loss": 1.4846,
+      "step": 6095
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.9733280539512634,
+      "learning_rate": 4.897981705072105e-05,
+      "loss": 1.4492,
+      "step": 6100
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.5876336097717285,
+      "learning_rate": 4.897815515364967e-05,
+      "loss": 1.3106,
+      "step": 6105
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.3915202617645264,
+      "learning_rate": 4.89764919322897e-05,
+      "loss": 1.2493,
+      "step": 6110
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.1110719442367554,
+      "learning_rate": 4.897482738673301e-05,
+      "loss": 1.5083,
+      "step": 6115
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.9961562752723694,
+      "learning_rate": 4.897316151707152e-05,
+      "loss": 1.5106,
+      "step": 6120
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7797302603721619,
+      "learning_rate": 4.897149432339724e-05,
+      "loss": 1.3252,
+      "step": 6125
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.6558862924575806,
+      "learning_rate": 4.896982580580224e-05,
+      "loss": 1.5234,
+      "step": 6130
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8228052854537964,
+      "learning_rate": 4.896815596437868e-05,
+      "loss": 1.3884,
+      "step": 6135
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7382698059082031,
+      "learning_rate": 4.896648479921878e-05,
+      "loss": 1.3653,
+      "step": 6140
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.1932933330535889,
+      "learning_rate": 4.896481231041483e-05,
+      "loss": 1.6213,
+      "step": 6145
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.208572506904602,
+      "learning_rate": 4.896313849805921e-05,
+      "loss": 1.4082,
+      "step": 6150
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.3084590435028076,
+      "learning_rate": 4.896146336224436e-05,
+      "loss": 1.5231,
+      "step": 6155
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.2458943128585815,
+      "learning_rate": 4.89597869030628e-05,
+      "loss": 1.2657,
+      "step": 6160
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.221347689628601,
+      "learning_rate": 4.8958109120607117e-05,
+      "loss": 1.4708,
+      "step": 6165
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9371591210365295,
+      "learning_rate": 4.895643001496996e-05,
+      "loss": 1.3741,
+      "step": 6170
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7532001733779907,
+      "learning_rate": 4.8954749586244074e-05,
+      "loss": 1.3621,
+      "step": 6175
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.8896582126617432,
+      "learning_rate": 4.895306783452228e-05,
+      "loss": 1.398,
+      "step": 6180
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.096148133277893,
+      "learning_rate": 4.895138475989743e-05,
+      "loss": 1.2032,
+      "step": 6185
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7891479730606079,
+      "learning_rate": 4.894970036246251e-05,
+      "loss": 1.3164,
+      "step": 6190
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9778878092765808,
+      "learning_rate": 4.894801464231053e-05,
+      "loss": 1.4422,
+      "step": 6195
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.088839054107666,
+      "learning_rate": 4.894632759953459e-05,
+      "loss": 1.3869,
+      "step": 6200
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.0717918872833252,
+      "learning_rate": 4.894463923422787e-05,
+      "loss": 1.3477,
+      "step": 6205
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.6605066061019897,
+      "learning_rate": 4.894294954648362e-05,
+      "loss": 1.2396,
+      "step": 6210
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.642807126045227,
+      "learning_rate": 4.894125853639514e-05,
+      "loss": 1.3233,
+      "step": 6215
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.835897207260132,
+      "learning_rate": 4.893956620405585e-05,
+      "loss": 1.4308,
+      "step": 6220
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6643049120903015,
+      "learning_rate": 4.893787254955919e-05,
+      "loss": 1.5178,
+      "step": 6225
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7968789935112,
+      "learning_rate": 4.893617757299872e-05,
+      "loss": 1.2129,
+      "step": 6230
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.484046459197998,
+      "learning_rate": 4.893448127446805e-05,
+      "loss": 1.2537,
+      "step": 6235
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.99761962890625,
+      "learning_rate": 4.8932783654060844e-05,
+      "loss": 1.3536,
+      "step": 6240
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.965552806854248,
+      "learning_rate": 4.8931084711870876e-05,
+      "loss": 1.3365,
+      "step": 6245
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8811823129653931,
+      "learning_rate": 4.8929384447991974e-05,
+      "loss": 1.4075,
+      "step": 6250
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.0002772808074951,
+      "learning_rate": 4.8927682862518044e-05,
+      "loss": 1.4315,
+      "step": 6255
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.4986729919910431,
+      "learning_rate": 4.8925979955543067e-05,
+      "loss": 1.1747,
+      "step": 6260
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.492148995399475,
+      "learning_rate": 4.892427572716108e-05,
+      "loss": 1.3318,
+      "step": 6265
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.448102593421936,
+      "learning_rate": 4.892257017746621e-05,
+      "loss": 1.2832,
+      "step": 6270
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.4307514429092407,
+      "learning_rate": 4.892086330655266e-05,
+      "loss": 1.5149,
+      "step": 6275
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8795152306556702,
+      "learning_rate": 4.8919155114514695e-05,
+      "loss": 1.4482,
+      "step": 6280
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9869243502616882,
+      "learning_rate": 4.891744560144666e-05,
+      "loss": 1.3061,
+      "step": 6285
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6511804461479187,
+      "learning_rate": 4.891573476744295e-05,
+      "loss": 1.4904,
+      "step": 6290
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.5018410086631775,
+      "learning_rate": 4.891402261259807e-05,
+      "loss": 1.374,
+      "step": 6295
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.5579232573509216,
+      "learning_rate": 4.891230913700659e-05,
+      "loss": 1.2526,
+      "step": 6300
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.1983088254928589,
+      "learning_rate": 4.8910594340763126e-05,
+      "loss": 1.5322,
+      "step": 6305
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.0148495435714722,
+      "learning_rate": 4.89088782239624e-05,
+      "loss": 1.4543,
+      "step": 6310
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.739857017993927,
+      "learning_rate": 4.890716078669917e-05,
+      "loss": 1.373,
+      "step": 6315
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.4641748666763306,
+      "learning_rate": 4.8905442029068296e-05,
+      "loss": 1.3945,
+      "step": 6320
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.2002661228179932,
+      "learning_rate": 4.890372195116471e-05,
+      "loss": 1.3935,
+      "step": 6325
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.9854932427406311,
+      "learning_rate": 4.890200055308342e-05,
+      "loss": 1.4561,
+      "step": 6330
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8449004888534546,
+      "learning_rate": 4.8900277834919475e-05,
+      "loss": 1.3978,
+      "step": 6335
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.8262367248535156,
+      "learning_rate": 4.889855379676802e-05,
+      "loss": 1.3593,
+      "step": 6340
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.1936458349227905,
+      "learning_rate": 4.889682843872429e-05,
+      "loss": 1.3747,
+      "step": 6345
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.5638678073883057,
+      "learning_rate": 4.8895101760883566e-05,
+      "loss": 1.2618,
+      "step": 6350
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.084822654724121,
+      "learning_rate": 4.88933737633412e-05,
+      "loss": 1.4142,
+      "step": 6355
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.615086793899536,
+      "learning_rate": 4.889164444619264e-05,
+      "loss": 1.3423,
+      "step": 6360
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.1368343830108643,
+      "learning_rate": 4.88899138095334e-05,
+      "loss": 1.325,
+      "step": 6365
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.9688156247138977,
+      "learning_rate": 4.8888181853459046e-05,
+      "loss": 1.3005,
+      "step": 6370
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.1333907842636108,
+      "learning_rate": 4.8886448578065236e-05,
+      "loss": 1.5213,
+      "step": 6375
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.5067886114120483,
+      "learning_rate": 4.8884713983447704e-05,
+      "loss": 1.4821,
+      "step": 6380
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.2759270668029785,
+      "learning_rate": 4.8882978069702246e-05,
+      "loss": 1.2604,
+      "step": 6385
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.2852680683135986,
+      "learning_rate": 4.888124083692473e-05,
+      "loss": 1.4365,
+      "step": 6390
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.2797343730926514,
+      "learning_rate": 4.887950228521111e-05,
+      "loss": 1.3997,
+      "step": 6395
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6342377662658691,
+      "learning_rate": 4.8877762414657394e-05,
+      "loss": 1.3325,
+      "step": 6400
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.4419025182724,
+      "learning_rate": 4.8876021225359684e-05,
+      "loss": 1.3404,
+      "step": 6405
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.9858295321464539,
+      "learning_rate": 4.887427871741414e-05,
+      "loss": 1.3812,
+      "step": 6410
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.2529702186584473,
+      "learning_rate": 4.8872534890916996e-05,
+      "loss": 1.3797,
+      "step": 6415
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.7771295309066772,
+      "learning_rate": 4.8870789745964566e-05,
+      "loss": 1.3254,
+      "step": 6420
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7619970440864563,
+      "learning_rate": 4.8869043282653234e-05,
+      "loss": 1.3069,
+      "step": 6425
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.5197542905807495,
+      "learning_rate": 4.886729550107945e-05,
+      "loss": 1.4575,
+      "step": 6430
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.9382418990135193,
+      "learning_rate": 4.8865546401339736e-05,
+      "loss": 1.4566,
+      "step": 6435
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7788457274436951,
+      "learning_rate": 4.886379598353071e-05,
+      "loss": 1.242,
+      "step": 6440
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.4081120491027832,
+      "learning_rate": 4.8862044247749034e-05,
+      "loss": 1.4399,
+      "step": 6445
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1004774570465088,
+      "learning_rate": 4.886029119409146e-05,
+      "loss": 1.4171,
+      "step": 6450
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.2400892972946167,
+      "learning_rate": 4.88585368226548e-05,
+      "loss": 1.3873,
+      "step": 6455
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6717667579650879,
+      "learning_rate": 4.8856781133535955e-05,
+      "loss": 1.1926,
+      "step": 6460
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7015379667282104,
+      "learning_rate": 4.8855024126831886e-05,
+      "loss": 1.2202,
+      "step": 6465
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.3034001588821411,
+      "learning_rate": 4.8853265802639625e-05,
+      "loss": 1.2759,
+      "step": 6470
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.2087301015853882,
+      "learning_rate": 4.8851506161056296e-05,
+      "loss": 1.1914,
+      "step": 6475
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8940001726150513,
+      "learning_rate": 4.8849745202179064e-05,
+      "loss": 1.4391,
+      "step": 6480
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.4564568996429443,
+      "learning_rate": 4.8847982926105195e-05,
+      "loss": 1.4835,
+      "step": 6485
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.3092164993286133,
+      "learning_rate": 4.884621933293203e-05,
+      "loss": 1.3326,
+      "step": 6490
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.8018982410430908,
+      "learning_rate": 4.8844454422756946e-05,
+      "loss": 1.4825,
+      "step": 6495
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8499402403831482,
+      "learning_rate": 4.884268819567743e-05,
+      "loss": 1.38,
+      "step": 6500
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.2163305282592773,
+      "learning_rate": 4.8840920651791036e-05,
+      "loss": 1.2982,
+      "step": 6505
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5904464721679688,
+      "learning_rate": 4.883915179119537e-05,
+      "loss": 1.3852,
+      "step": 6510
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.2518161535263062,
+      "learning_rate": 4.883738161398813e-05,
+      "loss": 1.4104,
+      "step": 6515
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.155625581741333,
+      "learning_rate": 4.883561012026708e-05,
+      "loss": 1.3967,
+      "step": 6520
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6556622982025146,
+      "learning_rate": 4.883383731013007e-05,
+      "loss": 1.3625,
+      "step": 6525
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9793677926063538,
+      "learning_rate": 4.883206318367499e-05,
+      "loss": 1.4226,
+      "step": 6530
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.0961387157440186,
+      "learning_rate": 4.883028774099983e-05,
+      "loss": 1.4849,
+      "step": 6535
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9702796936035156,
+      "learning_rate": 4.882851098220265e-05,
+      "loss": 1.3409,
+      "step": 6540
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.4797945022583008,
+      "learning_rate": 4.882673290738158e-05,
+      "loss": 1.3482,
+      "step": 6545
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1663824319839478,
+      "learning_rate": 4.8824953516634816e-05,
+      "loss": 1.2391,
+      "step": 6550
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7633179426193237,
+      "learning_rate": 4.882317281006064e-05,
+      "loss": 1.2115,
+      "step": 6555
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6997314095497131,
+      "learning_rate": 4.8821390787757384e-05,
+      "loss": 1.3422,
+      "step": 6560
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.2793208360671997,
+      "learning_rate": 4.8819607449823476e-05,
+      "loss": 1.4855,
+      "step": 6565
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.2282752990722656,
+      "learning_rate": 4.881782279635741e-05,
+      "loss": 1.5351,
+      "step": 6570
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.7947735786437988,
+      "learning_rate": 4.8816036827457745e-05,
+      "loss": 1.2795,
+      "step": 6575
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.136925458908081,
+      "learning_rate": 4.8814249543223125e-05,
+      "loss": 1.3417,
+      "step": 6580
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9631902575492859,
+      "learning_rate": 4.8812460943752256e-05,
+      "loss": 1.4113,
+      "step": 6585
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7148537039756775,
+      "learning_rate": 4.881067102914392e-05,
+      "loss": 1.3135,
+      "step": 6590
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.7874670028686523,
+      "learning_rate": 4.880887979949698e-05,
+      "loss": 1.3566,
+      "step": 6595
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9266506433486938,
+      "learning_rate": 4.8807087254910344e-05,
+      "loss": 1.3909,
+      "step": 6600
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.1458144187927246,
+      "learning_rate": 4.880529339548303e-05,
+      "loss": 1.4995,
+      "step": 6605
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6967890858650208,
+      "learning_rate": 4.8803498221314106e-05,
+      "loss": 1.5207,
+      "step": 6610
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.142874836921692,
+      "learning_rate": 4.880170173250272e-05,
+      "loss": 1.335,
+      "step": 6615
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.030205726623535,
+      "learning_rate": 4.879990392914809e-05,
+      "loss": 1.2267,
+      "step": 6620
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.5845741033554077,
+      "learning_rate": 4.8798104811349496e-05,
+      "loss": 1.4528,
+      "step": 6625
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.0889317989349365,
+      "learning_rate": 4.879630437920631e-05,
+      "loss": 1.3085,
+      "step": 6630
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.1721889972686768,
+      "learning_rate": 4.8794502632817983e-05,
+      "loss": 1.3531,
+      "step": 6635
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.2524924278259277,
+      "learning_rate": 4.8792699572283996e-05,
+      "loss": 1.482,
+      "step": 6640
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.9633756875991821,
+      "learning_rate": 4.879089519770395e-05,
+      "loss": 1.3769,
+      "step": 6645
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.5347559452056885,
+      "learning_rate": 4.8789089509177485e-05,
+      "loss": 1.3181,
+      "step": 6650
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.992938220500946,
+      "learning_rate": 4.878728250680433e-05,
+      "loss": 1.376,
+      "step": 6655
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.4422589540481567,
+      "learning_rate": 4.87854741906843e-05,
+      "loss": 1.2264,
+      "step": 6660
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.2222764492034912,
+      "learning_rate": 4.878366456091724e-05,
+      "loss": 1.3726,
+      "step": 6665
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5847399830818176,
+      "learning_rate": 4.8781853617603116e-05,
+      "loss": 1.3855,
+      "step": 6670
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1310397386550903,
+      "learning_rate": 4.878004136084194e-05,
+      "loss": 1.4062,
+      "step": 6675
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.0542705059051514,
+      "learning_rate": 4.877822779073379e-05,
+      "loss": 1.641,
+      "step": 6680
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.2997204065322876,
+      "learning_rate": 4.877641290737884e-05,
+      "loss": 1.4196,
+      "step": 6685
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9382727146148682,
+      "learning_rate": 4.8774596710877315e-05,
+      "loss": 1.3389,
+      "step": 6690
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9488927125930786,
+      "learning_rate": 4.877277920132953e-05,
+      "loss": 1.1815,
+      "step": 6695
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7831886410713196,
+      "learning_rate": 4.877096037883586e-05,
+      "loss": 1.2372,
+      "step": 6700
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.236717939376831,
+      "learning_rate": 4.876914024349676e-05,
+      "loss": 1.3009,
+      "step": 6705
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.9065530896186829,
+      "learning_rate": 4.8767318795412746e-05,
+      "loss": 1.4108,
+      "step": 6710
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7689698934555054,
+      "learning_rate": 4.876549603468442e-05,
+      "loss": 1.3159,
+      "step": 6715
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.1612657308578491,
+      "learning_rate": 4.876367196141245e-05,
+      "loss": 1.4331,
+      "step": 6720
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7282540798187256,
+      "learning_rate": 4.876184657569758e-05,
+      "loss": 1.2104,
+      "step": 6725
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.3885889053344727,
+      "learning_rate": 4.876001987764063e-05,
+      "loss": 1.3829,
+      "step": 6730
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.1671245098114014,
+      "learning_rate": 4.8758191867342465e-05,
+      "loss": 1.3192,
+      "step": 6735
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.369300365447998,
+      "learning_rate": 4.875636254490406e-05,
+      "loss": 1.202,
+      "step": 6740
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.8616713881492615,
+      "learning_rate": 4.875453191042646e-05,
+      "loss": 1.3398,
+      "step": 6745
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.1366583108901978,
+      "learning_rate": 4.875269996401074e-05,
+      "loss": 1.306,
+      "step": 6750
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.2424169778823853,
+      "learning_rate": 4.8750866705758106e-05,
+      "loss": 1.2174,
+      "step": 6755
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.3406630754470825,
+      "learning_rate": 4.874903213576977e-05,
+      "loss": 1.2905,
+      "step": 6760
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.9850149750709534,
+      "learning_rate": 4.874719625414709e-05,
+      "loss": 1.4,
+      "step": 6765
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.6209102869033813,
+      "learning_rate": 4.874535906099144e-05,
+      "loss": 1.347,
+      "step": 6770
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.3967558145523071,
+      "learning_rate": 4.874352055640429e-05,
+      "loss": 1.2893,
+      "step": 6775
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.4221364259719849,
+      "learning_rate": 4.874204880857213e-05,
+      "loss": 1.3171,
+      "step": 6780
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.1001510620117188,
+      "learning_rate": 4.8740207943664204e-05,
+      "loss": 1.2465,
+      "step": 6785
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.5954471230506897,
+      "learning_rate": 4.8738365767609275e-05,
+      "loss": 1.2439,
+      "step": 6790
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.8812218308448792,
+      "learning_rate": 4.873652228050908e-05,
+      "loss": 1.3691,
+      "step": 6795
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.045620083808899,
+      "learning_rate": 4.873467748246543e-05,
+      "loss": 1.2654,
+      "step": 6800
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.3172338008880615,
+      "learning_rate": 4.8732831373580216e-05,
+      "loss": 1.3099,
+      "step": 6805
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.8830695748329163,
+      "learning_rate": 4.873098395395539e-05,
+      "loss": 1.1458,
+      "step": 6810
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.0735907554626465,
+      "learning_rate": 4.872913522369299e-05,
+      "loss": 1.2193,
+      "step": 6815
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.7402642965316772,
+      "learning_rate": 4.8727285182895124e-05,
+      "loss": 1.3219,
+      "step": 6820
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.2762502431869507,
+      "learning_rate": 4.8725433831663944e-05,
+      "loss": 1.2906,
+      "step": 6825
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.894281804561615,
+      "learning_rate": 4.872358117010173e-05,
+      "loss": 1.4766,
+      "step": 6830
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.1440064907073975,
+      "learning_rate": 4.872172719831078e-05,
+      "loss": 1.2786,
+      "step": 6835
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.659610629081726,
+      "learning_rate": 4.8719871916393495e-05,
+      "loss": 1.4152,
+      "step": 6840
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.5597167015075684,
+      "learning_rate": 4.8718015324452336e-05,
+      "loss": 1.2945,
+      "step": 6845
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.578195571899414,
+      "learning_rate": 4.8716157422589855e-05,
+      "loss": 1.4838,
+      "step": 6850
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.8329493403434753,
+      "learning_rate": 4.8714298210908646e-05,
+      "loss": 1.2861,
+      "step": 6855
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.347894310951233,
+      "learning_rate": 4.8712437689511395e-05,
+      "loss": 1.3531,
+      "step": 6860
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.246103286743164,
+      "learning_rate": 4.871057585850085e-05,
+      "loss": 1.3114,
+      "step": 6865
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.8596112132072449,
+      "learning_rate": 4.870871271797986e-05,
+      "loss": 1.4278,
+      "step": 6870
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.986096203327179,
+      "learning_rate": 4.8706848268051305e-05,
+      "loss": 1.3325,
+      "step": 6875
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.3321117162704468,
+      "learning_rate": 4.870498250881816e-05,
+      "loss": 1.3678,
+      "step": 6880
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.3195933103561401,
+      "learning_rate": 4.8703115440383474e-05,
+      "loss": 1.2902,
+      "step": 6885
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.8268415927886963,
+      "learning_rate": 4.8701247062850355e-05,
+      "loss": 1.2304,
+      "step": 6890
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.8425890207290649,
+      "learning_rate": 4.8699377376322e-05,
+      "loss": 1.3488,
+      "step": 6895
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.5140923261642456,
+      "learning_rate": 4.869750638090167e-05,
+      "loss": 1.3143,
+      "step": 6900
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.7727601528167725,
+      "learning_rate": 4.8695634076692696e-05,
+      "loss": 1.4197,
+      "step": 6905
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.5425249338150024,
+      "learning_rate": 4.8693760463798476e-05,
+      "loss": 1.2215,
+      "step": 6910
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.0112886428833008,
+      "learning_rate": 4.869188554232249e-05,
+      "loss": 1.4421,
+      "step": 6915
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.9172187447547913,
+      "learning_rate": 4.86900093123683e-05,
+      "loss": 1.4047,
+      "step": 6920
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.2613557577133179,
+      "learning_rate": 4.868813177403952e-05,
+      "loss": 1.1333,
+      "step": 6925
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.811103880405426,
+      "learning_rate": 4.8686252927439844e-05,
+      "loss": 1.2167,
+      "step": 6930
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.1426024436950684,
+      "learning_rate": 4.868437277267304e-05,
+      "loss": 1.4477,
+      "step": 6935
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6741098165512085,
+      "learning_rate": 4.868249130984294e-05,
+      "loss": 1.4966,
+      "step": 6940
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6223343014717102,
+      "learning_rate": 4.868060853905346e-05,
+      "loss": 1.3415,
+      "step": 6945
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.3729140758514404,
+      "learning_rate": 4.86787244604086e-05,
+      "loss": 1.2362,
+      "step": 6950
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.3088891506195068,
+      "learning_rate": 4.8676839074012385e-05,
+      "loss": 1.4647,
+      "step": 6955
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.7941012382507324,
+      "learning_rate": 4.867495237996897e-05,
+      "loss": 1.2942,
+      "step": 6960
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.6377042531967163,
+      "learning_rate": 4.867306437838254e-05,
+      "loss": 1.4394,
+      "step": 6965
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.298292636871338,
+      "learning_rate": 4.867117506935737e-05,
+      "loss": 1.3146,
+      "step": 6970
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.6398062705993652,
+      "learning_rate": 4.8669284452997795e-05,
+      "loss": 1.1931,
+      "step": 6975
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.8299776911735535,
+      "learning_rate": 4.866739252940826e-05,
+      "loss": 1.3339,
+      "step": 6980
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.9005075693130493,
+      "learning_rate": 4.866549929869323e-05,
+      "loss": 1.289,
+      "step": 6985
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.9177480340003967,
+      "learning_rate": 4.866360476095727e-05,
+      "loss": 1.4014,
+      "step": 6990
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.6924612522125244,
+      "learning_rate": 4.866170891630502e-05,
+      "loss": 1.2609,
+      "step": 6995
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.8772937059402466,
+      "learning_rate": 4.865981176484118e-05,
+      "loss": 1.2517,
+      "step": 7000
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.7310841083526611,
+      "learning_rate": 4.865791330667053e-05,
+      "loss": 1.3634,
+      "step": 7005
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.4149776697158813,
+      "learning_rate": 4.865601354189792e-05,
+      "loss": 1.28,
+      "step": 7010
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.9106690287590027,
+      "learning_rate": 4.865411247062827e-05,
+      "loss": 1.4158,
+      "step": 7015
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.424204707145691,
+      "learning_rate": 4.865221009296657e-05,
+      "loss": 1.4474,
+      "step": 7020
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.823806881904602,
+      "learning_rate": 4.86503064090179e-05,
+      "loss": 1.1959,
+      "step": 7025
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.7646584510803223,
+      "learning_rate": 4.864840141888739e-05,
+      "loss": 1.2428,
+      "step": 7030
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7892530560493469,
+      "learning_rate": 4.864649512268024e-05,
+      "loss": 1.4816,
+      "step": 7035
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.802615761756897,
+      "learning_rate": 4.864458752050175e-05,
+      "loss": 1.2827,
+      "step": 7040
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.3489775657653809,
+      "learning_rate": 4.864267861245727e-05,
+      "loss": 1.3001,
+      "step": 7045
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.7209609746932983,
+      "learning_rate": 4.8640768398652224e-05,
+      "loss": 1.2558,
+      "step": 7050
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.2249550819396973,
+      "learning_rate": 4.863885687919212e-05,
+      "loss": 1.4765,
+      "step": 7055
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.2855435609817505,
+      "learning_rate": 4.863694405418251e-05,
+      "loss": 1.1862,
+      "step": 7060
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.4414563179016113,
+      "learning_rate": 4.863502992372906e-05,
+      "loss": 1.1531,
+      "step": 7065
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.9723905920982361,
+      "learning_rate": 4.863311448793746e-05,
+      "loss": 1.3402,
+      "step": 7070
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.274014949798584,
+      "learning_rate": 4.8631197746913525e-05,
+      "loss": 1.5264,
+      "step": 7075
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.9938822388648987,
+      "learning_rate": 4.86292797007631e-05,
+      "loss": 1.2517,
+      "step": 7080
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.461939811706543,
+      "learning_rate": 4.862736034959211e-05,
+      "loss": 1.3553,
+      "step": 7085
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.8186405897140503,
+      "learning_rate": 4.8625439693506576e-05,
+      "loss": 1.3186,
+      "step": 7090
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.9292609691619873,
+      "learning_rate": 4.862351773261256e-05,
+      "loss": 1.3257,
+      "step": 7095
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6807072758674622,
+      "learning_rate": 4.8621594467016216e-05,
+      "loss": 1.2733,
+      "step": 7100
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.9890395402908325,
+      "learning_rate": 4.8619669896823766e-05,
+      "loss": 1.4189,
+      "step": 7105
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6927660703659058,
+      "learning_rate": 4.8617744022141496e-05,
+      "loss": 1.0253,
+      "step": 7110
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.2300829887390137,
+      "learning_rate": 4.861581684307577e-05,
+      "loss": 1.3179,
+      "step": 7115
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.7834036350250244,
+      "learning_rate": 4.8613888359733035e-05,
+      "loss": 1.4093,
+      "step": 7120
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.634334683418274,
+      "learning_rate": 4.861195857221978e-05,
+      "loss": 1.3292,
+      "step": 7125
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.0878982543945312,
+      "learning_rate": 4.86100274806426e-05,
+      "loss": 1.2545,
+      "step": 7130
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.6050611734390259,
+      "learning_rate": 4.8608095085108155e-05,
+      "loss": 1.2622,
+      "step": 7135
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.9072321653366089,
+      "learning_rate": 4.8606161385723147e-05,
+      "loss": 1.2726,
+      "step": 7140
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.5780624151229858,
+      "learning_rate": 4.860422638259439e-05,
+      "loss": 1.3947,
+      "step": 7145
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.0767362117767334,
+      "learning_rate": 4.860229007582874e-05,
+      "loss": 1.4344,
+      "step": 7150
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.11172354221344,
+      "learning_rate": 4.860035246553314e-05,
+      "loss": 1.3602,
+      "step": 7155
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.8127298951148987,
+      "learning_rate": 4.859841355181461e-05,
+      "loss": 1.351,
+      "step": 7160
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.1209958791732788,
+      "learning_rate": 4.8596473334780225e-05,
+      "loss": 1.279,
+      "step": 7165
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.7135673761367798,
+      "learning_rate": 4.859453181453715e-05,
+      "loss": 1.5372,
+      "step": 7170
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5923312902450562,
+      "learning_rate": 4.85925889911926e-05,
+      "loss": 1.3879,
+      "step": 7175
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.2418073415756226,
+      "learning_rate": 4.8590644864853886e-05,
+      "loss": 1.3691,
+      "step": 7180
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.9829604625701904,
+      "learning_rate": 4.858869943562838e-05,
+      "loss": 1.3597,
+      "step": 7185
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.1962437629699707,
+      "learning_rate": 4.8586752703623516e-05,
+      "loss": 1.3715,
+      "step": 7190
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.7202482223510742,
+      "learning_rate": 4.8584804668946825e-05,
+      "loss": 1.446,
+      "step": 7195
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.16395902633667,
+      "learning_rate": 4.858285533170589e-05,
+      "loss": 1.2798,
+      "step": 7200
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6793248057365417,
+      "learning_rate": 4.858090469200835e-05,
+      "loss": 1.3275,
+      "step": 7205
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6636670231819153,
+      "learning_rate": 4.8578952749961974e-05,
+      "loss": 1.433,
+      "step": 7210
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.3391855955123901,
+      "learning_rate": 4.8576999505674546e-05,
+      "loss": 1.4,
+      "step": 7215
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.8590707182884216,
+      "learning_rate": 4.857504495925393e-05,
+      "loss": 1.4144,
+      "step": 7220
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.8004813194274902,
+      "learning_rate": 4.85730891108081e-05,
+      "loss": 1.2784,
+      "step": 7225
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.125076413154602,
+      "learning_rate": 4.8571131960445046e-05,
+      "loss": 1.3068,
+      "step": 7230
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.909472644329071,
+      "learning_rate": 4.856917350827289e-05,
+      "loss": 1.3304,
+      "step": 7235
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.6477268934249878,
+      "learning_rate": 4.8567213754399764e-05,
+      "loss": 1.2217,
+      "step": 7240
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.9648934006690979,
+      "learning_rate": 4.856525269893393e-05,
+      "loss": 1.4223,
+      "step": 7245
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.0322964191436768,
+      "learning_rate": 4.856329034198368e-05,
+      "loss": 1.4918,
+      "step": 7250
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.6450649499893188,
+      "learning_rate": 4.8561326683657405e-05,
+      "loss": 1.3366,
+      "step": 7255
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.0612916946411133,
+      "learning_rate": 4.855936172406354e-05,
+      "loss": 1.2437,
+      "step": 7260
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.8631542921066284,
+      "learning_rate": 4.855739546331062e-05,
+      "loss": 1.3482,
+      "step": 7265
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.2706685066223145,
+      "learning_rate": 4.855542790150723e-05,
+      "loss": 1.3031,
+      "step": 7270
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.1304792165756226,
+      "learning_rate": 4.855345903876204e-05,
+      "loss": 1.4319,
+      "step": 7275
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.4051589965820312,
+      "learning_rate": 4.8551488875183794e-05,
+      "loss": 1.2979,
+      "step": 7280
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.5046559572219849,
+      "learning_rate": 4.8549517410881296e-05,
+      "loss": 1.3051,
+      "step": 7285
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.2949678897857666,
+      "learning_rate": 4.8547544645963435e-05,
+      "loss": 1.4025,
+      "step": 7290
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.3546310663223267,
+      "learning_rate": 4.854557058053915e-05,
+      "loss": 1.2204,
+      "step": 7295
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.055916666984558,
+      "learning_rate": 4.8543595214717486e-05,
+      "loss": 1.3216,
+      "step": 7300
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.6973356008529663,
+      "learning_rate": 4.8541618548607525e-05,
+      "loss": 1.1717,
+      "step": 7305
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.9131316542625427,
+      "learning_rate": 4.853964058231844e-05,
+      "loss": 1.224,
+      "step": 7310
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.3941514492034912,
+      "learning_rate": 4.853766131595948e-05,
+      "loss": 1.4302,
+      "step": 7315
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6351519823074341,
+      "learning_rate": 4.853568074963994e-05,
+      "loss": 1.3049,
+      "step": 7320
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.1367075443267822,
+      "learning_rate": 4.853369888346923e-05,
+      "loss": 1.4376,
+      "step": 7325
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.2908804416656494,
+      "learning_rate": 4.853171571755679e-05,
+      "loss": 1.3539,
+      "step": 7330
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7789058089256287,
+      "learning_rate": 4.8529731252012145e-05,
+      "loss": 1.409,
+      "step": 7335
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.526838779449463,
+      "learning_rate": 4.8527745486944906e-05,
+      "loss": 1.3515,
+      "step": 7340
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.1574357748031616,
+      "learning_rate": 4.852575842246474e-05,
+      "loss": 1.3984,
+      "step": 7345
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.089668035507202,
+      "learning_rate": 4.852377005868138e-05,
+      "loss": 1.3556,
+      "step": 7350
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.4345306158065796,
+      "learning_rate": 4.852178039570466e-05,
+      "loss": 1.2993,
+      "step": 7355
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.0445202589035034,
+      "learning_rate": 4.851978943364446e-05,
+      "loss": 1.4464,
+      "step": 7360
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.4098684787750244,
+      "learning_rate": 4.851779717261072e-05,
+      "loss": 1.4884,
+      "step": 7365
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.9191934466362,
+      "learning_rate": 4.851580361271351e-05,
+      "loss": 1.5069,
+      "step": 7370
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.705700397491455,
+      "learning_rate": 4.8513808754062894e-05,
+      "loss": 1.2528,
+      "step": 7375
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.1692149639129639,
+      "learning_rate": 4.851181259676907e-05,
+      "loss": 1.3765,
+      "step": 7380
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.2977681159973145,
+      "learning_rate": 4.850981514094228e-05,
+      "loss": 1.2544,
+      "step": 7385
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.1639879941940308,
+      "learning_rate": 4.850781638669283e-05,
+      "loss": 1.342,
+      "step": 7390
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.9928296208381653,
+      "learning_rate": 4.8505816334131116e-05,
+      "loss": 1.2802,
+      "step": 7395
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.9788908958435059,
+      "learning_rate": 4.85038149833676e-05,
+      "loss": 1.39,
+      "step": 7400
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.9940471649169922,
+      "learning_rate": 4.850181233451281e-05,
+      "loss": 1.354,
+      "step": 7405
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.0245496034622192,
+      "learning_rate": 4.849980838767736e-05,
+      "loss": 1.3212,
+      "step": 7410
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.179013967514038,
+      "learning_rate": 4.849780314297191e-05,
+      "loss": 1.5085,
+      "step": 7415
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.598319411277771,
+      "learning_rate": 4.8495796600507226e-05,
+      "loss": 1.4813,
+      "step": 7420
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.519370675086975,
+      "learning_rate": 4.8493788760394115e-05,
+      "loss": 1.1946,
+      "step": 7425
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.5374704599380493,
+      "learning_rate": 4.849177962274347e-05,
+      "loss": 1.5137,
+      "step": 7430
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7265501022338867,
+      "learning_rate": 4.8489769187666255e-05,
+      "loss": 1.3113,
+      "step": 7435
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.2072696685791016,
+      "learning_rate": 4.848775745527351e-05,
+      "loss": 1.3973,
+      "step": 7440
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.156033992767334,
+      "learning_rate": 4.848574442567633e-05,
+      "loss": 1.4471,
+      "step": 7445
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.4983317852020264,
+      "learning_rate": 4.848373009898589e-05,
+      "loss": 1.3487,
+      "step": 7450
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.2913556098937988,
+      "learning_rate": 4.848171447531346e-05,
+      "loss": 1.3145,
+      "step": 7455
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.8600425720214844,
+      "learning_rate": 4.847969755477034e-05,
+      "loss": 1.3808,
+      "step": 7460
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.028699278831482,
+      "learning_rate": 4.847767933746793e-05,
+      "loss": 1.3718,
+      "step": 7465
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.1770877838134766,
+      "learning_rate": 4.8475659823517695e-05,
+      "loss": 1.4073,
+      "step": 7470
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6877673268318176,
+      "learning_rate": 4.847363901303117e-05,
+      "loss": 1.3549,
+      "step": 7475
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6186959743499756,
+      "learning_rate": 4.847161690611996e-05,
+      "loss": 1.2288,
+      "step": 7480
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.7008882761001587,
+      "learning_rate": 4.846959350289575e-05,
+      "loss": 1.1874,
+      "step": 7485
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.5733880996704102,
+      "learning_rate": 4.846756880347029e-05,
+      "loss": 1.3694,
+      "step": 7490
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7959824800491333,
+      "learning_rate": 4.846554280795539e-05,
+      "loss": 1.3386,
+      "step": 7495
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.6827507019042969,
+      "learning_rate": 4.8463515516462946e-05,
+      "loss": 1.3071,
+      "step": 7500
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.2505954504013062,
+      "learning_rate": 4.8461486929104936e-05,
+      "loss": 1.3838,
+      "step": 7505
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.9889876246452332,
+      "learning_rate": 4.8459457045993396e-05,
+      "loss": 1.4403,
+      "step": 7510
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.9707186222076416,
+      "learning_rate": 4.845742586724042e-05,
+      "loss": 1.3676,
+      "step": 7515
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.145528554916382,
+      "learning_rate": 4.845539339295819e-05,
+      "loss": 1.5171,
+      "step": 7520
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.8875707387924194,
+      "learning_rate": 4.845335962325897e-05,
+      "loss": 1.236,
+      "step": 7525
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.5600254535675049,
+      "learning_rate": 4.845132455825508e-05,
+      "loss": 1.1585,
+      "step": 7530
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.4640624523162842,
+      "learning_rate": 4.844928819805892e-05,
+      "loss": 1.2519,
+      "step": 7535
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7027156949043274,
+      "learning_rate": 4.844725054278293e-05,
+      "loss": 1.1937,
+      "step": 7540
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.2013248205184937,
+      "learning_rate": 4.8445211592539674e-05,
+      "loss": 1.3886,
+      "step": 7545
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7057489156723022,
+      "learning_rate": 4.844317134744174e-05,
+      "loss": 1.3281,
+      "step": 7550
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.8883240818977356,
+      "learning_rate": 4.8441129807601834e-05,
+      "loss": 1.3181,
+      "step": 7555
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.322933554649353,
+      "learning_rate": 4.8439086973132684e-05,
+      "loss": 1.4801,
+      "step": 7560
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.008849859237671,
+      "learning_rate": 4.843704284414713e-05,
+      "loss": 1.4635,
+      "step": 7565
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.0462682247161865,
+      "learning_rate": 4.8434997420758065e-05,
+      "loss": 1.4163,
+      "step": 7570
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.419507622718811,
+      "learning_rate": 4.843295070307844e-05,
+      "loss": 1.366,
+      "step": 7575
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.5404951572418213,
+      "learning_rate": 4.8430902691221314e-05,
+      "loss": 1.4692,
+      "step": 7580
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.5255093574523926,
+      "learning_rate": 4.842885338529979e-05,
+      "loss": 1.5013,
+      "step": 7585
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.1905360221862793,
+      "learning_rate": 4.842680278542704e-05,
+      "loss": 1.3308,
+      "step": 7590
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.8828074932098389,
+      "learning_rate": 4.842475089171632e-05,
+      "loss": 1.2902,
+      "step": 7595
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.31754732131958,
+      "learning_rate": 4.842269770428096e-05,
+      "loss": 1.2712,
+      "step": 7600
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.3499248027801514,
+      "learning_rate": 4.842064322323436e-05,
+      "loss": 1.419,
+      "step": 7605
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.2709991931915283,
+      "learning_rate": 4.841858744868997e-05,
+      "loss": 1.3574,
+      "step": 7610
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6626339554786682,
+      "learning_rate": 4.8416530380761335e-05,
+      "loss": 1.3227,
+      "step": 7615
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.2082407474517822,
+      "learning_rate": 4.841447201956208e-05,
+      "loss": 1.1635,
+      "step": 7620
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.1640417575836182,
+      "learning_rate": 4.841241236520586e-05,
+      "loss": 1.2477,
+      "step": 7625
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.2419222593307495,
+      "learning_rate": 4.8410351417806454e-05,
+      "loss": 1.1595,
+      "step": 7630
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.1427175998687744,
+      "learning_rate": 4.840828917747766e-05,
+      "loss": 1.3203,
+      "step": 7635
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4415032863616943,
+      "learning_rate": 4.8406225644333395e-05,
+      "loss": 1.2786,
+      "step": 7640
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.9205933809280396,
+      "learning_rate": 4.8404160818487615e-05,
+      "loss": 1.3197,
+      "step": 7645
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.3150992393493652,
+      "learning_rate": 4.840209470005436e-05,
+      "loss": 1.3841,
+      "step": 7650
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.114997148513794,
+      "learning_rate": 4.8400027289147746e-05,
+      "loss": 1.4485,
+      "step": 7655
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.2688688039779663,
+      "learning_rate": 4.8397958585881934e-05,
+      "loss": 1.2421,
+      "step": 7660
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.3500120639801025,
+      "learning_rate": 4.83958885903712e-05,
+      "loss": 1.4009,
+      "step": 7665
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.87721848487854,
+      "learning_rate": 4.839381730272985e-05,
+      "loss": 1.2508,
+      "step": 7670
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.1031057834625244,
+      "learning_rate": 4.83917447230723e-05,
+      "loss": 1.5297,
+      "step": 7675
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.6989408731460571,
+      "learning_rate": 4.838967085151299e-05,
+      "loss": 1.1791,
+      "step": 7680
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.4262398481369019,
+      "learning_rate": 4.8387595688166474e-05,
+      "loss": 1.4372,
+      "step": 7685
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.9426538944244385,
+      "learning_rate": 4.8385519233147355e-05,
+      "loss": 1.2325,
+      "step": 7690
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.5246938467025757,
+      "learning_rate": 4.838344148657033e-05,
+      "loss": 1.3084,
+      "step": 7695
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.8816734552383423,
+      "learning_rate": 4.8381362448550126e-05,
+      "loss": 1.3766,
+      "step": 7700
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.2604855298995972,
+      "learning_rate": 4.837928211920159e-05,
+      "loss": 1.2206,
+      "step": 7705
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.575439691543579,
+      "learning_rate": 4.837720049863958e-05,
+      "loss": 1.4169,
+      "step": 7710
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.9932951331138611,
+      "learning_rate": 4.837511758697911e-05,
+      "loss": 1.3783,
+      "step": 7715
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.9457315802574158,
+      "learning_rate": 4.8373033384335185e-05,
+      "loss": 1.3412,
+      "step": 7720
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.7361222505569458,
+      "learning_rate": 4.8370947890822914e-05,
+      "loss": 1.3247,
+      "step": 7725
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6409812569618225,
+      "learning_rate": 4.8368861106557494e-05,
+      "loss": 1.3878,
+      "step": 7730
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.8458154201507568,
+      "learning_rate": 4.8366773031654155e-05,
+      "loss": 1.3113,
+      "step": 7735
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.0293338298797607,
+      "learning_rate": 4.836468366622824e-05,
+      "loss": 1.5338,
+      "step": 7740
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.2134134769439697,
+      "learning_rate": 4.836259301039513e-05,
+      "loss": 1.4984,
+      "step": 7745
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.1873372793197632,
+      "learning_rate": 4.8360501064270293e-05,
+      "loss": 1.2362,
+      "step": 7750
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.671053647994995,
+      "learning_rate": 4.835840782796925e-05,
+      "loss": 1.4207,
+      "step": 7755
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.1975551843643188,
+      "learning_rate": 4.835631330160764e-05,
+      "loss": 1.3198,
+      "step": 7760
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.6073672771453857,
+      "learning_rate": 4.835421748530112e-05,
+      "loss": 1.2571,
+      "step": 7765
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.3273591995239258,
+      "learning_rate": 4.8352120379165444e-05,
+      "loss": 1.5365,
+      "step": 7770
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.3509129285812378,
+      "learning_rate": 4.835002198331643e-05,
+      "loss": 1.4124,
+      "step": 7775
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.4344903230667114,
+      "learning_rate": 4.834792229786997e-05,
+      "loss": 1.2365,
+      "step": 7780
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.8779138922691345,
+      "learning_rate": 4.834582132294203e-05,
+      "loss": 1.3173,
+      "step": 7785
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.2786595821380615,
+      "learning_rate": 4.834371905864865e-05,
+      "loss": 1.5267,
+      "step": 7790
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.5323184728622437,
+      "learning_rate": 4.834161550510593e-05,
+      "loss": 1.2912,
+      "step": 7795
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7492801547050476,
+      "learning_rate": 4.8339510662430046e-05,
+      "loss": 1.3253,
+      "step": 7800
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.9587426781654358,
+      "learning_rate": 4.833740453073725e-05,
+      "loss": 1.3364,
+      "step": 7805
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.607540488243103,
+      "learning_rate": 4.833529711014386e-05,
+      "loss": 1.2618,
+      "step": 7810
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.6319369077682495,
+      "learning_rate": 4.833318840076626e-05,
+      "loss": 1.2573,
+      "step": 7815
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.2608624696731567,
+      "learning_rate": 4.833107840272092e-05,
+      "loss": 1.1074,
+      "step": 7820
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8057272434234619,
+      "learning_rate": 4.832896711612438e-05,
+      "loss": 1.3739,
+      "step": 7825
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.2614953517913818,
+      "learning_rate": 4.832685454109323e-05,
+      "loss": 1.4891,
+      "step": 7830
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.4242806434631348,
+      "learning_rate": 4.8324740677744154e-05,
+      "loss": 1.254,
+      "step": 7835
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.3085395097732544,
+      "learning_rate": 4.832262552619389e-05,
+      "loss": 1.3165,
+      "step": 7840
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.3229748010635376,
+      "learning_rate": 4.832050908655926e-05,
+      "loss": 1.4134,
+      "step": 7845
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.4007372856140137,
+      "learning_rate": 4.831839135895716e-05,
+      "loss": 1.3104,
+      "step": 7850
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.3286653757095337,
+      "learning_rate": 4.831627234350453e-05,
+      "loss": 1.2269,
+      "step": 7855
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.3296312093734741,
+      "learning_rate": 4.831415204031843e-05,
+      "loss": 1.3776,
+      "step": 7860
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.8926234841346741,
+      "learning_rate": 4.831203044951593e-05,
+      "loss": 1.3842,
+      "step": 7865
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.3321502208709717,
+      "learning_rate": 4.8309907571214234e-05,
+      "loss": 1.2039,
+      "step": 7870
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7737197875976562,
+      "learning_rate": 4.830778340553057e-05,
+      "loss": 1.3269,
+      "step": 7875
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.7216429710388184,
+      "learning_rate": 4.830565795258225e-05,
+      "loss": 1.4176,
+      "step": 7880
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.1848835945129395,
+      "learning_rate": 4.830353121248667e-05,
+      "loss": 1.429,
+      "step": 7885
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.1508842706680298,
+      "learning_rate": 4.830140318536128e-05,
+      "loss": 1.361,
+      "step": 7890
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.9978459477424622,
+      "learning_rate": 4.829927387132362e-05,
+      "loss": 1.2397,
+      "step": 7895
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.3493157625198364,
+      "learning_rate": 4.829714327049127e-05,
+      "loss": 1.3463,
+      "step": 7900
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.3078923225402832,
+      "learning_rate": 4.829501138298192e-05,
+      "loss": 1.3451,
+      "step": 7905
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.330299735069275,
+      "learning_rate": 4.829287820891332e-05,
+      "loss": 1.4147,
+      "step": 7910
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.1133418083190918,
+      "learning_rate": 4.829074374840325e-05,
+      "loss": 1.2857,
+      "step": 7915
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.8395391702651978,
+      "learning_rate": 4.828860800156961e-05,
+      "loss": 1.262,
+      "step": 7920
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.1168737411499023,
+      "learning_rate": 4.8286470968530375e-05,
+      "loss": 1.364,
+      "step": 7925
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.8225456476211548,
+      "learning_rate": 4.8284332649403534e-05,
+      "loss": 1.3554,
+      "step": 7930
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.4843734502792358,
+      "learning_rate": 4.8282193044307213e-05,
+      "loss": 1.3841,
+      "step": 7935
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.07279896736145,
+      "learning_rate": 4.8280052153359565e-05,
+      "loss": 1.1941,
+      "step": 7940
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.881043016910553,
+      "learning_rate": 4.8277909976678847e-05,
+      "loss": 1.2924,
+      "step": 7945
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.8441745042800903,
+      "learning_rate": 4.8275766514383346e-05,
+      "loss": 1.3775,
+      "step": 7950
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.2822072505950928,
+      "learning_rate": 4.827362176659146e-05,
+      "loss": 1.3072,
+      "step": 7955
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.9727440476417542,
+      "learning_rate": 4.8271475733421636e-05,
+      "loss": 1.4031,
+      "step": 7960
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.616499900817871,
+      "learning_rate": 4.826932841499239e-05,
+      "loss": 1.383,
+      "step": 7965
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.3687835931777954,
+      "learning_rate": 4.826717981142233e-05,
+      "loss": 1.2824,
+      "step": 7970
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.598733901977539,
+      "learning_rate": 4.826502992283011e-05,
+      "loss": 1.2172,
+      "step": 7975
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.1800307035446167,
+      "learning_rate": 4.826287874933446e-05,
+      "loss": 1.2832,
+      "step": 7980
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.8195118308067322,
+      "learning_rate": 4.826072629105422e-05,
+      "loss": 1.2309,
+      "step": 7985
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.522782564163208,
+      "learning_rate": 4.8258572548108226e-05,
+      "loss": 1.4097,
+      "step": 7990
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.1727079153060913,
+      "learning_rate": 4.8256417520615446e-05,
+      "loss": 1.507,
+      "step": 7995
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7329881191253662,
+      "learning_rate": 4.825426120869491e-05,
+      "loss": 1.2724,
+      "step": 8000
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.2820346355438232,
+      "learning_rate": 4.825210361246569e-05,
+      "loss": 1.4193,
+      "step": 8005
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.400033712387085,
+      "learning_rate": 4.824994473204697e-05,
+      "loss": 1.2408,
+      "step": 8010
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.259271502494812,
+      "learning_rate": 4.824778456755796e-05,
+      "loss": 1.1783,
+      "step": 8015
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.4319287538528442,
+      "learning_rate": 4.824562311911798e-05,
+      "loss": 1.2565,
+      "step": 8020
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.389338493347168,
+      "learning_rate": 4.824346038684638e-05,
+      "loss": 1.4043,
+      "step": 8025
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.8132510185241699,
+      "learning_rate": 4.824129637086264e-05,
+      "loss": 1.3739,
+      "step": 8030
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.4358786344528198,
+      "learning_rate": 4.823913107128626e-05,
+      "loss": 1.2356,
+      "step": 8035
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.380957841873169,
+      "learning_rate": 4.823696448823681e-05,
+      "loss": 1.41,
+      "step": 8040
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.9512829780578613,
+      "learning_rate": 4.823479662183398e-05,
+      "loss": 1.1268,
+      "step": 8045
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.8032479286193848,
+      "learning_rate": 4.823262747219749e-05,
+      "loss": 1.3039,
+      "step": 8050
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.2832984924316406,
+      "learning_rate": 4.823045703944712e-05,
+      "loss": 1.3076,
+      "step": 8055
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.5915247201919556,
+      "learning_rate": 4.8228285323702754e-05,
+      "loss": 1.3724,
+      "step": 8060
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7308753132820129,
+      "learning_rate": 4.8226112325084335e-05,
+      "loss": 1.1823,
+      "step": 8065
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.5579122304916382,
+      "learning_rate": 4.822393804371188e-05,
+      "loss": 1.2714,
+      "step": 8070
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.8810973167419434,
+      "learning_rate": 4.822176247970547e-05,
+      "loss": 1.3433,
+      "step": 8075
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7145785689353943,
+      "learning_rate": 4.821958563318524e-05,
+      "loss": 1.2667,
+      "step": 8080
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.8525540828704834,
+      "learning_rate": 4.8217407504271446e-05,
+      "loss": 1.3189,
+      "step": 8085
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.6017142534255981,
+      "learning_rate": 4.821522809308436e-05,
+      "loss": 1.2954,
+      "step": 8090
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.1823608875274658,
+      "learning_rate": 4.821304739974437e-05,
+      "loss": 1.4232,
+      "step": 8095
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.2978624105453491,
+      "learning_rate": 4.821086542437189e-05,
+      "loss": 1.3639,
+      "step": 8100
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.7997733354568481,
+      "learning_rate": 4.8208682167087436e-05,
+      "loss": 1.3463,
+      "step": 8105
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.6938416957855225,
+      "learning_rate": 4.820649762801159e-05,
+      "loss": 1.3213,
+      "step": 8110
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.9703941941261292,
+      "learning_rate": 4.820431180726501e-05,
+      "loss": 1.2537,
+      "step": 8115
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.5119740962982178,
+      "learning_rate": 4.820212470496841e-05,
+      "loss": 1.2918,
+      "step": 8120
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.7154700756072998,
+      "learning_rate": 4.8199936321242576e-05,
+      "loss": 1.4548,
+      "step": 8125
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7919926643371582,
+      "learning_rate": 4.819774665620837e-05,
+      "loss": 1.2193,
+      "step": 8130
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.4892973899841309,
+      "learning_rate": 4.819555570998673e-05,
+      "loss": 1.325,
+      "step": 8135
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.446234107017517,
+      "learning_rate": 4.819336348269866e-05,
+      "loss": 1.371,
+      "step": 8140
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.9358090162277222,
+      "learning_rate": 4.8191169974465235e-05,
+      "loss": 1.3127,
+      "step": 8145
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.372162938117981,
+      "learning_rate": 4.818897518540759e-05,
+      "loss": 1.3957,
+      "step": 8150
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.8464466333389282,
+      "learning_rate": 4.818677911564696e-05,
+      "loss": 1.3159,
+      "step": 8155
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7592960000038147,
+      "learning_rate": 4.8184581765304616e-05,
+      "loss": 1.3153,
+      "step": 8160
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.8226920366287231,
+      "learning_rate": 4.8182383134501915e-05,
+      "loss": 1.4496,
+      "step": 8165
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.1266101598739624,
+      "learning_rate": 4.81801832233603e-05,
+      "loss": 1.2566,
+      "step": 8170
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.9116572141647339,
+      "learning_rate": 4.817798203200126e-05,
+      "loss": 1.2411,
+      "step": 8175
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.1480777263641357,
+      "learning_rate": 4.8175779560546357e-05,
+      "loss": 1.4541,
+      "step": 8180
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.545530080795288,
+      "learning_rate": 4.8173575809117246e-05,
+      "loss": 1.2647,
+      "step": 8185
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.0556013584136963,
+      "learning_rate": 4.817137077783562e-05,
+      "loss": 1.4301,
+      "step": 8190
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.490013599395752,
+      "learning_rate": 4.816916446682328e-05,
+      "loss": 1.5971,
+      "step": 8195
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.620077133178711,
+      "learning_rate": 4.8166956876202066e-05,
+      "loss": 1.2097,
+      "step": 8200
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.2473911046981812,
+      "learning_rate": 4.816474800609391e-05,
+      "loss": 1.4232,
+      "step": 8205
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.5744521617889404,
+      "learning_rate": 4.816253785662079e-05,
+      "loss": 1.2997,
+      "step": 8210
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.0682075023651123,
+      "learning_rate": 4.816032642790479e-05,
+      "loss": 1.2491,
+      "step": 8215
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.989682674407959,
+      "learning_rate": 4.815811372006803e-05,
+      "loss": 1.2638,
+      "step": 8220
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.9861974120140076,
+      "learning_rate": 4.8155899733232724e-05,
+      "loss": 1.3134,
+      "step": 8225
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.6601946353912354,
+      "learning_rate": 4.8153684467521145e-05,
+      "loss": 1.2995,
+      "step": 8230
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.8159310817718506,
+      "learning_rate": 4.8151467923055636e-05,
+      "loss": 1.4021,
+      "step": 8235
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.435726523399353,
+      "learning_rate": 4.814925009995862e-05,
+      "loss": 1.2735,
+      "step": 8240
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.530027985572815,
+      "learning_rate": 4.8147030998352585e-05,
+      "loss": 1.436,
+      "step": 8245
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.6053141355514526,
+      "learning_rate": 4.814481061836008e-05,
+      "loss": 1.2369,
+      "step": 8250
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.5459723472595215,
+      "learning_rate": 4.814258896010375e-05,
+      "loss": 1.2969,
+      "step": 8255
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.4241658449172974,
+      "learning_rate": 4.814036602370628e-05,
+      "loss": 1.2725,
+      "step": 8260
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.7884745597839355,
+      "learning_rate": 4.813814180929046e-05,
+      "loss": 1.3672,
+      "step": 8265
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.1821926832199097,
+      "learning_rate": 4.8135916316979114e-05,
+      "loss": 1.2665,
+      "step": 8270
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.176249623298645,
+      "learning_rate": 4.813368954689516e-05,
+      "loss": 1.2877,
+      "step": 8275
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.3001742362976074,
+      "learning_rate": 4.813146149916157e-05,
+      "loss": 1.1879,
+      "step": 8280
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.2863318920135498,
+      "learning_rate": 4.812923217390141e-05,
+      "loss": 1.2886,
+      "step": 8285
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.8792412281036377,
+      "learning_rate": 4.81270015712378e-05,
+      "loss": 1.3733,
+      "step": 8290
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.0877119302749634,
+      "learning_rate": 4.8124769691293925e-05,
+      "loss": 1.2741,
+      "step": 8295
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.0546271800994873,
+      "learning_rate": 4.812253653419306e-05,
+      "loss": 1.2784,
+      "step": 8300
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.8137794137001038,
+      "learning_rate": 4.8120302100058545e-05,
+      "loss": 1.3681,
+      "step": 8305
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.380187749862671,
+      "learning_rate": 4.811806638901377e-05,
+      "loss": 1.2771,
+      "step": 8310
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.1586872339248657,
+      "learning_rate": 4.8115829401182224e-05,
+      "loss": 1.3558,
+      "step": 8315
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.9364005327224731,
+      "learning_rate": 4.811359113668744e-05,
+      "loss": 1.4381,
+      "step": 8320
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.160302758216858,
+      "learning_rate": 4.8111351595653044e-05,
+      "loss": 1.4166,
+      "step": 8325
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.486975908279419,
+      "learning_rate": 4.810911077820273e-05,
+      "loss": 1.3687,
+      "step": 8330
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.307428002357483,
+      "learning_rate": 4.810686868446024e-05,
+      "loss": 1.2001,
+      "step": 8335
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.7174419164657593,
+      "learning_rate": 4.8104625314549414e-05,
+      "loss": 1.3438,
+      "step": 8340
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.201108455657959,
+      "learning_rate": 4.810238066859415e-05,
+      "loss": 1.3922,
+      "step": 8345
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.2420109510421753,
+      "learning_rate": 4.8100134746718405e-05,
+      "loss": 1.352,
+      "step": 8350
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.2646502256393433,
+      "learning_rate": 4.809788754904624e-05,
+      "loss": 1.291,
+      "step": 8355
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.1232274770736694,
+      "learning_rate": 4.809563907570175e-05,
+      "loss": 1.1746,
+      "step": 8360
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.6200834512710571,
+      "learning_rate": 4.809338932680912e-05,
+      "loss": 1.3263,
+      "step": 8365
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.9596359729766846,
+      "learning_rate": 4.809113830249261e-05,
+      "loss": 1.4203,
+      "step": 8370
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.5598084926605225,
+      "learning_rate": 4.808888600287652e-05,
+      "loss": 1.2957,
+      "step": 8375
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.9596273899078369,
+      "learning_rate": 4.808663242808526e-05,
+      "loss": 1.3509,
+      "step": 8380
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.8248343467712402,
+      "learning_rate": 4.80843775782433e-05,
+      "loss": 1.4082,
+      "step": 8385
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.0814073085784912,
+      "learning_rate": 4.808212145347515e-05,
+      "loss": 1.382,
+      "step": 8390
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.3564740419387817,
+      "learning_rate": 4.807986405390543e-05,
+      "loss": 1.3745,
+      "step": 8395
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.0324515104293823,
+      "learning_rate": 4.8077605379658804e-05,
+      "loss": 1.2376,
+      "step": 8400
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.1277930736541748,
+      "learning_rate": 4.807534543086002e-05,
+      "loss": 1.3123,
+      "step": 8405
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.6605470180511475,
+      "learning_rate": 4.807308420763389e-05,
+      "loss": 1.3997,
+      "step": 8410
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.9557091593742371,
+      "learning_rate": 4.807082171010531e-05,
+      "loss": 1.3075,
+      "step": 8415
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.9519081115722656,
+      "learning_rate": 4.8068557938399225e-05,
+      "loss": 1.3437,
+      "step": 8420
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.2350136041641235,
+      "learning_rate": 4.8066292892640666e-05,
+      "loss": 1.3418,
+      "step": 8425
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.178934931755066,
+      "learning_rate": 4.8064026572954726e-05,
+      "loss": 1.2916,
+      "step": 8430
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.9107993841171265,
+      "learning_rate": 4.806175897946657e-05,
+      "loss": 1.1943,
+      "step": 8435
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.4889490604400635,
+      "learning_rate": 4.805949011230144e-05,
+      "loss": 1.3919,
+      "step": 8440
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8420235514640808,
+      "learning_rate": 4.805721997158463e-05,
+      "loss": 1.323,
+      "step": 8445
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.4694463014602661,
+      "learning_rate": 4.8054948557441535e-05,
+      "loss": 1.2482,
+      "step": 8450
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.4326045513153076,
+      "learning_rate": 4.8052675869997596e-05,
+      "loss": 1.2654,
+      "step": 8455
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.6901538372039795,
+      "learning_rate": 4.805040190937833e-05,
+      "loss": 1.2435,
+      "step": 8460
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.1050915718078613,
+      "learning_rate": 4.804812667570933e-05,
+      "loss": 1.2554,
+      "step": 8465
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.468774676322937,
+      "learning_rate": 4.8045850169116244e-05,
+      "loss": 1.319,
+      "step": 8470
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.1394233703613281,
+      "learning_rate": 4.804357238972482e-05,
+      "loss": 1.3279,
+      "step": 8475
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.781924843788147,
+      "learning_rate": 4.804129333766083e-05,
+      "loss": 1.3171,
+      "step": 8480
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.5067170858383179,
+      "learning_rate": 4.803901301305017e-05,
+      "loss": 1.3916,
+      "step": 8485
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.2595572471618652,
+      "learning_rate": 4.803673141601877e-05,
+      "loss": 1.3188,
+      "step": 8490
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.1094878911972046,
+      "learning_rate": 4.803444854669262e-05,
+      "loss": 1.346,
+      "step": 8495
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.4350841045379639,
+      "learning_rate": 4.803216440519784e-05,
+      "loss": 1.3852,
+      "step": 8500
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.89030122756958,
+      "learning_rate": 4.8029878991660556e-05,
+      "loss": 1.4023,
+      "step": 8505
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.8968062996864319,
+      "learning_rate": 4.802759230620699e-05,
+      "loss": 1.329,
+      "step": 8510
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.008400559425354,
+      "learning_rate": 4.802530434896344e-05,
+      "loss": 1.321,
+      "step": 8515
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.945785641670227,
+      "learning_rate": 4.802301512005626e-05,
+      "loss": 1.4198,
+      "step": 8520
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.0313044786453247,
+      "learning_rate": 4.802072461961189e-05,
+      "loss": 1.1841,
+      "step": 8525
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.9936201572418213,
+      "learning_rate": 4.8018432847756823e-05,
+      "loss": 1.4208,
+      "step": 8530
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.7192137241363525,
+      "learning_rate": 4.8016139804617646e-05,
+      "loss": 1.4593,
+      "step": 8535
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.4999003410339355,
+      "learning_rate": 4.801384549032099e-05,
+      "loss": 1.388,
+      "step": 8540
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.0595996379852295,
+      "learning_rate": 4.8011549904993555e-05,
+      "loss": 1.3635,
+      "step": 8545
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.0259523391723633,
+      "learning_rate": 4.800925304876215e-05,
+      "loss": 1.4075,
+      "step": 8550
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.3080471754074097,
+      "learning_rate": 4.800695492175361e-05,
+      "loss": 1.4913,
+      "step": 8555
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.0340911149978638,
+      "learning_rate": 4.800465552409487e-05,
+      "loss": 1.3514,
+      "step": 8560
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.8910084366798401,
+      "learning_rate": 4.800235485591291e-05,
+      "loss": 1.3424,
+      "step": 8565
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.5599783658981323,
+      "learning_rate": 4.8000052917334815e-05,
+      "loss": 1.4032,
+      "step": 8570
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.667799949645996,
+      "learning_rate": 4.7997749708487695e-05,
+      "loss": 1.4332,
+      "step": 8575
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.000908374786377,
+      "learning_rate": 4.799544522949876e-05,
+      "loss": 1.4121,
+      "step": 8580
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7706380486488342,
+      "learning_rate": 4.799313948049529e-05,
+      "loss": 1.1469,
+      "step": 8585
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.065397024154663,
+      "learning_rate": 4.799083246160463e-05,
+      "loss": 1.4589,
+      "step": 8590
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.4384722709655762,
+      "learning_rate": 4.798852417295418e-05,
+      "loss": 1.3352,
+      "step": 8595
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.4065508842468262,
+      "learning_rate": 4.798621461467146e-05,
+      "loss": 1.4188,
+      "step": 8600
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.0686558485031128,
+      "learning_rate": 4.798390378688398e-05,
+      "loss": 1.5157,
+      "step": 8605
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.01014244556427,
+      "learning_rate": 4.798159168971938e-05,
+      "loss": 1.305,
+      "step": 8610
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.5456489324569702,
+      "learning_rate": 4.7979278323305364e-05,
+      "loss": 1.4341,
+      "step": 8615
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.8445463180541992,
+      "learning_rate": 4.7976963687769696e-05,
+      "loss": 1.3751,
+      "step": 8620
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.879667043685913,
+      "learning_rate": 4.797464778324021e-05,
+      "loss": 1.3944,
+      "step": 8625
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.7700638175010681,
+      "learning_rate": 4.79723306098448e-05,
+      "loss": 1.2801,
+      "step": 8630
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.3582429885864258,
+      "learning_rate": 4.7970012167711456e-05,
+      "loss": 1.4009,
+      "step": 8635
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.5631026029586792,
+      "learning_rate": 4.7967692456968207e-05,
+      "loss": 1.3642,
+      "step": 8640
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.840862512588501,
+      "learning_rate": 4.7965371477743185e-05,
+      "loss": 1.1852,
+      "step": 8645
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.0810298919677734,
+      "learning_rate": 4.7963049230164556e-05,
+      "loss": 1.3303,
+      "step": 8650
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.0809688568115234,
+      "learning_rate": 4.7960725714360596e-05,
+      "loss": 1.4119,
+      "step": 8655
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.8407636880874634,
+      "learning_rate": 4.7958400930459626e-05,
+      "loss": 1.2693,
+      "step": 8660
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.3845536708831787,
+      "learning_rate": 4.795607487859003e-05,
+      "loss": 1.3055,
+      "step": 8665
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.7638427019119263,
+      "learning_rate": 4.795374755888028e-05,
+      "loss": 1.2628,
+      "step": 8670
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.1101248264312744,
+      "learning_rate": 4.7951418971458915e-05,
+      "loss": 1.3121,
+      "step": 8675
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5876396894454956,
+      "learning_rate": 4.794908911645453e-05,
+      "loss": 1.3242,
+      "step": 8680
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.759340763092041,
+      "learning_rate": 4.7946757993995815e-05,
+      "loss": 1.2821,
+      "step": 8685
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5042836666107178,
+      "learning_rate": 4.794442560421151e-05,
+      "loss": 1.423,
+      "step": 8690
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.0872600078582764,
+      "learning_rate": 4.794209194723042e-05,
+      "loss": 1.2986,
+      "step": 8695
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.748469352722168,
+      "learning_rate": 4.7939757023181435e-05,
+      "loss": 1.364,
+      "step": 8700
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.6730393171310425,
+      "learning_rate": 4.793742083219353e-05,
+      "loss": 1.3123,
+      "step": 8705
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.2897133827209473,
+      "learning_rate": 4.793508337439569e-05,
+      "loss": 1.3758,
+      "step": 8710
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.8004686832427979,
+      "learning_rate": 4.793274464991706e-05,
+      "loss": 1.3204,
+      "step": 8715
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.192774772644043,
+      "learning_rate": 4.7930404658886766e-05,
+      "loss": 1.25,
+      "step": 8720
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.638178825378418,
+      "learning_rate": 4.7928063401434065e-05,
+      "loss": 1.4043,
+      "step": 8725
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.2886130809783936,
+      "learning_rate": 4.792572087768825e-05,
+      "loss": 1.4547,
+      "step": 8730
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.1815823316574097,
+      "learning_rate": 4.7923377087778695e-05,
+      "loss": 1.4566,
+      "step": 8735
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.7591919898986816,
+      "learning_rate": 4.7921032031834864e-05,
+      "loss": 1.3245,
+      "step": 8740
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.386547565460205,
+      "learning_rate": 4.7918685709986254e-05,
+      "loss": 1.3942,
+      "step": 8745
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.2307301759719849,
+      "learning_rate": 4.791633812236245e-05,
+      "loss": 1.2166,
+      "step": 8750
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.1974438428878784,
+      "learning_rate": 4.791398926909312e-05,
+      "loss": 1.361,
+      "step": 8755
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.449912190437317,
+      "learning_rate": 4.791163915030797e-05,
+      "loss": 1.4183,
+      "step": 8760
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.0868198871612549,
+      "learning_rate": 4.790928776613682e-05,
+      "loss": 1.4563,
+      "step": 8765
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4005632400512695,
+      "learning_rate": 4.790693511670951e-05,
+      "loss": 1.189,
+      "step": 8770
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.2266290187835693,
+      "learning_rate": 4.7904581202155983e-05,
+      "loss": 1.435,
+      "step": 8775
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.5857153534889221,
+      "learning_rate": 4.790222602260625e-05,
+      "loss": 1.3531,
+      "step": 8780
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.5321526527404785,
+      "learning_rate": 4.789986957819037e-05,
+      "loss": 1.361,
+      "step": 8785
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.288081407546997,
+      "learning_rate": 4.78975118690385e-05,
+      "loss": 1.3323,
+      "step": 8790
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.5145295858383179,
+      "learning_rate": 4.7895152895280856e-05,
+      "loss": 1.3989,
+      "step": 8795
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.247381329536438,
+      "learning_rate": 4.7892792657047714e-05,
+      "loss": 1.2564,
+      "step": 8800
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.8097814917564392,
+      "learning_rate": 4.7890903556126435e-05,
+      "loss": 1.44,
+      "step": 8805
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.263486385345459,
+      "learning_rate": 4.7888541042165937e-05,
+      "loss": 1.4437,
+      "step": 8810
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.0362577438354492,
+      "learning_rate": 4.78861772640951e-05,
+      "loss": 1.3168,
+      "step": 8815
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.632360816001892,
+      "learning_rate": 4.7883812222044486e-05,
+      "loss": 1.287,
+      "step": 8820
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.122438669204712,
+      "learning_rate": 4.788144591614472e-05,
+      "loss": 1.3109,
+      "step": 8825
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.8862602710723877,
+      "learning_rate": 4.7879078346526464e-05,
+      "loss": 1.3944,
+      "step": 8830
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.2232091426849365,
+      "learning_rate": 4.7876709513320506e-05,
+      "loss": 1.4264,
+      "step": 8835
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.3420339822769165,
+      "learning_rate": 4.787433941665765e-05,
+      "loss": 1.2224,
+      "step": 8840
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.9875779151916504,
+      "learning_rate": 4.787196805666881e-05,
+      "loss": 1.1951,
+      "step": 8845
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.106386661529541,
+      "learning_rate": 4.7869595433484946e-05,
+      "loss": 1.4364,
+      "step": 8850
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.7247858047485352,
+      "learning_rate": 4.78672215472371e-05,
+      "loss": 1.2084,
+      "step": 8855
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.5715019702911377,
+      "learning_rate": 4.786484639805637e-05,
+      "loss": 1.4345,
+      "step": 8860
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.0408717393875122,
+      "learning_rate": 4.7862469986073954e-05,
+      "loss": 1.4077,
+      "step": 8865
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.5695548057556152,
+      "learning_rate": 4.786009231142108e-05,
+      "loss": 1.2914,
+      "step": 8870
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.7669672966003418,
+      "learning_rate": 4.7857713374229066e-05,
+      "loss": 1.2795,
+      "step": 8875
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.5824629068374634,
+      "learning_rate": 4.78553331746293e-05,
+      "loss": 1.2562,
+      "step": 8880
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.8411540389060974,
+      "learning_rate": 4.7852951712753244e-05,
+      "loss": 1.4505,
+      "step": 8885
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.6968480348587036,
+      "learning_rate": 4.7850568988732416e-05,
+      "loss": 1.3871,
+      "step": 8890
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.4560775756835938,
+      "learning_rate": 4.7848185002698416e-05,
+      "loss": 1.4416,
+      "step": 8895
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.1410988569259644,
+      "learning_rate": 4.7845799754782907e-05,
+      "loss": 1.3336,
+      "step": 8900
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.136461615562439,
+      "learning_rate": 4.784341324511762e-05,
+      "loss": 1.3175,
+      "step": 8905
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.0604544878005981,
+      "learning_rate": 4.784102547383437e-05,
+      "loss": 1.3411,
+      "step": 8910
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.7411584258079529,
+      "learning_rate": 4.783863644106502e-05,
+      "loss": 1.1918,
+      "step": 8915
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.8481019735336304,
+      "learning_rate": 4.783624614694153e-05,
+      "loss": 1.4247,
+      "step": 8920
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.549062967300415,
+      "learning_rate": 4.7833854591595895e-05,
+      "loss": 1.2813,
+      "step": 8925
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.4914495944976807,
+      "learning_rate": 4.78314617751602e-05,
+      "loss": 1.3358,
+      "step": 8930
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.6367831230163574,
+      "learning_rate": 4.782906769776661e-05,
+      "loss": 1.2798,
+      "step": 8935
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.661944031715393,
+      "learning_rate": 4.7826672359547343e-05,
+      "loss": 1.3176,
+      "step": 8940
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.858271598815918,
+      "learning_rate": 4.782427576063468e-05,
+      "loss": 1.1787,
+      "step": 8945
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.1787148714065552,
+      "learning_rate": 4.7821877901160996e-05,
+      "loss": 1.3318,
+      "step": 8950
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.9741628170013428,
+      "learning_rate": 4.781947878125872e-05,
+      "loss": 1.3084,
+      "step": 8955
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.8977104425430298,
+      "learning_rate": 4.781707840106034e-05,
+      "loss": 1.3595,
+      "step": 8960
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.1290730237960815,
+      "learning_rate": 4.781467676069845e-05,
+      "loss": 1.3126,
+      "step": 8965
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.0687497854232788,
+      "learning_rate": 4.7812273860305665e-05,
+      "loss": 1.2979,
+      "step": 8970
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.463879108428955,
+      "learning_rate": 4.780986970001472e-05,
+      "loss": 1.3988,
+      "step": 8975
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7218008637428284,
+      "learning_rate": 4.780746427995837e-05,
+      "loss": 1.3782,
+      "step": 8980
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7819427847862244,
+      "learning_rate": 4.7805057600269485e-05,
+      "loss": 1.2902,
+      "step": 8985
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7912634015083313,
+      "learning_rate": 4.780264966108097e-05,
+      "loss": 1.2542,
+      "step": 8990
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.160401701927185,
+      "learning_rate": 4.780024046252581e-05,
+      "loss": 1.2758,
+      "step": 8995
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7142820954322815,
+      "learning_rate": 4.779783000473707e-05,
+      "loss": 1.278,
+      "step": 9000
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.7503546476364136,
+      "learning_rate": 4.779541828784788e-05,
+      "loss": 1.3523,
+      "step": 9005
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.522162914276123,
+      "learning_rate": 4.779300531199143e-05,
+      "loss": 1.3737,
+      "step": 9010
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.4219088554382324,
+      "learning_rate": 4.779059107730099e-05,
+      "loss": 1.4303,
+      "step": 9015
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.6250290870666504,
+      "learning_rate": 4.778817558390989e-05,
+      "loss": 1.366,
+      "step": 9020
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.2301737070083618,
+      "learning_rate": 4.7785758831951543e-05,
+      "loss": 1.3006,
+      "step": 9025
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.753127932548523,
+      "learning_rate": 4.778334082155942e-05,
+      "loss": 1.4303,
+      "step": 9030
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.423354148864746,
+      "learning_rate": 4.778092155286707e-05,
+      "loss": 1.2672,
+      "step": 9035
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.4239237308502197,
+      "learning_rate": 4.777850102600809e-05,
+      "loss": 1.2822,
+      "step": 9040
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.1485992670059204,
+      "learning_rate": 4.777607924111619e-05,
+      "loss": 1.3411,
+      "step": 9045
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7235462665557861,
+      "learning_rate": 4.77736561983251e-05,
+      "loss": 1.3456,
+      "step": 9050
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.1545169353485107,
+      "learning_rate": 4.7771231897768655e-05,
+      "loss": 1.4223,
+      "step": 9055
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 2.123680353164673,
+      "learning_rate": 4.776880633958073e-05,
+      "loss": 1.4271,
+      "step": 9060
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.6497604846954346,
+      "learning_rate": 4.776637952389531e-05,
+      "loss": 1.4639,
+      "step": 9065
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.2287499904632568,
+      "learning_rate": 4.776395145084641e-05,
+      "loss": 1.2522,
+      "step": 9070
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.5195343494415283,
+      "learning_rate": 4.7761522120568134e-05,
+      "loss": 1.4156,
+      "step": 9075
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.1889725923538208,
+      "learning_rate": 4.775909153319465e-05,
+      "loss": 1.2898,
+      "step": 9080
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.7549879550933838,
+      "learning_rate": 4.775665968886019e-05,
+      "loss": 1.2376,
+      "step": 9085
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.9946390390396118,
+      "learning_rate": 4.775422658769908e-05,
+      "loss": 1.4544,
+      "step": 9090
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.594741940498352,
+      "learning_rate": 4.775179222984568e-05,
+      "loss": 1.2683,
+      "step": 9095
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.0693196058273315,
+      "learning_rate": 4.774935661543445e-05,
+      "loss": 1.3477,
+      "step": 9100
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.610431671142578,
+      "learning_rate": 4.774691974459989e-05,
+      "loss": 1.321,
+      "step": 9105
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.1639457941055298,
+      "learning_rate": 4.774448161747661e-05,
+      "loss": 1.3614,
+      "step": 9110
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.8405585289001465,
+      "learning_rate": 4.774204223419925e-05,
+      "loss": 1.2398,
+      "step": 9115
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.058262586593628,
+      "learning_rate": 4.773960159490253e-05,
+      "loss": 1.2883,
+      "step": 9120
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 6.126988410949707,
+      "learning_rate": 4.773715969972125e-05,
+      "loss": 1.4032,
+      "step": 9125
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.726212739944458,
+      "learning_rate": 4.7734716548790274e-05,
+      "loss": 1.362,
+      "step": 9130
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.678607940673828,
+      "learning_rate": 4.773227214224454e-05,
+      "loss": 1.3634,
+      "step": 9135
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.9776396155357361,
+      "learning_rate": 4.7729826480219044e-05,
+      "loss": 1.274,
+      "step": 9140
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.055379033088684,
+      "learning_rate": 4.772737956284885e-05,
+      "loss": 1.219,
+      "step": 9145
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.7453423738479614,
+      "learning_rate": 4.7724931390269115e-05,
+      "loss": 1.1836,
+      "step": 9150
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.390795350074768,
+      "learning_rate": 4.772248196261504e-05,
+      "loss": 1.4451,
+      "step": 9155
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.7880897521972656,
+      "learning_rate": 4.7720031280021905e-05,
+      "loss": 1.3696,
+      "step": 9160
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.2243832349777222,
+      "learning_rate": 4.771757934262505e-05,
+      "loss": 1.2609,
+      "step": 9165
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.5606729984283447,
+      "learning_rate": 4.771512615055991e-05,
+      "loss": 1.4425,
+      "step": 9170
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.7605032920837402,
+      "learning_rate": 4.771267170396196e-05,
+      "loss": 1.4839,
+      "step": 9175
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.1422829627990723,
+      "learning_rate": 4.771021600296676e-05,
+      "loss": 1.4384,
+      "step": 9180
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.3212814331054688,
+      "learning_rate": 4.770775904770994e-05,
+      "loss": 1.4812,
+      "step": 9185
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.6888114213943481,
+      "learning_rate": 4.770530083832719e-05,
+      "loss": 1.3274,
+      "step": 9190
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.3165194988250732,
+      "learning_rate": 4.770284137495428e-05,
+      "loss": 1.3288,
+      "step": 9195
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.9350600242614746,
+      "learning_rate": 4.7700380657727027e-05,
+      "loss": 1.459,
+      "step": 9200
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.657118797302246,
+      "learning_rate": 4.769791868678135e-05,
+      "loss": 1.277,
+      "step": 9205
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7231934666633606,
+      "learning_rate": 4.769545546225322e-05,
+      "loss": 1.3761,
+      "step": 9210
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.0625771284103394,
+      "learning_rate": 4.7692990984278676e-05,
+      "loss": 1.294,
+      "step": 9215
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.9042518138885498,
+      "learning_rate": 4.769052525299383e-05,
+      "loss": 1.4183,
+      "step": 9220
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.654555082321167,
+      "learning_rate": 4.7688058268534855e-05,
+      "loss": 1.365,
+      "step": 9225
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.9852314591407776,
+      "learning_rate": 4.768559003103801e-05,
+      "loss": 1.4423,
+      "step": 9230
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.300337791442871,
+      "learning_rate": 4.76831205406396e-05,
+      "loss": 1.2752,
+      "step": 9235
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.255034327507019,
+      "learning_rate": 4.768064979747603e-05,
+      "loss": 1.3097,
+      "step": 9240
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.3751298189163208,
+      "learning_rate": 4.767817780168374e-05,
+      "loss": 1.2666,
+      "step": 9245
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.7448214888572693,
+      "learning_rate": 4.7675704553399265e-05,
+      "loss": 1.4518,
+      "step": 9250
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.0621886253356934,
+      "learning_rate": 4.76732300527592e-05,
+      "loss": 1.4063,
+      "step": 9255
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.179983139038086,
+      "learning_rate": 4.76707542999002e-05,
+      "loss": 1.425,
+      "step": 9260
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.047968864440918,
+      "learning_rate": 4.7668277294959006e-05,
+      "loss": 1.3343,
+      "step": 9265
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.8152706623077393,
+      "learning_rate": 4.766579903807242e-05,
+      "loss": 1.3582,
+      "step": 9270
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.3376197814941406,
+      "learning_rate": 4.7663319529377323e-05,
+      "loss": 1.4053,
+      "step": 9275
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.8099796772003174,
+      "learning_rate": 4.7660838769010635e-05,
+      "loss": 1.5009,
+      "step": 9280
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.4460123777389526,
+      "learning_rate": 4.765835675710938e-05,
+      "loss": 1.5084,
+      "step": 9285
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 2.1671342849731445,
+      "learning_rate": 4.765587349381063e-05,
+      "loss": 1.3576,
+      "step": 9290
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.9347168803215027,
+      "learning_rate": 4.765338897925154e-05,
+      "loss": 1.4602,
+      "step": 9295
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.7557331323623657,
+      "learning_rate": 4.765090321356932e-05,
+      "loss": 1.4104,
+      "step": 9300
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.8540169596672058,
+      "learning_rate": 4.764841619690127e-05,
+      "loss": 1.3433,
+      "step": 9305
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.3423655033111572,
+      "learning_rate": 4.764592792938473e-05,
+      "loss": 1.3206,
+      "step": 9310
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.7521955966949463,
+      "learning_rate": 4.764343841115712e-05,
+      "loss": 1.2059,
+      "step": 9315
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.0334832668304443,
+      "learning_rate": 4.764094764235595e-05,
+      "loss": 1.3869,
+      "step": 9320
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 6.126003265380859,
+      "learning_rate": 4.763845562311877e-05,
+      "loss": 1.4294,
+      "step": 9325
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.9254896640777588,
+      "learning_rate": 4.763596235358323e-05,
+      "loss": 1.3488,
+      "step": 9330
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.4646904468536377,
+      "learning_rate": 4.7633467833887017e-05,
+      "loss": 1.397,
+      "step": 9335
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.3618175983428955,
+      "learning_rate": 4.763097206416789e-05,
+      "loss": 1.4709,
+      "step": 9340
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.8631429672241211,
+      "learning_rate": 4.76284750445637e-05,
+      "loss": 1.4577,
+      "step": 9345
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.02370285987854,
+      "learning_rate": 4.762597677521237e-05,
+      "loss": 1.4129,
+      "step": 9350
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7775421142578125,
+      "learning_rate": 4.762347725625184e-05,
+      "loss": 1.3969,
+      "step": 9355
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.6427661180496216,
+      "learning_rate": 4.762097648782019e-05,
+      "loss": 1.4851,
+      "step": 9360
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.482051968574524,
+      "learning_rate": 4.761847447005552e-05,
+      "loss": 1.4408,
+      "step": 9365
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.8775594234466553,
+      "learning_rate": 4.761597120309602e-05,
+      "loss": 1.338,
+      "step": 9370
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.230642795562744,
+      "learning_rate": 4.7613466687079924e-05,
+      "loss": 1.3657,
+      "step": 9375
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.0118967294692993,
+      "learning_rate": 4.7610960922145585e-05,
+      "loss": 1.3863,
+      "step": 9380
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.517298936843872,
+      "learning_rate": 4.7608453908431365e-05,
+      "loss": 1.337,
+      "step": 9385
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.0825563669204712,
+      "learning_rate": 4.760594564607574e-05,
+      "loss": 1.1625,
+      "step": 9390
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.539588212966919,
+      "learning_rate": 4.760343613521724e-05,
+      "loss": 1.3975,
+      "step": 9395
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7479655742645264,
+      "learning_rate": 4.760092537599445e-05,
+      "loss": 1.3449,
+      "step": 9400
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.136042594909668,
+      "learning_rate": 4.7598413368546045e-05,
+      "loss": 1.3229,
+      "step": 9405
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.4666575193405151,
+      "learning_rate": 4.759590011301076e-05,
+      "loss": 1.382,
+      "step": 9410
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.5006098747253418,
+      "learning_rate": 4.7593385609527406e-05,
+      "loss": 1.1975,
+      "step": 9415
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.135817527770996,
+      "learning_rate": 4.7590869858234837e-05,
+      "loss": 1.3296,
+      "step": 9420
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.2089482545852661,
+      "learning_rate": 4.758835285927201e-05,
+      "loss": 1.195,
+      "step": 9425
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.0782078504562378,
+      "learning_rate": 4.758583461277794e-05,
+      "loss": 1.295,
+      "step": 9430
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.9145467877388,
+      "learning_rate": 4.75833151188917e-05,
+      "loss": 1.2709,
+      "step": 9435
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.9970406293869019,
+      "learning_rate": 4.7580794377752436e-05,
+      "loss": 1.1399,
+      "step": 9440
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8484485149383545,
+      "learning_rate": 4.7578272389499375e-05,
+      "loss": 1.3806,
+      "step": 9445
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.205751895904541,
+      "learning_rate": 4.75757491542718e-05,
+      "loss": 1.4711,
+      "step": 9450
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.451490640640259,
+      "learning_rate": 4.757322467220905e-05,
+      "loss": 1.5658,
+      "step": 9455
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.7217803001403809,
+      "learning_rate": 4.757069894345058e-05,
+      "loss": 1.5043,
+      "step": 9460
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.288859248161316,
+      "learning_rate": 4.756817196813587e-05,
+      "loss": 1.4451,
+      "step": 9465
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.3524889945983887,
+      "learning_rate": 4.756564374640447e-05,
+      "loss": 1.1978,
+      "step": 9470
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.2573511600494385,
+      "learning_rate": 4.756311427839602e-05,
+      "loss": 1.4399,
+      "step": 9475
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.9026219844818115,
+      "learning_rate": 4.756058356425024e-05,
+      "loss": 1.262,
+      "step": 9480
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.9231290817260742,
+      "learning_rate": 4.755805160410686e-05,
+      "loss": 1.2867,
+      "step": 9485
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 2.3469276428222656,
+      "learning_rate": 4.755551839810575e-05,
+      "loss": 1.425,
+      "step": 9490
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.1398067474365234,
+      "learning_rate": 4.75529839463868e-05,
+      "loss": 1.3015,
+      "step": 9495
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.265406847000122,
+      "learning_rate": 4.755044824908998e-05,
+      "loss": 1.4007,
+      "step": 9500
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.610261082649231,
+      "learning_rate": 4.754791130635537e-05,
+      "loss": 1.4184,
+      "step": 9505
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.8379100561141968,
+      "learning_rate": 4.754537311832303e-05,
+      "loss": 1.3009,
+      "step": 9510
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 2.690635919570923,
+      "learning_rate": 4.754283368513317e-05,
+      "loss": 1.3863,
+      "step": 9515
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.1724915504455566,
+      "learning_rate": 4.754029300692604e-05,
+      "loss": 1.236,
+      "step": 9520
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.0479933023452759,
+      "learning_rate": 4.753775108384196e-05,
+      "loss": 1.2245,
+      "step": 9525
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 2.2428202629089355,
+      "learning_rate": 4.753520791602132e-05,
+      "loss": 1.3393,
+      "step": 9530
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.5128904581069946,
+      "learning_rate": 4.753266350360456e-05,
+      "loss": 1.3574,
+      "step": 9535
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.9393275380134583,
+      "learning_rate": 4.7530117846732224e-05,
+      "loss": 1.4423,
+      "step": 9540
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.992976427078247,
+      "learning_rate": 4.75275709455449e-05,
+      "loss": 1.4572,
+      "step": 9545
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.182762622833252,
+      "learning_rate": 4.752502280018324e-05,
+      "loss": 1.3492,
+      "step": 9550
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.1004512310028076,
+      "learning_rate": 4.7522473410787985e-05,
+      "loss": 1.3265,
+      "step": 9555
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.9491164684295654,
+      "learning_rate": 4.751992277749994e-05,
+      "loss": 1.2525,
+      "step": 9560
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.315471887588501,
+      "learning_rate": 4.751737090045996e-05,
+      "loss": 1.3748,
+      "step": 9565
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.0210347175598145,
+      "learning_rate": 4.7514817779809e-05,
+      "loss": 1.3718,
+      "step": 9570
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.1683263778686523,
+      "learning_rate": 4.7512263415688054e-05,
+      "loss": 1.3103,
+      "step": 9575
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.9375793933868408,
+      "learning_rate": 4.75097078082382e-05,
+      "loss": 1.4674,
+      "step": 9580
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.5116485357284546,
+      "learning_rate": 4.750715095760058e-05,
+      "loss": 1.3561,
+      "step": 9585
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.965453028678894,
+      "learning_rate": 4.7504592863916405e-05,
+      "loss": 1.4002,
+      "step": 9590
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.2619178295135498,
+      "learning_rate": 4.750203352732696e-05,
+      "loss": 1.2696,
+      "step": 9595
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.385756731033325,
+      "learning_rate": 4.749947294797359e-05,
+      "loss": 1.4395,
+      "step": 9600
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.714622139930725,
+      "learning_rate": 4.749691112599772e-05,
+      "loss": 1.1337,
+      "step": 9605
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.6344813704490662,
+      "learning_rate": 4.7494348061540835e-05,
+      "loss": 1.3205,
+      "step": 9610
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.1575703620910645,
+      "learning_rate": 4.749178375474448e-05,
+      "loss": 1.325,
+      "step": 9615
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.1499351263046265,
+      "learning_rate": 4.7489218205750295e-05,
+      "loss": 1.3551,
+      "step": 9620
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.3958709239959717,
+      "learning_rate": 4.7486651414699965e-05,
+      "loss": 1.2264,
+      "step": 9625
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.359483242034912,
+      "learning_rate": 4.748408338173525e-05,
+      "loss": 1.3342,
+      "step": 9630
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.6914485692977905,
+      "learning_rate": 4.7481514106997975e-05,
+      "loss": 1.4204,
+      "step": 9635
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 3.1051025390625,
+      "learning_rate": 4.747894359063005e-05,
+      "loss": 1.273,
+      "step": 9640
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.179664134979248,
+      "learning_rate": 4.747637183277343e-05,
+      "loss": 1.1155,
+      "step": 9645
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.4367886781692505,
+      "learning_rate": 4.747379883357016e-05,
+      "loss": 1.2716,
+      "step": 9650
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8933409452438354,
+      "learning_rate": 4.7471224593162346e-05,
+      "loss": 1.2237,
+      "step": 9655
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.1432609558105469,
+      "learning_rate": 4.7468649111692145e-05,
+      "loss": 1.4141,
+      "step": 9660
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.6368004083633423,
+      "learning_rate": 4.746607238930182e-05,
+      "loss": 1.3289,
+      "step": 9665
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.7630443572998047,
+      "learning_rate": 4.746349442613366e-05,
+      "loss": 1.2229,
+      "step": 9670
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.9641880989074707,
+      "learning_rate": 4.7460915222330054e-05,
+      "loss": 1.2851,
+      "step": 9675
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.504604458808899,
+      "learning_rate": 4.7458334778033446e-05,
+      "loss": 1.2702,
+      "step": 9680
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.346604585647583,
+      "learning_rate": 4.745575309338636e-05,
+      "loss": 1.3684,
+      "step": 9685
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.1776371002197266,
+      "learning_rate": 4.745317016853137e-05,
+      "loss": 1.2718,
+      "step": 9690
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8132845163345337,
+      "learning_rate": 4.745058600361112e-05,
+      "loss": 1.5028,
+      "step": 9695
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.2717533111572266,
+      "learning_rate": 4.7448000598768346e-05,
+      "loss": 1.2721,
+      "step": 9700
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.1330522298812866,
+      "learning_rate": 4.7445413954145834e-05,
+      "loss": 1.1869,
+      "step": 9705
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.326873779296875,
+      "learning_rate": 4.744282606988645e-05,
+      "loss": 1.2496,
+      "step": 9710
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.6587218642234802,
+      "learning_rate": 4.74402369461331e-05,
+      "loss": 1.2123,
+      "step": 9715
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4744004011154175,
+      "learning_rate": 4.7437646583028784e-05,
+      "loss": 1.4327,
+      "step": 9720
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 2.3345212936401367,
+      "learning_rate": 4.7435054980716576e-05,
+      "loss": 1.2863,
+      "step": 9725
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 2.557159185409546,
+      "learning_rate": 4.74324621393396e-05,
+      "loss": 1.4434,
+      "step": 9730
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.1000850200653076,
+      "learning_rate": 4.742986805904106e-05,
+      "loss": 1.4368,
+      "step": 9735
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.9850451946258545,
+      "learning_rate": 4.742727273996422e-05,
+      "loss": 1.3663,
+      "step": 9740
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.2970614433288574,
+      "learning_rate": 4.7424676182252414e-05,
+      "loss": 1.1515,
+      "step": 9745
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.1286096572875977,
+      "learning_rate": 4.742207838604906e-05,
+      "loss": 1.4357,
+      "step": 9750
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 2.1564157009124756,
+      "learning_rate": 4.741947935149762e-05,
+      "loss": 1.29,
+      "step": 9755
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.436423897743225,
+      "learning_rate": 4.741687907874164e-05,
+      "loss": 1.3875,
+      "step": 9760
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 2.328244924545288,
+      "learning_rate": 4.7414277567924723e-05,
+      "loss": 1.2351,
+      "step": 9765
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.7096843719482422,
+      "learning_rate": 4.741167481919056e-05,
+      "loss": 1.3325,
+      "step": 9770
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.5687915086746216,
+      "learning_rate": 4.740907083268289e-05,
+      "loss": 1.3996,
+      "step": 9775
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8579280376434326,
+      "learning_rate": 4.7406465608545534e-05,
+      "loss": 1.4976,
+      "step": 9780
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.4446935653686523,
+      "learning_rate": 4.740385914692237e-05,
+      "loss": 1.2704,
+      "step": 9785
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.7776928544044495,
+      "learning_rate": 4.7401251447957354e-05,
+      "loss": 1.4133,
+      "step": 9790
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.9897927641868591,
+      "learning_rate": 4.73986425117945e-05,
+      "loss": 1.3766,
+      "step": 9795
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.700108528137207,
+      "learning_rate": 4.739603233857791e-05,
+      "loss": 1.451,
+      "step": 9800
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8862264156341553,
+      "learning_rate": 4.7393420928451733e-05,
+      "loss": 1.3687,
+      "step": 9805
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.4384057521820068,
+      "learning_rate": 4.739080828156019e-05,
+      "loss": 1.3752,
+      "step": 9810
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7352772951126099,
+      "learning_rate": 4.738819439804758e-05,
+      "loss": 1.2511,
+      "step": 9815
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.937332272529602,
+      "learning_rate": 4.738557927805827e-05,
+      "loss": 1.3718,
+      "step": 9820
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.5250344276428223,
+      "learning_rate": 4.738296292173668e-05,
+      "loss": 1.3754,
+      "step": 9825
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.4641368389129639,
+      "learning_rate": 4.7380345329227315e-05,
+      "loss": 1.4706,
+      "step": 9830
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 2.3265063762664795,
+      "learning_rate": 4.737772650067474e-05,
+      "loss": 1.3016,
+      "step": 9835
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.2815556526184082,
+      "learning_rate": 4.737510643622359e-05,
+      "loss": 1.4134,
+      "step": 9840
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.8234738111495972,
+      "learning_rate": 4.7372485136018577e-05,
+      "loss": 1.2904,
+      "step": 9845
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.526788353919983,
+      "learning_rate": 4.736986260020445e-05,
+      "loss": 1.222,
+      "step": 9850
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 2.026932716369629,
+      "learning_rate": 4.736723882892607e-05,
+      "loss": 1.3834,
+      "step": 9855
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.5664438009262085,
+      "learning_rate": 4.736461382232835e-05,
+      "loss": 1.4893,
+      "step": 9860
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8656167387962341,
+      "learning_rate": 4.736198758055624e-05,
+      "loss": 1.2126,
+      "step": 9865
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 2.0590291023254395,
+      "learning_rate": 4.735936010375481e-05,
+      "loss": 1.504,
+      "step": 9870
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.066401481628418,
+      "learning_rate": 4.7356731392069154e-05,
+      "loss": 1.3136,
+      "step": 9875
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.326509952545166,
+      "learning_rate": 4.7354101445644475e-05,
+      "loss": 1.2444,
+      "step": 9880
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.4632080793380737,
+      "learning_rate": 4.7351470264625995e-05,
+      "loss": 1.4441,
+      "step": 9885
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 3.273226261138916,
+      "learning_rate": 4.734883784915905e-05,
+      "loss": 1.2485,
+      "step": 9890
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.4411187171936035,
+      "learning_rate": 4.734620419938902e-05,
+      "loss": 1.2447,
+      "step": 9895
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.597489833831787,
+      "learning_rate": 4.734356931546137e-05,
+      "loss": 1.3136,
+      "step": 9900
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.6306302547454834,
+      "learning_rate": 4.7340933197521595e-05,
+      "loss": 1.4123,
+      "step": 9905
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.817524790763855,
+      "learning_rate": 4.7338295845715316e-05,
+      "loss": 1.4538,
+      "step": 9910
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.2677677869796753,
+      "learning_rate": 4.733565726018816e-05,
+      "loss": 1.2595,
+      "step": 9915
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.2904586791992188,
+      "learning_rate": 4.7333017441085884e-05,
+      "loss": 1.3257,
+      "step": 9920
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.0129778385162354,
+      "learning_rate": 4.733037638855427e-05,
+      "loss": 1.2028,
+      "step": 9925
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7746032476425171,
+      "learning_rate": 4.732773410273917e-05,
+      "loss": 1.174,
+      "step": 9930
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 2.065194845199585,
+      "learning_rate": 4.732509058378653e-05,
+      "loss": 1.2716,
+      "step": 9935
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.3352677822113037,
+      "learning_rate": 4.732244583184234e-05,
+      "loss": 1.4513,
+      "step": 9940
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.7781025171279907,
+      "learning_rate": 4.731979984705267e-05,
+      "loss": 1.3069,
+      "step": 9945
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.0501564741134644,
+      "learning_rate": 4.731715262956365e-05,
+      "loss": 1.3894,
+      "step": 9950
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 2.136993885040283,
+      "learning_rate": 4.7314504179521505e-05,
+      "loss": 1.3747,
+      "step": 9955
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.8982741832733154,
+      "learning_rate": 4.7311854497072474e-05,
+      "loss": 1.4344,
+      "step": 9960
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 2.735962390899658,
+      "learning_rate": 4.730920358236291e-05,
+      "loss": 1.3851,
+      "step": 9965
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.7903727293014526,
+      "learning_rate": 4.730655143553922e-05,
+      "loss": 1.3806,
+      "step": 9970
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.4778308868408203,
+      "learning_rate": 4.7303898056747895e-05,
+      "loss": 1.312,
+      "step": 9975
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.839424133300781,
+      "learning_rate": 4.730124344613545e-05,
+      "loss": 1.3264,
+      "step": 9980
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.6125905513763428,
+      "learning_rate": 4.729858760384851e-05,
+      "loss": 1.3275,
+      "step": 9985
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.407729148864746,
+      "learning_rate": 4.7295930530033765e-05,
+      "loss": 1.401,
+      "step": 9990
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.9895592331886292,
+      "learning_rate": 4.729327222483795e-05,
+      "loss": 1.3163,
+      "step": 9995
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.6113483905792236,
+      "learning_rate": 4.729061268840788e-05,
+      "loss": 1.3118,
+      "step": 10000
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 2.3144493103027344,
+      "learning_rate": 4.728795192089044e-05,
+      "loss": 1.4342,
+      "step": 10005
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 2.2775607109069824,
+      "learning_rate": 4.728528992243258e-05,
+      "loss": 1.3856,
+      "step": 10010
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.822791576385498,
+      "learning_rate": 4.728262669318132e-05,
+      "loss": 1.3132,
+      "step": 10015
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.9759201407432556,
+      "learning_rate": 4.727996223328376e-05,
+      "loss": 1.3979,
+      "step": 10020
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.3269437551498413,
+      "learning_rate": 4.727729654288704e-05,
+      "loss": 1.3228,
+      "step": 10025
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.9397942423820496,
+      "learning_rate": 4.7274629622138384e-05,
+      "loss": 1.2653,
+      "step": 10030
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8297321200370789,
+      "learning_rate": 4.7271961471185086e-05,
+      "loss": 1.2575,
+      "step": 10035
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.417336106300354,
+      "learning_rate": 4.7269292090174514e-05,
+      "loss": 1.2264,
+      "step": 10040
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 2.3243861198425293,
+      "learning_rate": 4.7266621479254084e-05,
+      "loss": 1.2769,
+      "step": 10045
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.8861342072486877,
+      "learning_rate": 4.726394963857129e-05,
+      "loss": 1.137,
+      "step": 10050
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 1.0285983085632324,
+      "learning_rate": 4.72612765682737e-05,
+      "loss": 1.3641,
+      "step": 10055
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 1.3051286935806274,
+      "learning_rate": 4.725860226850896e-05,
+      "loss": 1.2869,
+      "step": 10060
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 1.4701013565063477,
+      "learning_rate": 4.725592673942473e-05,
+      "loss": 1.3181,
+      "step": 10065
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 1.0764509439468384,
+      "learning_rate": 4.725324998116881e-05,
+      "loss": 1.1691,
+      "step": 10070
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.9032073616981506,
+      "learning_rate": 4.7250571993889025e-05,
+      "loss": 1.181,
+      "step": 10075
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.8329408168792725,
+      "learning_rate": 4.724789277773328e-05,
+      "loss": 1.2904,
+      "step": 10080
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.2701138257980347,
+      "learning_rate": 4.7245212332849544e-05,
+      "loss": 1.3036,
+      "step": 10085
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.293056607246399,
+      "learning_rate": 4.7242530659385845e-05,
+      "loss": 1.4167,
+      "step": 10090
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.7726661562919617,
+      "learning_rate": 4.72398477574903e-05,
+      "loss": 1.3607,
+      "step": 10095
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.5324546098709106,
+      "learning_rate": 4.7237163627311084e-05,
+      "loss": 1.475,
+      "step": 10100
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.288621187210083,
+      "learning_rate": 4.723447826899644e-05,
+      "loss": 1.3092,
+      "step": 10105
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.084633469581604,
+      "learning_rate": 4.723179168269466e-05,
+      "loss": 1.2224,
+      "step": 10110
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 1.096497654914856,
+      "learning_rate": 4.7229103868554136e-05,
+      "loss": 1.2714,
+      "step": 10115
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 1.8230210542678833,
+      "learning_rate": 4.722641482672332e-05,
+      "loss": 1.3204,
+      "step": 10120
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 2.226288080215454,
+      "learning_rate": 4.7223724557350714e-05,
+      "loss": 1.3604,
+      "step": 10125
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.8786224722862244,
+      "learning_rate": 4.7221033060584897e-05,
+      "loss": 1.3312,
+      "step": 10130
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.790179431438446,
+      "learning_rate": 4.7218340336574516e-05,
+      "loss": 1.2433,
+      "step": 10135
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.7193413376808167,
+      "learning_rate": 4.7215646385468304e-05,
+      "loss": 1.1751,
+      "step": 10140
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 1.1704049110412598,
+      "learning_rate": 4.721295120741503e-05,
+      "loss": 1.2481,
+      "step": 10145
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 1.1936416625976562,
+      "learning_rate": 4.7210254802563547e-05,
+      "loss": 1.382,
+      "step": 10150
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 1.2714142799377441,
+      "learning_rate": 4.7207557171062784e-05,
+      "loss": 1.3151,
+      "step": 10155
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 1.973885178565979,
+      "learning_rate": 4.7204858313061715e-05,
+      "loss": 1.3476,
+      "step": 10160
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 1.704947829246521,
+      "learning_rate": 4.7202158228709404e-05,
+      "loss": 1.2753,
+      "step": 10165
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 1.6987488269805908,
+      "learning_rate": 4.719945691815498e-05,
+      "loss": 1.3038,
+      "step": 10170
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 1.8484985828399658,
+      "learning_rate": 4.719675438154761e-05,
+      "loss": 1.2721,
+      "step": 10175
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.5991469621658325,
+      "learning_rate": 4.719405061903658e-05,
+      "loss": 1.2864,
+      "step": 10180
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.9577586054801941,
+      "learning_rate": 4.71913456307712e-05,
+      "loss": 1.2446,
+      "step": 10185
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.2344039678573608,
+      "learning_rate": 4.718863941690087e-05,
+      "loss": 1.3046,
+      "step": 10190
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.5823438167572021,
+      "learning_rate": 4.7185931977575046e-05,
+      "loss": 1.2316,
+      "step": 10195
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.7910969257354736,
+      "learning_rate": 4.7183223312943257e-05,
+      "loss": 1.3293,
+      "step": 10200
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.5367566347122192,
+      "learning_rate": 4.7180513423155105e-05,
+      "loss": 1.2248,
+      "step": 10205
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.397963047027588,
+      "learning_rate": 4.717780230836025e-05,
+      "loss": 1.3568,
+      "step": 10210
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 1.2412103414535522,
+      "learning_rate": 4.717508996870843e-05,
+      "loss": 1.2943,
+      "step": 10215
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 2.0877139568328857,
+      "learning_rate": 4.7172376404349436e-05,
+      "loss": 1.415,
+      "step": 10220
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 1.540336012840271,
+      "learning_rate": 4.716966161543315e-05,
+      "loss": 1.1981,
+      "step": 10225
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 1.3535550832748413,
+      "learning_rate": 4.716694560210949e-05,
+      "loss": 1.2493,
+      "step": 10230
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 1.6786630153656006,
+      "learning_rate": 4.7164228364528464e-05,
+      "loss": 1.2774,
+      "step": 10235
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 1.0593065023422241,
+      "learning_rate": 4.716150990284015e-05,
+      "loss": 1.1643,
+      "step": 10240
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 1.656699299812317,
+      "learning_rate": 4.715879021719467e-05,
+      "loss": 1.2421,
+      "step": 10245
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 1.0437387228012085,
+      "learning_rate": 4.7156069307742244e-05,
+      "loss": 1.2925,
+      "step": 10250
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 1.7811616659164429,
+      "learning_rate": 4.7153347174633145e-05,
+      "loss": 1.3529,
+      "step": 10255
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 1.5187066793441772,
+      "learning_rate": 4.71506238180177e-05,
+      "loss": 1.5044,
+      "step": 10260
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 2.582772731781006,
+      "learning_rate": 4.714789923804633e-05,
+      "loss": 1.4396,
+      "step": 10265
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 1.7767497301101685,
+      "learning_rate": 4.714517343486951e-05,
+      "loss": 1.2826,
+      "step": 10270
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 1.2945784330368042,
+      "learning_rate": 4.7142446408637774e-05,
+      "loss": 1.2212,
+      "step": 10275
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.8587417006492615,
+      "learning_rate": 4.7139718159501747e-05,
+      "loss": 1.2171,
+      "step": 10280
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 3.030939817428589,
+      "learning_rate": 4.71369886876121e-05,
+      "loss": 1.3301,
+      "step": 10285
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 1.777191162109375,
+      "learning_rate": 4.7134257993119564e-05,
+      "loss": 1.2228,
+      "step": 10290
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 1.1202077865600586,
+      "learning_rate": 4.713152607617497e-05,
+      "loss": 1.2215,
+      "step": 10295
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.6653339862823486,
+      "learning_rate": 4.71287929369292e-05,
+      "loss": 1.4115,
+      "step": 10300
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 1.3566327095031738,
+      "learning_rate": 4.712605857553319e-05,
+      "loss": 1.332,
+      "step": 10305
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 1.673338532447815,
+      "learning_rate": 4.7123322992137975e-05,
+      "loss": 1.4227,
+      "step": 10310
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.8734028339385986,
+      "learning_rate": 4.712058618689462e-05,
+      "loss": 1.2329,
+      "step": 10315
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 1.6159021854400635,
+      "learning_rate": 4.7117848159954294e-05,
+      "loss": 1.1735,
+      "step": 10320
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 1.456439733505249,
+      "learning_rate": 4.7115108911468194e-05,
+      "loss": 1.3611,
+      "step": 10325
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 1.3023260831832886,
+      "learning_rate": 4.7112368441587615e-05,
+      "loss": 1.3157,
+      "step": 10330
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.9196366667747498,
+      "learning_rate": 4.710962675046391e-05,
+      "loss": 1.3092,
+      "step": 10335
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.8758362531661987,
+      "learning_rate": 4.7106883838248505e-05,
+      "loss": 1.4856,
+      "step": 10340
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 1.404257893562317,
+      "learning_rate": 4.710413970509289e-05,
+      "loss": 1.3546,
+      "step": 10345
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 2.7018842697143555,
+      "learning_rate": 4.710139435114861e-05,
+      "loss": 1.196,
+      "step": 10350
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 1.2335234880447388,
+      "learning_rate": 4.709864777656729e-05,
+      "loss": 1.145,
+      "step": 10355
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 2.3632590770721436,
+      "learning_rate": 4.709589998150063e-05,
+      "loss": 1.4927,
+      "step": 10360
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.9068330526351929,
+      "learning_rate": 4.709315096610038e-05,
+      "loss": 1.2923,
+      "step": 10365
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 1.093493103981018,
+      "learning_rate": 4.709040073051837e-05,
+      "loss": 1.2462,
+      "step": 10370
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.9117778539657593,
+      "learning_rate": 4.7087649274906475e-05,
+      "loss": 1.2301,
+      "step": 10375
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 1.3600761890411377,
+      "learning_rate": 4.7084896599416685e-05,
+      "loss": 1.3659,
+      "step": 10380
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 2.149157762527466,
+      "learning_rate": 4.7082142704200996e-05,
+      "loss": 1.2298,
+      "step": 10385
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 1.4319911003112793,
+      "learning_rate": 4.707938758941153e-05,
+      "loss": 1.2099,
+      "step": 10390
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.8007309436798096,
+      "learning_rate": 4.7076631255200436e-05,
+      "loss": 1.208,
+      "step": 10395
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 2.5758299827575684,
+      "learning_rate": 4.707387370171995e-05,
+      "loss": 1.457,
+      "step": 10400
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 1.3891574144363403,
+      "learning_rate": 4.707111492912235e-05,
+      "loss": 1.1751,
+      "step": 10405
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 1.3793649673461914,
+      "learning_rate": 4.7068354937560026e-05,
+      "loss": 1.2248,
+      "step": 10410
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 1.7429753541946411,
+      "learning_rate": 4.7065593727185395e-05,
+      "loss": 1.3831,
+      "step": 10415
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.9534806609153748,
+      "learning_rate": 4.706283129815095e-05,
+      "loss": 1.3393,
+      "step": 10420
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.9170734286308289,
+      "learning_rate": 4.706006765060928e-05,
+      "loss": 1.2613,
+      "step": 10425
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 1.9768500328063965,
+      "learning_rate": 4.7057302784713e-05,
+      "loss": 1.3073,
+      "step": 10430
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 1.7326947450637817,
+      "learning_rate": 4.705453670061481e-05,
+      "loss": 1.2648,
+      "step": 10435
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 4.0265069007873535,
+      "learning_rate": 4.7051769398467484e-05,
+      "loss": 1.3263,
+      "step": 10440
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 1.0905771255493164,
+      "learning_rate": 4.7049000878423856e-05,
+      "loss": 1.1149,
+      "step": 10445
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 2.596532106399536,
+      "learning_rate": 4.7046231140636826e-05,
+      "loss": 1.3235,
+      "step": 10450
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 1.497429370880127,
+      "learning_rate": 4.704346018525937e-05,
+      "loss": 1.3615,
+      "step": 10455
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.9473041892051697,
+      "learning_rate": 4.704068801244452e-05,
+      "loss": 1.2733,
+      "step": 10460
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.628065824508667,
+      "learning_rate": 4.703791462234537e-05,
+      "loss": 1.3351,
+      "step": 10465
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.9832010865211487,
+      "learning_rate": 4.703514001511512e-05,
+      "loss": 1.3246,
+      "step": 10470
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 1.0642294883728027,
+      "learning_rate": 4.7032364190906985e-05,
+      "loss": 1.309,
+      "step": 10475
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 2.5547728538513184,
+      "learning_rate": 4.702958714987427e-05,
+      "loss": 1.2542,
+      "step": 10480
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 2.006378173828125,
+      "learning_rate": 4.702680889217036e-05,
+      "loss": 1.2813,
+      "step": 10485
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.6523733735084534,
+      "learning_rate": 4.702402941794869e-05,
+      "loss": 1.2206,
+      "step": 10490
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 1.642492651939392,
+      "learning_rate": 4.702124872736277e-05,
+      "loss": 1.3215,
+      "step": 10495
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 3.4225122928619385,
+      "learning_rate": 4.7018466820566174e-05,
+      "loss": 1.2342,
+      "step": 10500
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 1.2977136373519897,
+      "learning_rate": 4.701568369771254e-05,
+      "loss": 1.3561,
+      "step": 10505
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 2.203577995300293,
+      "learning_rate": 4.701289935895558e-05,
+      "loss": 1.2129,
+      "step": 10510
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.4925588369369507,
+      "learning_rate": 4.701011380444907e-05,
+      "loss": 1.2579,
+      "step": 10515
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 2.624371290206909,
+      "learning_rate": 4.700732703434685e-05,
+      "loss": 1.4046,
+      "step": 10520
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.6034959554672241,
+      "learning_rate": 4.7004539048802834e-05,
+      "loss": 1.2614,
+      "step": 10525
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.3191356658935547,
+      "learning_rate": 4.7001749847971e-05,
+      "loss": 1.2469,
+      "step": 10530
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.2245357036590576,
+      "learning_rate": 4.6998959432005393e-05,
+      "loss": 1.3599,
+      "step": 10535
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.8731427192687988,
+      "learning_rate": 4.699616780106012e-05,
+      "loss": 1.2735,
+      "step": 10540
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 2.4979584217071533,
+      "learning_rate": 4.699337495528937e-05,
+      "loss": 1.3179,
+      "step": 10545
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.922391414642334,
+      "learning_rate": 4.699058089484737e-05,
+      "loss": 1.3218,
+      "step": 10550
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.5169897079467773,
+      "learning_rate": 4.6987785619888455e-05,
+      "loss": 1.3695,
+      "step": 10555
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.7197904586791992,
+      "learning_rate": 4.698498913056699e-05,
+      "loss": 1.3514,
+      "step": 10560
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.886755108833313,
+      "learning_rate": 4.698219142703743e-05,
+      "loss": 1.3594,
+      "step": 10565
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 2.066739082336426,
+      "learning_rate": 4.6979392509454286e-05,
+      "loss": 1.2485,
+      "step": 10570
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.8264986276626587,
+      "learning_rate": 4.697659237797214e-05,
+      "loss": 1.2989,
+      "step": 10575
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 1.5807194709777832,
+      "learning_rate": 4.697379103274564e-05,
+      "loss": 1.1532,
+      "step": 10580
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 2.379920244216919,
+      "learning_rate": 4.69709884739295e-05,
+      "loss": 1.1664,
+      "step": 10585
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 2.331644296646118,
+      "learning_rate": 4.696818470167851e-05,
+      "loss": 1.2713,
+      "step": 10590
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 2.402665376663208,
+      "learning_rate": 4.6965379716147504e-05,
+      "loss": 1.3309,
+      "step": 10595
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 3.28230881690979,
+      "learning_rate": 4.6962573517491414e-05,
+      "loss": 1.2569,
+      "step": 10600
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 1.6851478815078735,
+      "learning_rate": 4.6959766105865225e-05,
+      "loss": 1.4249,
+      "step": 10605
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 3.4568557739257812,
+      "learning_rate": 4.695695748142397e-05,
+      "loss": 1.4145,
+      "step": 10610
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.6388516426086426,
+      "learning_rate": 4.695414764432278e-05,
+      "loss": 1.3806,
+      "step": 10615
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 2.0879287719726562,
+      "learning_rate": 4.695133659471683e-05,
+      "loss": 1.2937,
+      "step": 10620
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 1.9023692607879639,
+      "learning_rate": 4.694852433276138e-05,
+      "loss": 1.3197,
+      "step": 10625
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.421626329421997,
+      "learning_rate": 4.6945710858611746e-05,
+      "loss": 1.2769,
+      "step": 10630
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 1.5215307474136353,
+      "learning_rate": 4.694289617242331e-05,
+      "loss": 1.3148,
+      "step": 10635
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 1.3588393926620483,
+      "learning_rate": 4.694008027435154e-05,
+      "loss": 1.2765,
+      "step": 10640
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 1.517864465713501,
+      "learning_rate": 4.6937263164551926e-05,
+      "loss": 1.1829,
+      "step": 10645
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 1.18218994140625,
+      "learning_rate": 4.693444484318008e-05,
+      "loss": 1.3342,
+      "step": 10650
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.9616156816482544,
+      "learning_rate": 4.693162531039164e-05,
+      "loss": 1.4308,
+      "step": 10655
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 1.3494954109191895,
+      "learning_rate": 4.692880456634233e-05,
+      "loss": 1.2483,
+      "step": 10660
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 1.6697038412094116,
+      "learning_rate": 4.6925982611187934e-05,
+      "loss": 1.2745,
+      "step": 10665
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 1.6610711812973022,
+      "learning_rate": 4.692315944508432e-05,
+      "loss": 1.3191,
+      "step": 10670
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 2.5130321979522705,
+      "learning_rate": 4.692033506818739e-05,
+      "loss": 1.0889,
+      "step": 10675
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.1333353519439697,
+      "learning_rate": 4.6917509480653146e-05,
+      "loss": 1.1744,
+      "step": 10680
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.9024247527122498,
+      "learning_rate": 4.6914682682637626e-05,
+      "loss": 1.2681,
+      "step": 10685
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.6913889646530151,
+      "learning_rate": 4.6911854674296964e-05,
+      "loss": 1.246,
+      "step": 10690
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 2.867292881011963,
+      "learning_rate": 4.690902545578735e-05,
+      "loss": 1.3205,
+      "step": 10695
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.3072094917297363,
+      "learning_rate": 4.690619502726502e-05,
+      "loss": 1.3059,
+      "step": 10700
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.9709635972976685,
+      "learning_rate": 4.6903363388886325e-05,
+      "loss": 1.321,
+      "step": 10705
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.7038654088974,
+      "learning_rate": 4.6900530540807624e-05,
+      "loss": 1.3357,
+      "step": 10710
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.885414183139801,
+      "learning_rate": 4.68976964831854e-05,
+      "loss": 1.2921,
+      "step": 10715
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.5345251560211182,
+      "learning_rate": 4.689486121617615e-05,
+      "loss": 1.4398,
+      "step": 10720
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.8840970993041992,
+      "learning_rate": 4.689202473993647e-05,
+      "loss": 1.1967,
+      "step": 10725
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.861816167831421,
+      "learning_rate": 4.6889187054623016e-05,
+      "loss": 1.2309,
+      "step": 10730
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.6276249885559082,
+      "learning_rate": 4.688634816039253e-05,
+      "loss": 1.2227,
+      "step": 10735
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.3406356573104858,
+      "learning_rate": 4.688350805740178e-05,
+      "loss": 1.3609,
+      "step": 10740
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 2.407961130142212,
+      "learning_rate": 4.6880666745807625e-05,
+      "loss": 1.0896,
+      "step": 10745
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.9864040613174438,
+      "learning_rate": 4.687782422576698e-05,
+      "loss": 1.2063,
+      "step": 10750
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.8912273645401,
+      "learning_rate": 4.687498049743685e-05,
+      "loss": 1.2833,
+      "step": 10755
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.490443468093872,
+      "learning_rate": 4.6872135560974285e-05,
+      "loss": 1.3106,
+      "step": 10760
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 2.050848960876465,
+      "learning_rate": 4.686928941653641e-05,
+      "loss": 1.2577,
+      "step": 10765
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.508426308631897,
+      "learning_rate": 4.686644206428041e-05,
+      "loss": 1.2863,
+      "step": 10770
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.549778699874878,
+      "learning_rate": 4.686359350436354e-05,
+      "loss": 1.1178,
+      "step": 10775
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 1.004459023475647,
+      "learning_rate": 4.6860743736943134e-05,
+      "loss": 1.3404,
+      "step": 10780
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 3.65075945854187,
+      "learning_rate": 4.685789276217658e-05,
+      "loss": 1.449,
+      "step": 10785
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.863338053226471,
+      "learning_rate": 4.6855040580221323e-05,
+      "loss": 1.1967,
+      "step": 10790
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 1.1816414594650269,
+      "learning_rate": 4.685218719123489e-05,
+      "loss": 1.2813,
+      "step": 10795
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 1.7856791019439697,
+      "learning_rate": 4.6849332595374864e-05,
+      "loss": 1.4305,
+      "step": 10800
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 1.4339046478271484,
+      "learning_rate": 4.684647679279892e-05,
+      "loss": 1.3854,
+      "step": 10805
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.8699505925178528,
+      "learning_rate": 4.684361978366477e-05,
+      "loss": 1.2915,
+      "step": 10810
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.3646979331970215,
+      "learning_rate": 4.6840761568130206e-05,
+      "loss": 1.3623,
+      "step": 10815
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 2.1360511779785156,
+      "learning_rate": 4.6837902146353076e-05,
+      "loss": 1.2835,
+      "step": 10820
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.1500669717788696,
+      "learning_rate": 4.683504151849132e-05,
+      "loss": 1.2727,
+      "step": 10825
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 4.0960187911987305,
+      "learning_rate": 4.68321796847029e-05,
+      "loss": 1.2047,
+      "step": 10830
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.9620679020881653,
+      "learning_rate": 4.68293166451459e-05,
+      "loss": 1.2046,
+      "step": 10835
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.1444592475891113,
+      "learning_rate": 4.6826452399978436e-05,
+      "loss": 1.391,
+      "step": 10840
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 2.369436740875244,
+      "learning_rate": 4.6823586949358686e-05,
+      "loss": 1.2937,
+      "step": 10845
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 1.327525019645691,
+      "learning_rate": 4.682072029344492e-05,
+      "loss": 1.4105,
+      "step": 10850
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 1.3739736080169678,
+      "learning_rate": 4.681785243239545e-05,
+      "loss": 1.4381,
+      "step": 10855
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 1.522316813468933,
+      "learning_rate": 4.681498336636867e-05,
+      "loss": 1.3176,
+      "step": 10860
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 1.123241901397705,
+      "learning_rate": 4.681211309552304e-05,
+      "loss": 1.2045,
+      "step": 10865
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.9612917900085449,
+      "learning_rate": 4.680924162001706e-05,
+      "loss": 1.3388,
+      "step": 10870
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 1.3160079717636108,
+      "learning_rate": 4.680636894000935e-05,
+      "loss": 1.338,
+      "step": 10875
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.9297557473182678,
+      "learning_rate": 4.680349505565854e-05,
+      "loss": 1.2748,
+      "step": 10880
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.926292359828949,
+      "learning_rate": 4.6800619967123373e-05,
+      "loss": 1.3359,
+      "step": 10885
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 1.0277527570724487,
+      "learning_rate": 4.679774367456261e-05,
+      "loss": 1.2692,
+      "step": 10890
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 1.3001703023910522,
+      "learning_rate": 4.679486617813513e-05,
+      "loss": 1.2465,
+      "step": 10895
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 1.934599757194519,
+      "learning_rate": 4.679198747799984e-05,
+      "loss": 1.318,
+      "step": 10900
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 1.1883105039596558,
+      "learning_rate": 4.678910757431574e-05,
+      "loss": 1.2774,
+      "step": 10905
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 2.791804075241089,
+      "learning_rate": 4.678622646724188e-05,
+      "loss": 1.3288,
+      "step": 10910
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 1.5142138004302979,
+      "learning_rate": 4.678334415693737e-05,
+      "loss": 1.444,
+      "step": 10915
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 1.1314914226531982,
+      "learning_rate": 4.678046064356141e-05,
+      "loss": 1.3735,
+      "step": 10920
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 2.2148430347442627,
+      "learning_rate": 4.677757592727325e-05,
+      "loss": 1.2055,
+      "step": 10925
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 1.7565327882766724,
+      "learning_rate": 4.67746900082322e-05,
+      "loss": 1.2262,
+      "step": 10930
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 4.066123962402344,
+      "learning_rate": 4.677180288659766e-05,
+      "loss": 1.3309,
+      "step": 10935
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 2.2226760387420654,
+      "learning_rate": 4.676891456252908e-05,
+      "loss": 1.3966,
+      "step": 10940
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 2.1616628170013428,
+      "learning_rate": 4.676602503618597e-05,
+      "loss": 1.1928,
+      "step": 10945
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 1.3349906206130981,
+      "learning_rate": 4.6763134307727916e-05,
+      "loss": 1.2268,
+      "step": 10950
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 1.6452038288116455,
+      "learning_rate": 4.676024237731459e-05,
+      "loss": 1.1698,
+      "step": 10955
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 1.8451074361801147,
+      "learning_rate": 4.675734924510569e-05,
+      "loss": 1.2763,
+      "step": 10960
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 1.9175277948379517,
+      "learning_rate": 4.6754454911261005e-05,
+      "loss": 1.3425,
+      "step": 10965
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 1.0430094003677368,
+      "learning_rate": 4.6751559375940384e-05,
+      "loss": 1.324,
+      "step": 10970
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 4.597285747528076,
+      "learning_rate": 4.674866263930375e-05,
+      "loss": 1.2731,
+      "step": 10975
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.3398327827453613,
+      "learning_rate": 4.674576470151109e-05,
+      "loss": 1.3967,
+      "step": 10980
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.1260203123092651,
+      "learning_rate": 4.674344548655217e-05,
+      "loss": 1.4702,
+      "step": 10985
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.8859392404556274,
+      "learning_rate": 4.674054538708202e-05,
+      "loss": 1.3604,
+      "step": 10990
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.0552682876586914,
+      "learning_rate": 4.6737644086904156e-05,
+      "loss": 1.2871,
+      "step": 10995
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.1921944618225098,
+      "learning_rate": 4.6734741586178794e-05,
+      "loss": 1.411,
+      "step": 11000
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.534670352935791,
+      "learning_rate": 4.673183788506625e-05,
+      "loss": 1.1321,
+      "step": 11005
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.8128979802131653,
+      "learning_rate": 4.672893298372688e-05,
+      "loss": 1.1416,
+      "step": 11010
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.2440390586853027,
+      "learning_rate": 4.672602688232114e-05,
+      "loss": 1.3895,
+      "step": 11015
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.7895771861076355,
+      "learning_rate": 4.672311958100951e-05,
+      "loss": 1.2575,
+      "step": 11020
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 8.737719535827637,
+      "learning_rate": 4.672021107995257e-05,
+      "loss": 1.2813,
+      "step": 11025
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.1101980209350586,
+      "learning_rate": 4.671730137931095e-05,
+      "loss": 1.3107,
+      "step": 11030
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.0142308473587036,
+      "learning_rate": 4.671439047924535e-05,
+      "loss": 1.2947,
+      "step": 11035
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.9950822591781616,
+      "learning_rate": 4.671147837991653e-05,
+      "loss": 1.4065,
+      "step": 11040
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.4179012775421143,
+      "learning_rate": 4.670856508148532e-05,
+      "loss": 1.3455,
+      "step": 11045
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 1.6330708265304565,
+      "learning_rate": 4.670565058411264e-05,
+      "loss": 1.3063,
+      "step": 11050
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 1.7123674154281616,
+      "learning_rate": 4.6702734887959434e-05,
+      "loss": 1.2452,
+      "step": 11055
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 1.1619246006011963,
+      "learning_rate": 4.669981799318674e-05,
+      "loss": 1.361,
+      "step": 11060
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 2.159376382827759,
+      "learning_rate": 4.669689989995565e-05,
+      "loss": 1.2923,
+      "step": 11065
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 1.8677573204040527,
+      "learning_rate": 4.6693980608427326e-05,
+      "loss": 1.2485,
+      "step": 11070
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 1.9316105842590332,
+      "learning_rate": 4.669106011876301e-05,
+      "loss": 1.2627,
+      "step": 11075
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.5687687397003174,
+      "learning_rate": 4.6688138431124e-05,
+      "loss": 1.5126,
+      "step": 11080
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.8287721872329712,
+      "learning_rate": 4.6685215545671634e-05,
+      "loss": 1.1865,
+      "step": 11085
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.737056851387024,
+      "learning_rate": 4.668229146256735e-05,
+      "loss": 1.2676,
+      "step": 11090
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.2314867973327637,
+      "learning_rate": 4.667936618197266e-05,
+      "loss": 1.2986,
+      "step": 11095
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 2.5323379039764404,
+      "learning_rate": 4.66764397040491e-05,
+      "loss": 1.2406,
+      "step": 11100
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 2.0463790893554688,
+      "learning_rate": 4.667351202895831e-05,
+      "loss": 1.3175,
+      "step": 11105
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.8490899801254272,
+      "learning_rate": 4.6670583156861984e-05,
+      "loss": 1.273,
+      "step": 11110
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 2.06270694732666,
+      "learning_rate": 4.666765308792187e-05,
+      "loss": 1.2421,
+      "step": 11115
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 2.100942611694336,
+      "learning_rate": 4.66647218222998e-05,
+      "loss": 1.3169,
+      "step": 11120
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 2.400592565536499,
+      "learning_rate": 4.666178936015767e-05,
+      "loss": 1.4987,
+      "step": 11125
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 1.5925202369689941,
+      "learning_rate": 4.665885570165742e-05,
+      "loss": 1.2487,
+      "step": 11130
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.8102724552154541,
+      "learning_rate": 4.665592084696108e-05,
+      "loss": 1.2773,
+      "step": 11135
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.9131859540939331,
+      "learning_rate": 4.665298479623075e-05,
+      "loss": 1.0861,
+      "step": 11140
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 2.433790445327759,
+      "learning_rate": 4.665004754962857e-05,
+      "loss": 1.1546,
+      "step": 11145
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 1.100435495376587,
+      "learning_rate": 4.664710910731677e-05,
+      "loss": 1.2528,
+      "step": 11150
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 3.4649596214294434,
+      "learning_rate": 4.6644169469457635e-05,
+      "loss": 1.3513,
+      "step": 11155
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 1.005937933921814,
+      "learning_rate": 4.664122863621352e-05,
+      "loss": 1.2872,
+      "step": 11160
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 1.2074639797210693,
+      "learning_rate": 4.663828660774684e-05,
+      "loss": 1.341,
+      "step": 11165
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 2.826364040374756,
+      "learning_rate": 4.663534338422009e-05,
+      "loss": 1.2848,
+      "step": 11170
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 2.759699821472168,
+      "learning_rate": 4.663239896579581e-05,
+      "loss": 1.4064,
+      "step": 11175
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 2.0414791107177734,
+      "learning_rate": 4.662945335263662e-05,
+      "loss": 1.3405,
+      "step": 11180
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.7230271100997925,
+      "learning_rate": 4.6626506544905194e-05,
+      "loss": 1.3063,
+      "step": 11185
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.979017972946167,
+      "learning_rate": 4.66235585427643e-05,
+      "loss": 1.2698,
+      "step": 11190
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.1741644144058228,
+      "learning_rate": 4.662060934637674e-05,
+      "loss": 1.1668,
+      "step": 11195
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.050943374633789,
+      "learning_rate": 4.661765895590541e-05,
+      "loss": 1.3493,
+      "step": 11200
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 3.3978419303894043,
+      "learning_rate": 4.661470737151323e-05,
+      "loss": 1.2133,
+      "step": 11205
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.509317398071289,
+      "learning_rate": 4.661175459336324e-05,
+      "loss": 1.3368,
+      "step": 11210
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 2.1096768379211426,
+      "learning_rate": 4.6608800621618506e-05,
+      "loss": 1.3254,
+      "step": 11215
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 1.4031864404678345,
+      "learning_rate": 4.660584545644218e-05,
+      "loss": 1.4675,
+      "step": 11220
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 1.8638592958450317,
+      "learning_rate": 4.660288909799746e-05,
+      "loss": 1.4674,
+      "step": 11225
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 1.681207299232483,
+      "learning_rate": 4.659993154644763e-05,
+      "loss": 1.1908,
+      "step": 11230
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 1.484300136566162,
+      "learning_rate": 4.659697280195604e-05,
+      "loss": 1.1835,
+      "step": 11235
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 1.0395090579986572,
+      "learning_rate": 4.6594012864686084e-05,
+      "loss": 1.2898,
+      "step": 11240
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 2.5014450550079346,
+      "learning_rate": 4.6591051734801246e-05,
+      "loss": 1.2174,
+      "step": 11245
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 2.0165107250213623,
+      "learning_rate": 4.658808941246506e-05,
+      "loss": 1.2818,
+      "step": 11250
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 1.2676451206207275,
+      "learning_rate": 4.658512589784114e-05,
+      "loss": 1.3866,
+      "step": 11255
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 1.7123560905456543,
+      "learning_rate": 4.658216119109315e-05,
+      "loss": 1.4062,
+      "step": 11260
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 2.4512739181518555,
+      "learning_rate": 4.6579195292384825e-05,
+      "loss": 1.2707,
+      "step": 11265
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 1.6923884153366089,
+      "learning_rate": 4.657622820187998e-05,
+      "loss": 1.3526,
+      "step": 11270
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 2.5449416637420654,
+      "learning_rate": 4.6573259919742484e-05,
+      "loss": 1.1957,
+      "step": 11275
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 1.0603396892547607,
+      "learning_rate": 4.657029044613626e-05,
+      "loss": 1.4085,
+      "step": 11280
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 2.8096063137054443,
+      "learning_rate": 4.6567319781225313e-05,
+      "loss": 1.4052,
+      "step": 11285
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 2.0202414989471436,
+      "learning_rate": 4.656434792517372e-05,
+      "loss": 1.3356,
+      "step": 11290
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.3925175666809082,
+      "learning_rate": 4.6561374878145606e-05,
+      "loss": 1.2098,
+      "step": 11295
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.2433990240097046,
+      "learning_rate": 4.655840064030517e-05,
+      "loss": 1.3263,
+      "step": 11300
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.0137583017349243,
+      "learning_rate": 4.6555425211816675e-05,
+      "loss": 1.2833,
+      "step": 11305
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 2.3041789531707764,
+      "learning_rate": 4.655244859284444e-05,
+      "loss": 1.4461,
+      "step": 11310
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 1.4320485591888428,
+      "learning_rate": 4.6549470783552886e-05,
+      "loss": 1.3514,
+      "step": 11315
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 1.6517833471298218,
+      "learning_rate": 4.654649178410645e-05,
+      "loss": 1.3607,
+      "step": 11320
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 2.3063910007476807,
+      "learning_rate": 4.6543511594669675e-05,
+      "loss": 1.2331,
+      "step": 11325
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 1.5468136072158813,
+      "learning_rate": 4.654053021540714e-05,
+      "loss": 1.207,
+      "step": 11330
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.7770300507545471,
+      "learning_rate": 4.653754764648352e-05,
+      "loss": 1.2422,
+      "step": 11335
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 1.037409782409668,
+      "learning_rate": 4.6534563888063534e-05,
+      "loss": 1.1782,
+      "step": 11340
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 1.4780035018920898,
+      "learning_rate": 4.653157894031196e-05,
+      "loss": 1.281,
+      "step": 11345
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.644164800643921,
+      "learning_rate": 4.652859280339366e-05,
+      "loss": 1.2121,
+      "step": 11350
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.3956807851791382,
+      "learning_rate": 4.6525605477473564e-05,
+      "loss": 1.2194,
+      "step": 11355
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.0469894409179688,
+      "learning_rate": 4.6522616962716646e-05,
+      "loss": 1.3257,
+      "step": 11360
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.9847218990325928,
+      "learning_rate": 4.651962725928797e-05,
+      "loss": 1.4013,
+      "step": 11365
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.7343077659606934,
+      "learning_rate": 4.6516636367352646e-05,
+      "loss": 1.3517,
+      "step": 11370
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 2.5026791095733643,
+      "learning_rate": 4.6513644287075866e-05,
+      "loss": 1.4136,
+      "step": 11375
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 2.037338972091675,
+      "learning_rate": 4.651065101862286e-05,
+      "loss": 1.3706,
+      "step": 11380
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 2.3901917934417725,
+      "learning_rate": 4.650765656215898e-05,
+      "loss": 1.2601,
+      "step": 11385
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 1.2326328754425049,
+      "learning_rate": 4.650466091784956e-05,
+      "loss": 1.4596,
+      "step": 11390
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 1.7991396188735962,
+      "learning_rate": 4.650166408586009e-05,
+      "loss": 1.2718,
+      "step": 11395
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 1.9020781517028809,
+      "learning_rate": 4.649866606635605e-05,
+      "loss": 1.2791,
+      "step": 11400
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 1.2564764022827148,
+      "learning_rate": 4.649566685950304e-05,
+      "loss": 1.2263,
+      "step": 11405
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 1.6249325275421143,
+      "learning_rate": 4.649266646546668e-05,
+      "loss": 1.3176,
+      "step": 11410
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 1.1263751983642578,
+      "learning_rate": 4.64896648844127e-05,
+      "loss": 1.3471,
+      "step": 11415
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 1.0438363552093506,
+      "learning_rate": 4.648666211650686e-05,
+      "loss": 1.3099,
+      "step": 11420
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 1.0559061765670776,
+      "learning_rate": 4.648365816191501e-05,
+      "loss": 1.2612,
+      "step": 11425
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.788719654083252,
+      "learning_rate": 4.648065302080305e-05,
+      "loss": 1.2446,
+      "step": 11430
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 2.290250539779663,
+      "learning_rate": 4.647764669333695e-05,
+      "loss": 1.2215,
+      "step": 11435
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 1.0487291812896729,
+      "learning_rate": 4.647463917968275e-05,
+      "loss": 1.1949,
+      "step": 11440
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 1.4382762908935547,
+      "learning_rate": 4.647163048000655e-05,
+      "loss": 1.2458,
+      "step": 11445
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 2.330712080001831,
+      "learning_rate": 4.6468620594474515e-05,
+      "loss": 1.2711,
+      "step": 11450
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 3.3765647411346436,
+      "learning_rate": 4.6465609523252884e-05,
+      "loss": 1.359,
+      "step": 11455
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.7210214138031006,
+      "learning_rate": 4.646259726650794e-05,
+      "loss": 1.2642,
+      "step": 11460
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.9674427509307861,
+      "learning_rate": 4.645958382440607e-05,
+      "loss": 1.2704,
+      "step": 11465
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.1404820680618286,
+      "learning_rate": 4.645656919711369e-05,
+      "loss": 1.3052,
+      "step": 11470
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.92625892162323,
+      "learning_rate": 4.645355338479729e-05,
+      "loss": 1.1419,
+      "step": 11475
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 2.301802396774292,
+      "learning_rate": 4.645053638762344e-05,
+      "loss": 1.2296,
+      "step": 11480
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 2.9896774291992188,
+      "learning_rate": 4.6447518205758765e-05,
+      "loss": 1.3242,
+      "step": 11485
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 1.1062313318252563,
+      "learning_rate": 4.6444498839369956e-05,
+      "loss": 1.3635,
+      "step": 11490
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 1.2494499683380127,
+      "learning_rate": 4.644147828862375e-05,
+      "loss": 1.3554,
+      "step": 11495
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 4.519656658172607,
+      "learning_rate": 4.6438456553687005e-05,
+      "loss": 1.3204,
+      "step": 11500
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 1.678928256034851,
+      "learning_rate": 4.6435433634726575e-05,
+      "loss": 1.2169,
+      "step": 11505
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 1.1864182949066162,
+      "learning_rate": 4.6432409531909434e-05,
+      "loss": 1.2554,
+      "step": 11510
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.6995583772659302,
+      "learning_rate": 4.6429384245402585e-05,
+      "loss": 1.3414,
+      "step": 11515
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.1473424434661865,
+      "learning_rate": 4.642635777537312e-05,
+      "loss": 1.1982,
+      "step": 11520
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.1320407390594482,
+      "learning_rate": 4.6423330121988196e-05,
+      "loss": 1.3277,
+      "step": 11525
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.7268291115760803,
+      "learning_rate": 4.6420301285415005e-05,
+      "loss": 1.4298,
+      "step": 11530
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.2748498916625977,
+      "learning_rate": 4.641727126582085e-05,
+      "loss": 1.3528,
+      "step": 11535
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.8894537687301636,
+      "learning_rate": 4.641424006337306e-05,
+      "loss": 1.3222,
+      "step": 11540
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.8295863270759583,
+      "learning_rate": 4.641120767823905e-05,
+      "loss": 1.2138,
+      "step": 11545
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 2.017294406890869,
+      "learning_rate": 4.640817411058629e-05,
+      "loss": 1.3272,
+      "step": 11550
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 2.1451973915100098,
+      "learning_rate": 4.640513936058233e-05,
+      "loss": 1.2706,
+      "step": 11555
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 1.6634198427200317,
+      "learning_rate": 4.640210342839478e-05,
+      "loss": 1.2066,
+      "step": 11560
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 2.191540002822876,
+      "learning_rate": 4.6399066314191294e-05,
+      "loss": 1.2551,
+      "step": 11565
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 1.4875872135162354,
+      "learning_rate": 4.639602801813963e-05,
+      "loss": 1.4758,
+      "step": 11570
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 1.266520619392395,
+      "learning_rate": 4.6392988540407564e-05,
+      "loss": 1.2998,
+      "step": 11575
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 1.3954042196273804,
+      "learning_rate": 4.6389947881162984e-05,
+      "loss": 1.2636,
+      "step": 11580
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 1.5587852001190186,
+      "learning_rate": 4.6386906040573825e-05,
+      "loss": 1.2853,
+      "step": 11585
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 2.421710729598999,
+      "learning_rate": 4.638386301880807e-05,
+      "loss": 1.3729,
+      "step": 11590
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 1.5814123153686523,
+      "learning_rate": 4.638081881603378e-05,
+      "loss": 1.4307,
+      "step": 11595
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.9030851125717163,
+      "learning_rate": 4.63777734324191e-05,
+      "loss": 1.3591,
+      "step": 11600
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 2.414091110229492,
+      "learning_rate": 4.637472686813221e-05,
+      "loss": 1.2515,
+      "step": 11605
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 1.2490521669387817,
+      "learning_rate": 4.637167912334138e-05,
+      "loss": 1.3923,
+      "step": 11610
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.478395700454712,
+      "learning_rate": 4.6368630198214916e-05,
+      "loss": 1.3068,
+      "step": 11615
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.7249629497528076,
+      "learning_rate": 4.636558009292122e-05,
+      "loss": 1.4022,
+      "step": 11620
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.7434816360473633,
+      "learning_rate": 4.636252880762875e-05,
+      "loss": 1.2509,
+      "step": 11625
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.605729341506958,
+      "learning_rate": 4.6359476342506015e-05,
+      "loss": 1.2654,
+      "step": 11630
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.8081423044204712,
+      "learning_rate": 4.63564226977216e-05,
+      "loss": 1.2407,
+      "step": 11635
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.519515037536621,
+      "learning_rate": 4.6353367873444165e-05,
+      "loss": 1.427,
+      "step": 11640
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.4996379613876343,
+      "learning_rate": 4.635031186984241e-05,
+      "loss": 1.2309,
+      "step": 11645
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 1.2170788049697876,
+      "learning_rate": 4.634725468708513e-05,
+      "loss": 1.2613,
+      "step": 11650
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 1.4180010557174683,
+      "learning_rate": 4.634419632534116e-05,
+      "loss": 1.3331,
+      "step": 11655
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.9083313345909119,
+      "learning_rate": 4.6341136784779415e-05,
+      "loss": 1.3276,
+      "step": 11660
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 2.1086177825927734,
+      "learning_rate": 4.633807606556887e-05,
+      "loss": 1.2769,
+      "step": 11665
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 1.2872865200042725,
+      "learning_rate": 4.6335014167878557e-05,
+      "loss": 1.1635,
+      "step": 11670
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.8815200924873352,
+      "learning_rate": 4.633195109187759e-05,
+      "loss": 1.2593,
+      "step": 11675
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 2.032335042953491,
+      "learning_rate": 4.632888683773514e-05,
+      "loss": 1.3319,
+      "step": 11680
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 2.3916192054748535,
+      "learning_rate": 4.632582140562044e-05,
+      "loss": 1.2105,
+      "step": 11685
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.9669456481933594,
+      "learning_rate": 4.6322754795702795e-05,
+      "loss": 1.2198,
+      "step": 11690
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.4312375783920288,
+      "learning_rate": 4.6319687008151555e-05,
+      "loss": 1.2218,
+      "step": 11695
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.3405520915985107,
+      "learning_rate": 4.6316618043136175e-05,
+      "loss": 1.1403,
+      "step": 11700
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.2970833778381348,
+      "learning_rate": 4.6313547900826124e-05,
+      "loss": 1.4566,
+      "step": 11705
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.1170778274536133,
+      "learning_rate": 4.6310476581390985e-05,
+      "loss": 1.3428,
+      "step": 11710
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 2.171121835708618,
+      "learning_rate": 4.6307404085000374e-05,
+      "loss": 1.2992,
+      "step": 11715
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.2334961891174316,
+      "learning_rate": 4.630433041182398e-05,
+      "loss": 1.1957,
+      "step": 11720
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 2.142889976501465,
+      "learning_rate": 4.630125556203156e-05,
+      "loss": 1.3488,
+      "step": 11725
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.060781478881836,
+      "learning_rate": 4.629817953579295e-05,
+      "loss": 1.0665,
+      "step": 11730
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.4236716032028198,
+      "learning_rate": 4.629510233327802e-05,
+      "loss": 1.2998,
+      "step": 11735
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.210963249206543,
+      "learning_rate": 4.6292023954656716e-05,
+      "loss": 1.3734,
+      "step": 11740
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.2984216213226318,
+      "learning_rate": 4.628894440009906e-05,
+      "loss": 1.3029,
+      "step": 11745
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.2461925745010376,
+      "learning_rate": 4.628586366977513e-05,
+      "loss": 1.2483,
+      "step": 11750
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.2298219203948975,
+      "learning_rate": 4.628278176385509e-05,
+      "loss": 1.1885,
+      "step": 11755
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 2.885982036590576,
+      "learning_rate": 4.627969868250912e-05,
+      "loss": 1.2902,
+      "step": 11760
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.220379114151001,
+      "learning_rate": 4.6276614425907514e-05,
+      "loss": 1.2562,
+      "step": 11765
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.2900983095169067,
+      "learning_rate": 4.6273528994220616e-05,
+      "loss": 1.2722,
+      "step": 11770
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.6189454793930054,
+      "learning_rate": 4.627044238761882e-05,
+      "loss": 1.2755,
+      "step": 11775
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 2.1183040142059326,
+      "learning_rate": 4.62673546062726e-05,
+      "loss": 1.3765,
+      "step": 11780
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 1.5245985984802246,
+      "learning_rate": 4.6264265650352494e-05,
+      "loss": 1.3768,
+      "step": 11785
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 1.0661191940307617,
+      "learning_rate": 4.62611755200291e-05,
+      "loss": 1.2853,
+      "step": 11790
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 2.844651699066162,
+      "learning_rate": 4.625808421547307e-05,
+      "loss": 1.3109,
+      "step": 11795
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 1.5837798118591309,
+      "learning_rate": 4.6254991736855156e-05,
+      "loss": 1.192,
+      "step": 11800
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 1.2192292213439941,
+      "learning_rate": 4.625189808434614e-05,
+      "loss": 1.207,
+      "step": 11805
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 2.5890302658081055,
+      "learning_rate": 4.624880325811689e-05,
+      "loss": 1.3307,
+      "step": 11810
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 1.7649799585342407,
+      "learning_rate": 4.624570725833831e-05,
+      "loss": 1.3292,
+      "step": 11815
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 2.073655605316162,
+      "learning_rate": 4.624261008518141e-05,
+      "loss": 1.2353,
+      "step": 11820
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 1.5090842247009277,
+      "learning_rate": 4.623951173881723e-05,
+      "loss": 1.2454,
+      "step": 11825
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 1.7265774011611938,
+      "learning_rate": 4.62364122194169e-05,
+      "loss": 1.2835,
+      "step": 11830
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 2.3008291721343994,
+      "learning_rate": 4.62333115271516e-05,
+      "loss": 1.3273,
+      "step": 11835
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 1.107966661453247,
+      "learning_rate": 4.6230209662192565e-05,
+      "loss": 1.2891,
+      "step": 11840
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 1.2190146446228027,
+      "learning_rate": 4.622710662471112e-05,
+      "loss": 1.4084,
+      "step": 11845
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.5428764820098877,
+      "learning_rate": 4.6224002414878644e-05,
+      "loss": 1.1754,
+      "step": 11850
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.452290654182434,
+      "learning_rate": 4.6220897032866574e-05,
+      "loss": 1.2888,
+      "step": 11855
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.2167850732803345,
+      "learning_rate": 4.621779047884642e-05,
+      "loss": 1.3761,
+      "step": 11860
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.8653399348258972,
+      "learning_rate": 4.6214682752989746e-05,
+      "loss": 1.3106,
+      "step": 11865
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.1357594728469849,
+      "learning_rate": 4.6211573855468205e-05,
+      "loss": 1.3966,
+      "step": 11870
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 2.4457480907440186,
+      "learning_rate": 4.6208463786453485e-05,
+      "loss": 1.3881,
+      "step": 11875
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 2.0536937713623047,
+      "learning_rate": 4.6205352546117356e-05,
+      "loss": 1.3554,
+      "step": 11880
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.843001127243042,
+      "learning_rate": 4.6202240134631644e-05,
+      "loss": 1.3486,
+      "step": 11885
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.340734839439392,
+      "learning_rate": 4.619912655216825e-05,
+      "loss": 1.2276,
+      "step": 11890
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.9751700162887573,
+      "learning_rate": 4.619601179889913e-05,
+      "loss": 1.1469,
+      "step": 11895
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.659359335899353,
+      "learning_rate": 4.619289587499631e-05,
+      "loss": 1.271,
+      "step": 11900
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.9473052024841309,
+      "learning_rate": 4.618977878063188e-05,
+      "loss": 1.2939,
+      "step": 11905
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.0142465829849243,
+      "learning_rate": 4.6186660515978e-05,
+      "loss": 1.2489,
+      "step": 11910
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.311125636100769,
+      "learning_rate": 4.618354108120687e-05,
+      "loss": 1.2565,
+      "step": 11915
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 1.6890449523925781,
+      "learning_rate": 4.618042047649079e-05,
+      "loss": 1.2294,
+      "step": 11920
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 2.259535789489746,
+      "learning_rate": 4.6177298702002106e-05,
+      "loss": 1.3299,
+      "step": 11925
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 1.944066047668457,
+      "learning_rate": 4.6174800440290745e-05,
+      "loss": 1.4415,
+      "step": 11930
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 1.1252658367156982,
+      "learning_rate": 4.617167656064589e-05,
+      "loss": 1.271,
+      "step": 11935
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 1.6652323007583618,
+      "learning_rate": 4.616855151171134e-05,
+      "loss": 1.2467,
+      "step": 11940
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 1.1521717309951782,
+      "learning_rate": 4.61654252936597e-05,
+      "loss": 1.3624,
+      "step": 11945
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 1.9085320234298706,
+      "learning_rate": 4.616229790666362e-05,
+      "loss": 1.2027,
+      "step": 11950
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 1.8030284643173218,
+      "learning_rate": 4.6159169350895825e-05,
+      "loss": 1.2746,
+      "step": 11955
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 1.5365478992462158,
+      "learning_rate": 4.61560396265291e-05,
+      "loss": 1.4625,
+      "step": 11960
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 2.0654516220092773,
+      "learning_rate": 4.615290873373629e-05,
+      "loss": 1.2609,
+      "step": 11965
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 2.4061896800994873,
+      "learning_rate": 4.614977667269033e-05,
+      "loss": 1.2647,
+      "step": 11970
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 3.210092544555664,
+      "learning_rate": 4.614664344356417e-05,
+      "loss": 1.2957,
+      "step": 11975
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 2.317610025405884,
+      "learning_rate": 4.614350904653089e-05,
+      "loss": 1.2691,
+      "step": 11980
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.450660228729248,
+      "learning_rate": 4.614037348176358e-05,
+      "loss": 1.2757,
+      "step": 11985
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.0978591442108154,
+      "learning_rate": 4.6137236749435413e-05,
+      "loss": 1.2778,
+      "step": 11990
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.411629319190979,
+      "learning_rate": 4.613409884971963e-05,
+      "loss": 1.5244,
+      "step": 11995
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 4.375972270965576,
+      "learning_rate": 4.613095978278954e-05,
+      "loss": 1.3383,
+      "step": 12000
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.4129486083984375,
+      "learning_rate": 4.6127819548818506e-05,
+      "loss": 1.4209,
+      "step": 12005
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.0286214351654053,
+      "learning_rate": 4.612467814797996e-05,
+      "loss": 1.3274,
+      "step": 12010
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.4175000190734863,
+      "learning_rate": 4.61215355804474e-05,
+      "loss": 1.261,
+      "step": 12015
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.794857144355774,
+      "learning_rate": 4.611839184639437e-05,
+      "loss": 1.2601,
+      "step": 12020
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.3006887435913086,
+      "learning_rate": 4.611524694599452e-05,
+      "loss": 1.1114,
+      "step": 12025
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.3993430137634277,
+      "learning_rate": 4.6112100879421524e-05,
+      "loss": 1.3558,
+      "step": 12030
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.9399762153625488,
+      "learning_rate": 4.610895364684915e-05,
+      "loss": 1.3497,
+      "step": 12035
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.2342357635498047,
+      "learning_rate": 4.61058052484512e-05,
+      "loss": 1.3263,
+      "step": 12040
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.044203042984009,
+      "learning_rate": 4.610265568440157e-05,
+      "loss": 1.3344,
+      "step": 12045
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.6386117935180664,
+      "learning_rate": 4.609950495487419e-05,
+      "loss": 1.143,
+      "step": 12050
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 1.735022783279419,
+      "learning_rate": 4.60963530600431e-05,
+      "loss": 1.4065,
+      "step": 12055
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 1.7873737812042236,
+      "learning_rate": 4.6093200000082346e-05,
+      "loss": 1.3379,
+      "step": 12060
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 1.9546738862991333,
+      "learning_rate": 4.609004577516609e-05,
+      "loss": 1.1978,
+      "step": 12065
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 1.1246274709701538,
+      "learning_rate": 4.6086890385468526e-05,
+      "loss": 1.2626,
+      "step": 12070
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 1.5824791193008423,
+      "learning_rate": 4.6083733831163925e-05,
+      "loss": 1.3883,
+      "step": 12075
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 1.4226622581481934,
+      "learning_rate": 4.608057611242662e-05,
+      "loss": 1.2575,
+      "step": 12080
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 1.6612082719802856,
+      "learning_rate": 4.6077417229431e-05,
+      "loss": 1.107,
+      "step": 12085
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 1.722304105758667,
+      "learning_rate": 4.6074257182351546e-05,
+      "loss": 1.1742,
+      "step": 12090
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 2.5315911769866943,
+      "learning_rate": 4.607109597136277e-05,
+      "loss": 1.2671,
+      "step": 12095
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 1.4968048334121704,
+      "learning_rate": 4.606793359663926e-05,
+      "loss": 1.2445,
+      "step": 12100
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.9814133644104004,
+      "learning_rate": 4.6064770058355675e-05,
+      "loss": 1.2989,
+      "step": 12105
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 1.6419323682785034,
+      "learning_rate": 4.6061605356686746e-05,
+      "loss": 1.4119,
+      "step": 12110
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 1.662709355354309,
+      "learning_rate": 4.605843949180724e-05,
+      "loss": 1.4737,
+      "step": 12115
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.9931641817092896,
+      "learning_rate": 4.605527246389201e-05,
+      "loss": 1.3292,
+      "step": 12120
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 2.7940378189086914,
+      "learning_rate": 4.605210427311596e-05,
+      "loss": 1.3846,
+      "step": 12125
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.8893239498138428,
+      "learning_rate": 4.604893491965409e-05,
+      "loss": 1.2893,
+      "step": 12130
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 2.463541030883789,
+      "learning_rate": 4.6045764403681415e-05,
+      "loss": 1.2914,
+      "step": 12135
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 2.0132181644439697,
+      "learning_rate": 4.604259272537304e-05,
+      "loss": 1.4236,
+      "step": 12140
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.1866495609283447,
+      "learning_rate": 4.603941988490415e-05,
+      "loss": 1.3439,
+      "step": 12145
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 1.9764355421066284,
+      "learning_rate": 4.603624588244997e-05,
+      "loss": 1.3349,
+      "step": 12150
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 1.9496948719024658,
+      "learning_rate": 4.603307071818579e-05,
+      "loss": 1.2836,
+      "step": 12155
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 3.0015974044799805,
+      "learning_rate": 4.602989439228698e-05,
+      "loss": 1.367,
+      "step": 12160
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 1.3181309700012207,
+      "learning_rate": 4.6026716904928965e-05,
+      "loss": 1.4144,
+      "step": 12165
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 1.3011974096298218,
+      "learning_rate": 4.602353825628722e-05,
+      "loss": 1.2887,
+      "step": 12170
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 2.6620357036590576,
+      "learning_rate": 4.602035844653733e-05,
+      "loss": 1.3088,
+      "step": 12175
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 1.3964091539382935,
+      "learning_rate": 4.601717747585488e-05,
+      "loss": 1.2074,
+      "step": 12180
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 2.161696434020996,
+      "learning_rate": 4.601399534441556e-05,
+      "loss": 1.3595,
+      "step": 12185
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.0250306129455566,
+      "learning_rate": 4.601081205239512e-05,
+      "loss": 1.3982,
+      "step": 12190
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.1172339916229248,
+      "learning_rate": 4.6007627599969385e-05,
+      "loss": 1.1501,
+      "step": 12195
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.6155879497528076,
+      "learning_rate": 4.60044419873142e-05,
+      "loss": 1.3539,
+      "step": 12200
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.6924985647201538,
+      "learning_rate": 4.600125521460552e-05,
+      "loss": 1.2307,
+      "step": 12205
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.349064826965332,
+      "learning_rate": 4.599806728201935e-05,
+      "loss": 1.1824,
+      "step": 12210
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 2.545496940612793,
+      "learning_rate": 4.599487818973174e-05,
+      "loss": 1.3419,
+      "step": 12215
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 2.050647258758545,
+      "learning_rate": 4.599168793791884e-05,
+      "loss": 1.21,
+      "step": 12220
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 2.1457366943359375,
+      "learning_rate": 4.598849652675683e-05,
+      "loss": 1.3711,
+      "step": 12225
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 2.319671630859375,
+      "learning_rate": 4.598530395642197e-05,
+      "loss": 1.2425,
+      "step": 12230
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 1.073062777519226,
+      "learning_rate": 4.598211022709059e-05,
+      "loss": 1.3835,
+      "step": 12235
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 1.0914404392242432,
+      "learning_rate": 4.597891533893908e-05,
+      "loss": 1.5263,
+      "step": 12240
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.8850699663162231,
+      "learning_rate": 4.5975719292143865e-05,
+      "loss": 1.299,
+      "step": 12245
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.4798288345336914,
+      "learning_rate": 4.5972522086881485e-05,
+      "loss": 1.3877,
+      "step": 12250
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.8555247783660889,
+      "learning_rate": 4.5969323723328505e-05,
+      "loss": 1.4655,
+      "step": 12255
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.9914946556091309,
+      "learning_rate": 4.596612420166158e-05,
+      "loss": 1.2558,
+      "step": 12260
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.4527971744537354,
+      "learning_rate": 4.596292352205741e-05,
+      "loss": 1.2612,
+      "step": 12265
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.6714884042739868,
+      "learning_rate": 4.595972168469276e-05,
+      "loss": 1.2867,
+      "step": 12270
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.638249397277832,
+      "learning_rate": 4.595651868974447e-05,
+      "loss": 1.3167,
+      "step": 12275
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.3472903966903687,
+      "learning_rate": 4.5953314537389426e-05,
+      "loss": 1.2534,
+      "step": 12280
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 1.781898856163025,
+      "learning_rate": 4.59501092278046e-05,
+      "loss": 1.2795,
+      "step": 12285
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 2.2303428649902344,
+      "learning_rate": 4.594690276116703e-05,
+      "loss": 1.3091,
+      "step": 12290
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 1.383138656616211,
+      "learning_rate": 4.594369513765379e-05,
+      "loss": 1.2663,
+      "step": 12295
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 1.707878828048706,
+      "learning_rate": 4.594048635744203e-05,
+      "loss": 1.2528,
+      "step": 12300
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.8633110523223877,
+      "learning_rate": 4.5937276420708985e-05,
+      "loss": 1.3231,
+      "step": 12305
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 1.4325439929962158,
+      "learning_rate": 4.593406532763192e-05,
+      "loss": 1.21,
+      "step": 12310
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 1.455572247505188,
+      "learning_rate": 4.5930853078388185e-05,
+      "loss": 1.2322,
+      "step": 12315
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.189207911491394,
+      "learning_rate": 4.59276396731552e-05,
+      "loss": 1.3804,
+      "step": 12320
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.2645961046218872,
+      "learning_rate": 4.592442511211042e-05,
+      "loss": 1.3339,
+      "step": 12325
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.2472920417785645,
+      "learning_rate": 4.59212093954314e-05,
+      "loss": 1.2354,
+      "step": 12330
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.2860960960388184,
+      "learning_rate": 4.5917992523295716e-05,
+      "loss": 1.2905,
+      "step": 12335
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.902571439743042,
+      "learning_rate": 4.591477449588106e-05,
+      "loss": 1.2967,
+      "step": 12340
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 2.72182559967041,
+      "learning_rate": 4.591155531336514e-05,
+      "loss": 1.3203,
+      "step": 12345
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 4.221126079559326,
+      "learning_rate": 4.590833497592576e-05,
+      "loss": 1.3342,
+      "step": 12350
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 2.888122320175171,
+      "learning_rate": 4.590511348374078e-05,
+      "loss": 1.2249,
+      "step": 12355
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.8218414783477783,
+      "learning_rate": 4.5901890836988107e-05,
+      "loss": 1.3318,
+      "step": 12360
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.3401216268539429,
+      "learning_rate": 4.589866703584573e-05,
+      "loss": 1.355,
+      "step": 12365
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.530899167060852,
+      "learning_rate": 4.5895442080491694e-05,
+      "loss": 1.1148,
+      "step": 12370
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.1697745323181152,
+      "learning_rate": 4.589221597110411e-05,
+      "loss": 1.3208,
+      "step": 12375
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.9336832761764526,
+      "learning_rate": 4.588898870786116e-05,
+      "loss": 1.3834,
+      "step": 12380
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.2124789953231812,
+      "learning_rate": 4.588576029094106e-05,
+      "loss": 1.2128,
+      "step": 12385
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.4189821481704712,
+      "learning_rate": 4.588253072052214e-05,
+      "loss": 1.3005,
+      "step": 12390
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.8805235624313354,
+      "learning_rate": 4.5879299996782765e-05,
+      "loss": 1.3208,
+      "step": 12395
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.4041588306427,
+      "learning_rate": 4.587606811990134e-05,
+      "loss": 1.3952,
+      "step": 12400
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.0374820232391357,
+      "learning_rate": 4.5872835090056375e-05,
+      "loss": 1.3175,
+      "step": 12405
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.124577760696411,
+      "learning_rate": 4.586960090742643e-05,
+      "loss": 1.2484,
+      "step": 12410
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 4.781686305999756,
+      "learning_rate": 4.586636557219011e-05,
+      "loss": 1.2985,
+      "step": 12415
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 1.7843397855758667,
+      "learning_rate": 4.586312908452612e-05,
+      "loss": 1.2425,
+      "step": 12420
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 3.2184629440307617,
+      "learning_rate": 4.585989144461319e-05,
+      "loss": 1.3105,
+      "step": 12425
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 1.4963960647583008,
+      "learning_rate": 4.585665265263014e-05,
+      "loss": 1.2972,
+      "step": 12430
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 3.5858564376831055,
+      "learning_rate": 4.585341270875584e-05,
+      "loss": 1.3527,
+      "step": 12435
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 2.781733989715576,
+      "learning_rate": 4.5850171613169235e-05,
+      "loss": 1.3733,
+      "step": 12440
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.917702317237854,
+      "learning_rate": 4.5846929366049316e-05,
+      "loss": 1.2823,
+      "step": 12445
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 1.9473000764846802,
+      "learning_rate": 4.584368596757517e-05,
+      "loss": 1.2604,
+      "step": 12450
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 1.2440733909606934,
+      "learning_rate": 4.584044141792591e-05,
+      "loss": 1.3558,
+      "step": 12455
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 1.291435956954956,
+      "learning_rate": 4.5837195717280736e-05,
+      "loss": 1.2116,
+      "step": 12460
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 1.013763427734375,
+      "learning_rate": 4.583394886581889e-05,
+      "loss": 1.3011,
+      "step": 12465
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 1.0628958940505981,
+      "learning_rate": 4.583070086371971e-05,
+      "loss": 1.4285,
+      "step": 12470
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 1.822394847869873,
+      "learning_rate": 4.5827451711162575e-05,
+      "loss": 1.2755,
+      "step": 12475
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 1.7311921119689941,
+      "learning_rate": 4.5824201408326934e-05,
+      "loss": 1.4643,
+      "step": 12480
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 1.9452160596847534,
+      "learning_rate": 4.582094995539229e-05,
+      "loss": 1.2745,
+      "step": 12485
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 4.922383785247803,
+      "learning_rate": 4.581769735253822e-05,
+      "loss": 1.3349,
+      "step": 12490
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 2.8117499351501465,
+      "learning_rate": 4.581444359994437e-05,
+      "loss": 1.4012,
+      "step": 12495
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 1.1426706314086914,
+      "learning_rate": 4.581118869779043e-05,
+      "loss": 1.2974,
+      "step": 12500
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 1.857204556465149,
+      "learning_rate": 4.580793264625618e-05,
+      "loss": 1.2929,
+      "step": 12505
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 1.1504136323928833,
+      "learning_rate": 4.580467544552143e-05,
+      "loss": 1.3019,
+      "step": 12510
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 3.0354089736938477,
+      "learning_rate": 4.580141709576608e-05,
+      "loss": 1.3482,
+      "step": 12515
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 1.1074090003967285,
+      "learning_rate": 4.579815759717009e-05,
+      "loss": 1.4458,
+      "step": 12520
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 1.3502519130706787,
+      "learning_rate": 4.579489694991347e-05,
+      "loss": 1.1843,
+      "step": 12525
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.9303249716758728,
+      "learning_rate": 4.579163515417631e-05,
+      "loss": 1.3647,
+      "step": 12530
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.86869877576828,
+      "learning_rate": 4.578837221013875e-05,
+      "loss": 1.2498,
+      "step": 12535
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 1.8906731605529785,
+      "learning_rate": 4.5785108117981e-05,
+      "loss": 1.3277,
+      "step": 12540
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 1.9639962911605835,
+      "learning_rate": 4.578184287788334e-05,
+      "loss": 1.4021,
+      "step": 12545
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 2.0274109840393066,
+      "learning_rate": 4.5778576490026094e-05,
+      "loss": 1.2441,
+      "step": 12550
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 2.2819442749023438,
+      "learning_rate": 4.577530895458967e-05,
+      "loss": 1.4074,
+      "step": 12555
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 1.8426662683486938,
+      "learning_rate": 4.577204027175453e-05,
+      "loss": 1.2962,
+      "step": 12560
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 1.2728874683380127,
+      "learning_rate": 4.576877044170118e-05,
+      "loss": 1.2356,
+      "step": 12565
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 1.636061668395996,
+      "learning_rate": 4.576549946461024e-05,
+      "loss": 1.4087,
+      "step": 12570
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 1.0893962383270264,
+      "learning_rate": 4.576222734066235e-05,
+      "loss": 1.1445,
+      "step": 12575
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 1.898051381111145,
+      "learning_rate": 4.575895407003822e-05,
+      "loss": 1.3225,
+      "step": 12580
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.7080368995666504,
+      "learning_rate": 4.575567965291864e-05,
+      "loss": 1.3176,
+      "step": 12585
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 2.100494861602783,
+      "learning_rate": 4.575240408948443e-05,
+      "loss": 1.4143,
+      "step": 12590
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.711904764175415,
+      "learning_rate": 4.5749127379916536e-05,
+      "loss": 1.4185,
+      "step": 12595
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 3.1518118381500244,
+      "learning_rate": 4.57458495243959e-05,
+      "loss": 1.3182,
+      "step": 12600
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 2.016057252883911,
+      "learning_rate": 4.574257052310355e-05,
+      "loss": 1.3566,
+      "step": 12605
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.0968761444091797,
+      "learning_rate": 4.573929037622059e-05,
+      "loss": 1.2916,
+      "step": 12610
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.1773560047149658,
+      "learning_rate": 4.573600908392819e-05,
+      "loss": 1.2659,
+      "step": 12615
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.2752783298492432,
+      "learning_rate": 4.573272664640755e-05,
+      "loss": 1.1999,
+      "step": 12620
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.5268641710281372,
+      "learning_rate": 4.5729443063839986e-05,
+      "loss": 1.4445,
+      "step": 12625
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.8288689851760864,
+      "learning_rate": 4.572615833640681e-05,
+      "loss": 1.2565,
+      "step": 12630
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 2.030205726623535,
+      "learning_rate": 4.572287246428946e-05,
+      "loss": 1.2579,
+      "step": 12635
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 2.9597010612487793,
+      "learning_rate": 4.57195854476694e-05,
+      "loss": 1.1755,
+      "step": 12640
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.5973179340362549,
+      "learning_rate": 4.571629728672818e-05,
+      "loss": 1.1414,
+      "step": 12645
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.202883005142212,
+      "learning_rate": 4.5713007981647394e-05,
+      "loss": 1.3216,
+      "step": 12650
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.9232937097549438,
+      "learning_rate": 4.57097175326087e-05,
+      "loss": 1.306,
+      "step": 12655
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.9437748193740845,
+      "learning_rate": 4.570642593979384e-05,
+      "loss": 1.2824,
+      "step": 12660
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.9514751434326172,
+      "learning_rate": 4.5703133203384594e-05,
+      "loss": 1.404,
+      "step": 12665
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.337649941444397,
+      "learning_rate": 4.5699839323562824e-05,
+      "loss": 1.4712,
+      "step": 12670
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.9567248225212097,
+      "learning_rate": 4.569654430051045e-05,
+      "loss": 1.2779,
+      "step": 12675
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.1744178533554077,
+      "learning_rate": 4.5693248134409434e-05,
+      "loss": 1.3132,
+      "step": 12680
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 2.404231548309326,
+      "learning_rate": 4.568995082544184e-05,
+      "loss": 1.3972,
+      "step": 12685
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.4296953678131104,
+      "learning_rate": 4.5686652373789764e-05,
+      "loss": 1.3201,
+      "step": 12690
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.6986697912216187,
+      "learning_rate": 4.568335277963538e-05,
+      "loss": 1.2344,
+      "step": 12695
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.2852948904037476,
+      "learning_rate": 4.568005204316093e-05,
+      "loss": 1.3389,
+      "step": 12700
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.1272681951522827,
+      "learning_rate": 4.567675016454869e-05,
+      "loss": 1.3613,
+      "step": 12705
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 3.3429181575775146,
+      "learning_rate": 4.567344714398104e-05,
+      "loss": 1.3005,
+      "step": 12710
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.5739295482635498,
+      "learning_rate": 4.5670142981640384e-05,
+      "loss": 1.0753,
+      "step": 12715
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.3423209190368652,
+      "learning_rate": 4.566683767770923e-05,
+      "loss": 1.1526,
+      "step": 12720
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.0714255571365356,
+      "learning_rate": 4.5663531232370105e-05,
+      "loss": 1.4233,
+      "step": 12725
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.9389391541481018,
+      "learning_rate": 4.5660223645805624e-05,
+      "loss": 1.3871,
+      "step": 12730
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.9636523723602295,
+      "learning_rate": 4.5656914918198465e-05,
+      "loss": 1.2896,
+      "step": 12735
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 2.509476661682129,
+      "learning_rate": 4.565360504973138e-05,
+      "loss": 1.2616,
+      "step": 12740
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.8651650547981262,
+      "learning_rate": 4.565029404058715e-05,
+      "loss": 1.4291,
+      "step": 12745
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.635857105255127,
+      "learning_rate": 4.564698189094864e-05,
+      "loss": 1.2209,
+      "step": 12750
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 2.29955792427063,
+      "learning_rate": 4.564366860099879e-05,
+      "loss": 1.3041,
+      "step": 12755
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 1.347915768623352,
+      "learning_rate": 4.5640354170920575e-05,
+      "loss": 1.2319,
+      "step": 12760
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 3.2308144569396973,
+      "learning_rate": 4.563703860089705e-05,
+      "loss": 1.3922,
+      "step": 12765
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 1.212275505065918,
+      "learning_rate": 4.5633721891111336e-05,
+      "loss": 1.4356,
+      "step": 12770
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 1.5803014039993286,
+      "learning_rate": 4.563040404174662e-05,
+      "loss": 1.2631,
+      "step": 12775
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 2.851174831390381,
+      "learning_rate": 4.562708505298612e-05,
+      "loss": 1.3305,
+      "step": 12780
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.705056071281433,
+      "learning_rate": 4.5623764925013155e-05,
+      "loss": 1.1758,
+      "step": 12785
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.0315181016921997,
+      "learning_rate": 4.56204436580111e-05,
+      "loss": 1.1893,
+      "step": 12790
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.8870512247085571,
+      "learning_rate": 4.561712125216337e-05,
+      "loss": 1.4365,
+      "step": 12795
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 2.1169257164001465,
+      "learning_rate": 4.561379770765346e-05,
+      "loss": 1.3076,
+      "step": 12800
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 2.4839279651641846,
+      "learning_rate": 4.5610473024664935e-05,
+      "loss": 1.2516,
+      "step": 12805
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.1137679815292358,
+      "learning_rate": 4.560714720338141e-05,
+      "loss": 1.3052,
+      "step": 12810
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.6442543268203735,
+      "learning_rate": 4.560382024398655e-05,
+      "loss": 1.3419,
+      "step": 12815
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.8083311319351196,
+      "learning_rate": 4.560049214666413e-05,
+      "loss": 1.304,
+      "step": 12820
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.2691807746887207,
+      "learning_rate": 4.559716291159793e-05,
+      "loss": 1.2481,
+      "step": 12825
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.8940715789794922,
+      "learning_rate": 4.5593832538971846e-05,
+      "loss": 1.2207,
+      "step": 12830
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.9223408699035645,
+      "learning_rate": 4.5590501028969787e-05,
+      "loss": 1.313,
+      "step": 12835
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.92839777469635,
+      "learning_rate": 4.5587168381775756e-05,
+      "loss": 1.2413,
+      "step": 12840
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.7354521751403809,
+      "learning_rate": 4.5583834597573825e-05,
+      "loss": 1.2551,
+      "step": 12845
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.491490125656128,
+      "learning_rate": 4.55804996765481e-05,
+      "loss": 1.3245,
+      "step": 12850
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.8341639041900635,
+      "learning_rate": 4.5577163618882766e-05,
+      "loss": 1.3094,
+      "step": 12855
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.9003182053565979,
+      "learning_rate": 4.557382642476208e-05,
+      "loss": 1.3996,
+      "step": 12860
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 2.012786388397217,
+      "learning_rate": 4.557048809437035e-05,
+      "loss": 1.4447,
+      "step": 12865
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 1.0975584983825684,
+      "learning_rate": 4.556714862789193e-05,
+      "loss": 1.0958,
+      "step": 12870
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 1.1538339853286743,
+      "learning_rate": 4.556380802551128e-05,
+      "loss": 1.268,
+      "step": 12875
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 2.0543277263641357,
+      "learning_rate": 4.556046628741288e-05,
+      "loss": 1.4274,
+      "step": 12880
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.767110824584961,
+      "learning_rate": 4.555712341378131e-05,
+      "loss": 1.3229,
+      "step": 12885
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.8329064846038818,
+      "learning_rate": 4.555377940480118e-05,
+      "loss": 1.2361,
+      "step": 12890
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.9706895351409912,
+      "learning_rate": 4.5550434260657174e-05,
+      "loss": 1.0816,
+      "step": 12895
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.0541021823883057,
+      "learning_rate": 4.554708798153404e-05,
+      "loss": 1.1388,
+      "step": 12900
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.4227367639541626,
+      "learning_rate": 4.55437405676166e-05,
+      "loss": 1.3312,
+      "step": 12905
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.4533671140670776,
+      "learning_rate": 4.554039201908972e-05,
+      "loss": 1.2984,
+      "step": 12910
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.0998986959457397,
+      "learning_rate": 4.553704233613833e-05,
+      "loss": 1.2591,
+      "step": 12915
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.8056262731552124,
+      "learning_rate": 4.553369151894746e-05,
+      "loss": 1.113,
+      "step": 12920
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.9125102162361145,
+      "learning_rate": 4.553033956770214e-05,
+      "loss": 1.3231,
+      "step": 12925
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.3653115034103394,
+      "learning_rate": 4.55269864825875e-05,
+      "loss": 1.4139,
+      "step": 12930
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.144234299659729,
+      "learning_rate": 4.552363226378874e-05,
+      "loss": 1.2501,
+      "step": 12935
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 2.3261189460754395,
+      "learning_rate": 4.5520276911491104e-05,
+      "loss": 1.2052,
+      "step": 12940
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.1465058326721191,
+      "learning_rate": 4.5516920425879905e-05,
+      "loss": 1.3006,
+      "step": 12945
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.5299936532974243,
+      "learning_rate": 4.5513562807140515e-05,
+      "loss": 1.3346,
+      "step": 12950
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 1.8486464023590088,
+      "learning_rate": 4.551020405545837e-05,
+      "loss": 1.3022,
+      "step": 12955
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 1.537140130996704,
+      "learning_rate": 4.550684417101898e-05,
+      "loss": 1.3352,
+      "step": 12960
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 1.5041508674621582,
+      "learning_rate": 4.55034831540079e-05,
+      "loss": 1.3945,
+      "step": 12965
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 2.660659074783325,
+      "learning_rate": 4.550012100461075e-05,
+      "loss": 1.3366,
+      "step": 12970
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 1.6128523349761963,
+      "learning_rate": 4.549675772301323e-05,
+      "loss": 1.3376,
+      "step": 12975
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 1.45650315284729,
+      "learning_rate": 4.549339330940109e-05,
+      "loss": 1.5226,
+      "step": 12980
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 1.1668511629104614,
+      "learning_rate": 4.5490027763960144e-05,
+      "loss": 1.2782,
+      "step": 12985
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 1.2220723628997803,
+      "learning_rate": 4.548666108687625e-05,
+      "loss": 1.3738,
+      "step": 12990
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 2.1042776107788086,
+      "learning_rate": 4.548329327833537e-05,
+      "loss": 1.3442,
+      "step": 12995
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 1.7469139099121094,
+      "learning_rate": 4.547992433852349e-05,
+      "loss": 1.2097,
+      "step": 13000
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.8937451839447021,
+      "learning_rate": 4.547655426762668e-05,
+      "loss": 1.3553,
+      "step": 13005
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 2.8568267822265625,
+      "learning_rate": 4.5473183065831045e-05,
+      "loss": 1.4715,
+      "step": 13010
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 1.3825722932815552,
+      "learning_rate": 4.546981073332281e-05,
+      "loss": 1.2806,
+      "step": 13015
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 1.2610100507736206,
+      "learning_rate": 4.54664372702882e-05,
+      "loss": 1.1896,
+      "step": 13020
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 1.5070395469665527,
+      "learning_rate": 4.5463062676913527e-05,
+      "loss": 1.3712,
+      "step": 13025
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.2432565689086914,
+      "learning_rate": 4.545968695338518e-05,
+      "loss": 1.3497,
+      "step": 13030
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 1.4944096803665161,
+      "learning_rate": 4.545631009988958e-05,
+      "loss": 1.2663,
+      "step": 13035
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.186475992202759,
+      "learning_rate": 4.545293211661324e-05,
+      "loss": 1.3296,
+      "step": 13040
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 1.5511248111724854,
+      "learning_rate": 4.544955300374273e-05,
+      "loss": 1.1667,
+      "step": 13045
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.8364615440368652,
+      "learning_rate": 4.544617276146465e-05,
+      "loss": 1.2472,
+      "step": 13050
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.9596519470214844,
+      "learning_rate": 4.544279138996571e-05,
+      "loss": 1.3599,
+      "step": 13055
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.2000133991241455,
+      "learning_rate": 4.543940888943264e-05,
+      "loss": 1.229,
+      "step": 13060
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.129855751991272,
+      "learning_rate": 4.543602526005227e-05,
+      "loss": 1.2954,
+      "step": 13065
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.3284913301467896,
+      "learning_rate": 4.543264050201146e-05,
+      "loss": 1.3602,
+      "step": 13070
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.7732700109481812,
+      "learning_rate": 4.5429254615497165e-05,
+      "loss": 1.3526,
+      "step": 13075
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.5674593448638916,
+      "learning_rate": 4.542586760069637e-05,
+      "loss": 1.4001,
+      "step": 13080
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.350285291671753,
+      "learning_rate": 4.542247945779613e-05,
+      "loss": 1.3551,
+      "step": 13085
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.2830190658569336,
+      "learning_rate": 4.5419090186983587e-05,
+      "loss": 1.3441,
+      "step": 13090
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.5413031578063965,
+      "learning_rate": 4.541569978844591e-05,
+      "loss": 1.2436,
+      "step": 13095
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.2088831663131714,
+      "learning_rate": 4.541230826237036e-05,
+      "loss": 1.2817,
+      "step": 13100
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.644108533859253,
+      "learning_rate": 4.540891560894424e-05,
+      "loss": 1.3467,
+      "step": 13105
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.8455935716629028,
+      "learning_rate": 4.5405521828354924e-05,
+      "loss": 1.2956,
+      "step": 13110
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.0476539134979248,
+      "learning_rate": 4.5402126920789854e-05,
+      "loss": 1.2251,
+      "step": 13115
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.2740333080291748,
+      "learning_rate": 4.539873088643651e-05,
+      "loss": 1.4182,
+      "step": 13120
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.2473523616790771,
+      "learning_rate": 4.539533372548247e-05,
+      "loss": 1.4605,
+      "step": 13125
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.2522354125976562,
+      "learning_rate": 4.5391935438115355e-05,
+      "loss": 1.4085,
+      "step": 13130
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.8400856852531433,
+      "learning_rate": 4.538853602452283e-05,
+      "loss": 1.4076,
+      "step": 13135
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 3.1586833000183105,
+      "learning_rate": 4.538513548489265e-05,
+      "loss": 1.2214,
+      "step": 13140
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.3957751989364624,
+      "learning_rate": 4.538173381941264e-05,
+      "loss": 1.3048,
+      "step": 13145
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.6071856021881104,
+      "learning_rate": 4.537833102827065e-05,
+      "loss": 1.2303,
+      "step": 13150
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 1.384070873260498,
+      "learning_rate": 4.537492711165462e-05,
+      "loss": 1.3229,
+      "step": 13155
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 2.3858485221862793,
+      "learning_rate": 4.5371522069752544e-05,
+      "loss": 1.3097,
+      "step": 13160
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 1.255236029624939,
+      "learning_rate": 4.536811590275247e-05,
+      "loss": 1.3769,
+      "step": 13165
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 1.0633465051651,
+      "learning_rate": 4.5364708610842545e-05,
+      "loss": 1.2269,
+      "step": 13170
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 1.6456542015075684,
+      "learning_rate": 4.536130019421092e-05,
+      "loss": 1.4208,
+      "step": 13175
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 1.0649845600128174,
+      "learning_rate": 4.535789065304585e-05,
+      "loss": 1.2724,
+      "step": 13180
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 2.226824998855591,
+      "learning_rate": 4.535447998753564e-05,
+      "loss": 1.2443,
+      "step": 13185
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 1.2029057741165161,
+      "learning_rate": 4.535106819786866e-05,
+      "loss": 1.3256,
+      "step": 13190
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.9885445237159729,
+      "learning_rate": 4.534765528423333e-05,
+      "loss": 1.3866,
+      "step": 13195
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 4.255756855010986,
+      "learning_rate": 4.5344241246818154e-05,
+      "loss": 1.1524,
+      "step": 13200
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 1.9906721115112305,
+      "learning_rate": 4.5340826085811684e-05,
+      "loss": 1.3169,
+      "step": 13205
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 1.9341821670532227,
+      "learning_rate": 4.533740980140253e-05,
+      "loss": 1.4293,
+      "step": 13210
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 2.2988078594207764,
+      "learning_rate": 4.533399239377937e-05,
+      "loss": 1.2559,
+      "step": 13215
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 1.1931335926055908,
+      "learning_rate": 4.5330573863130946e-05,
+      "loss": 1.2239,
+      "step": 13220
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 1.9085408449172974,
+      "learning_rate": 4.5327154209646065e-05,
+      "loss": 1.3006,
+      "step": 13225
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 1.42606782913208,
+      "learning_rate": 4.532373343351358e-05,
+      "loss": 1.1926,
+      "step": 13230
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 1.9960684776306152,
+      "learning_rate": 4.5320311534922425e-05,
+      "loss": 1.4581,
+      "step": 13235
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 1.3302810192108154,
+      "learning_rate": 4.531688851406159e-05,
+      "loss": 1.2497,
+      "step": 13240
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 1.0140787363052368,
+      "learning_rate": 4.531346437112012e-05,
+      "loss": 1.2587,
+      "step": 13245
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 2.0420432090759277,
+      "learning_rate": 4.5310039106287115e-05,
+      "loss": 1.2299,
+      "step": 13250
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.9707965850830078,
+      "learning_rate": 4.530661271975177e-05,
+      "loss": 1.2625,
+      "step": 13255
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.9889094829559326,
+      "learning_rate": 4.530318521170332e-05,
+      "loss": 1.445,
+      "step": 13260
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 1.5554451942443848,
+      "learning_rate": 4.529975658233104e-05,
+      "loss": 1.1236,
+      "step": 13265
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 1.6578177213668823,
+      "learning_rate": 4.529632683182432e-05,
+      "loss": 1.1677,
+      "step": 13270
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 2.4614758491516113,
+      "learning_rate": 4.529289596037256e-05,
+      "loss": 1.2574,
+      "step": 13275
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 1.8018752336502075,
+      "learning_rate": 4.528946396816524e-05,
+      "loss": 1.3027,
+      "step": 13280
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 2.3157856464385986,
+      "learning_rate": 4.5286030855391924e-05,
+      "loss": 1.439,
+      "step": 13285
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 1.605214238166809,
+      "learning_rate": 4.528259662224221e-05,
+      "loss": 1.2793,
+      "step": 13290
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 3.1263720989227295,
+      "learning_rate": 4.527916126890576e-05,
+      "loss": 1.386,
+      "step": 13295
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.8434714078903198,
+      "learning_rate": 4.527572479557232e-05,
+      "loss": 1.3058,
+      "step": 13300
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 1.2560151815414429,
+      "learning_rate": 4.5272287202431674e-05,
+      "loss": 1.2894,
+      "step": 13305
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 2.4032607078552246,
+      "learning_rate": 4.526884848967368e-05,
+      "loss": 1.2046,
+      "step": 13310
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.8934091329574585,
+      "learning_rate": 4.526540865748824e-05,
+      "loss": 1.3464,
+      "step": 13315
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 1.3520596027374268,
+      "learning_rate": 4.526196770606536e-05,
+      "loss": 1.3311,
+      "step": 13320
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 1.0915952920913696,
+      "learning_rate": 4.5258525635595054e-05,
+      "loss": 1.3354,
+      "step": 13325
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 1.1796505451202393,
+      "learning_rate": 4.525508244626743e-05,
+      "loss": 1.3232,
+      "step": 13330
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 1.1852450370788574,
+      "learning_rate": 4.525163813827267e-05,
+      "loss": 1.1876,
+      "step": 13335
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 3.0009610652923584,
+      "learning_rate": 4.524819271180098e-05,
+      "loss": 1.1891,
+      "step": 13340
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.9333651065826416,
+      "learning_rate": 4.524474616704265e-05,
+      "loss": 1.3117,
+      "step": 13345
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 1.0289852619171143,
+      "learning_rate": 4.524129850418803e-05,
+      "loss": 1.3284,
+      "step": 13350
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.6752545833587646,
+      "learning_rate": 4.523784972342755e-05,
+      "loss": 1.4082,
+      "step": 13355
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.345482349395752,
+      "learning_rate": 4.523439982495166e-05,
+      "loss": 1.2208,
+      "step": 13360
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.1788907051086426,
+      "learning_rate": 4.5230948808950894e-05,
+      "loss": 1.2674,
+      "step": 13365
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.9993130564689636,
+      "learning_rate": 4.522749667561586e-05,
+      "loss": 1.1215,
+      "step": 13370
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.5533525943756104,
+      "learning_rate": 4.52240434251372e-05,
+      "loss": 1.409,
+      "step": 13375
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.6221991181373596,
+      "learning_rate": 4.522058905770564e-05,
+      "loss": 1.2527,
+      "step": 13380
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.9359819889068604,
+      "learning_rate": 4.521713357351198e-05,
+      "loss": 1.1879,
+      "step": 13385
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 1.0392394065856934,
+      "learning_rate": 4.521367697274704e-05,
+      "loss": 1.1468,
+      "step": 13390
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 1.0424681901931763,
+      "learning_rate": 4.5210219255601734e-05,
+      "loss": 1.2535,
+      "step": 13395
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 1.4294004440307617,
+      "learning_rate": 4.5206760422267025e-05,
+      "loss": 1.3226,
+      "step": 13400
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 1.275223731994629,
+      "learning_rate": 4.520330047293394e-05,
+      "loss": 1.3615,
+      "step": 13405
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 1.919305443763733,
+      "learning_rate": 4.519983940779357e-05,
+      "loss": 1.1126,
+      "step": 13410
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.8619056344032288,
+      "learning_rate": 4.519637722703707e-05,
+      "loss": 1.2955,
+      "step": 13415
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 3.1386544704437256,
+      "learning_rate": 4.519291393085564e-05,
+      "loss": 1.0872,
+      "step": 13420
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 1.826619029045105,
+      "learning_rate": 4.5189449519440575e-05,
+      "loss": 1.3617,
+      "step": 13425
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.8781746029853821,
+      "learning_rate": 4.5185983992983186e-05,
+      "loss": 1.1874,
+      "step": 13430
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 1.6987659931182861,
+      "learning_rate": 4.518251735167489e-05,
+      "loss": 1.5476,
+      "step": 13435
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 1.178309679031372,
+      "learning_rate": 4.517904959570714e-05,
+      "loss": 1.1798,
+      "step": 13440
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 1.655390977859497,
+      "learning_rate": 4.517558072527146e-05,
+      "loss": 1.3168,
+      "step": 13445
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 2.087698459625244,
+      "learning_rate": 4.517211074055942e-05,
+      "loss": 1.1618,
+      "step": 13450
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 1.096436858177185,
+      "learning_rate": 4.516863964176268e-05,
+      "loss": 1.1491,
+      "step": 13455
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 1.1264992952346802,
+      "learning_rate": 4.5165167429072924e-05,
+      "loss": 1.2237,
+      "step": 13460
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 4.63450813293457,
+      "learning_rate": 4.516169410268194e-05,
+      "loss": 1.2954,
+      "step": 13465
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 1.207798719406128,
+      "learning_rate": 4.5158219662781546e-05,
+      "loss": 1.3565,
+      "step": 13470
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 1.3179618120193481,
+      "learning_rate": 4.515474410956363e-05,
+      "loss": 1.358,
+      "step": 13475
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 2.3414742946624756,
+      "learning_rate": 4.5151267443220146e-05,
+      "loss": 1.2395,
+      "step": 13480
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 1.8621050119400024,
+      "learning_rate": 4.514778966394312e-05,
+      "loss": 1.318,
+      "step": 13485
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.7734013795852661,
+      "learning_rate": 4.5144310771924606e-05,
+      "loss": 1.261,
+      "step": 13490
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.8414225578308105,
+      "learning_rate": 4.514083076735674e-05,
+      "loss": 1.1795,
+      "step": 13495
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.5032846927642822,
+      "learning_rate": 4.5137349650431735e-05,
+      "loss": 1.2926,
+      "step": 13500
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 2.254154920578003,
+      "learning_rate": 4.513386742134183e-05,
+      "loss": 1.227,
+      "step": 13505
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.056298017501831,
+      "learning_rate": 4.5130384080279364e-05,
+      "loss": 1.1288,
+      "step": 13510
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.4214680194854736,
+      "learning_rate": 4.512689962743671e-05,
+      "loss": 1.317,
+      "step": 13515
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 2.0844547748565674,
+      "learning_rate": 4.512341406300631e-05,
+      "loss": 1.152,
+      "step": 13520
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.6370080709457397,
+      "learning_rate": 4.511992738718066e-05,
+      "loss": 1.2748,
+      "step": 13525
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.4384336471557617,
+      "learning_rate": 4.511643960015234e-05,
+      "loss": 1.3395,
+      "step": 13530
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.0424606800079346,
+      "learning_rate": 4.5112950702113975e-05,
+      "loss": 1.2358,
+      "step": 13535
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.1179556846618652,
+      "learning_rate": 4.5109460693258245e-05,
+      "loss": 1.1542,
+      "step": 13540
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.2349655628204346,
+      "learning_rate": 4.51059695737779e-05,
+      "loss": 1.2682,
+      "step": 13545
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 2.0068607330322266,
+      "learning_rate": 4.510247734386576e-05,
+      "loss": 1.2837,
+      "step": 13550
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 3.1415653228759766,
+      "learning_rate": 4.5098984003714686e-05,
+      "loss": 1.2172,
+      "step": 13555
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 2.988642454147339,
+      "learning_rate": 4.5095489553517625e-05,
+      "loss": 1.1815,
+      "step": 13560
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 2.215256690979004,
+      "learning_rate": 4.509199399346756e-05,
+      "loss": 1.2083,
+      "step": 13565
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 2.271375894546509,
+      "learning_rate": 4.508849732375755e-05,
+      "loss": 1.4811,
+      "step": 13570
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 2.4488816261291504,
+      "learning_rate": 4.5084999544580714e-05,
+      "loss": 1.3544,
+      "step": 13575
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 1.7777202129364014,
+      "learning_rate": 4.5081500656130225e-05,
+      "loss": 1.0847,
+      "step": 13580
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 1.3816187381744385,
+      "learning_rate": 4.507800065859934e-05,
+      "loss": 1.2128,
+      "step": 13585
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 1.5490256547927856,
+      "learning_rate": 4.507449955218134e-05,
+      "loss": 1.2547,
+      "step": 13590
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 1.968616247177124,
+      "learning_rate": 4.50709973370696e-05,
+      "loss": 1.2077,
+      "step": 13595
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 1.046992540359497,
+      "learning_rate": 4.506749401345754e-05,
+      "loss": 1.3679,
+      "step": 13600
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 2.8304526805877686,
+      "learning_rate": 4.506398958153864e-05,
+      "loss": 1.2516,
+      "step": 13605
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 2.89141845703125,
+      "learning_rate": 4.5060484041506454e-05,
+      "loss": 1.2331,
+      "step": 13610
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 1.2688446044921875,
+      "learning_rate": 4.505697739355459e-05,
+      "loss": 1.2566,
+      "step": 13615
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 2.1087491512298584,
+      "learning_rate": 4.505346963787671e-05,
+      "loss": 1.0886,
+      "step": 13620
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 2.119562864303589,
+      "learning_rate": 4.5049960774666546e-05,
+      "loss": 1.1783,
+      "step": 13625
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 1.7251992225646973,
+      "learning_rate": 4.504645080411789e-05,
+      "loss": 1.1867,
+      "step": 13630
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 2.285336494445801,
+      "learning_rate": 4.5042939726424585e-05,
+      "loss": 1.223,
+      "step": 13635
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 1.3887274265289307,
+      "learning_rate": 4.503942754178056e-05,
+      "loss": 1.1539,
+      "step": 13640
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 2.003840923309326,
+      "learning_rate": 4.503591425037978e-05,
+      "loss": 1.3083,
+      "step": 13645
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 1.603243112564087,
+      "learning_rate": 4.5032399852416285e-05,
+      "loss": 1.3067,
+      "step": 13650
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.1768264770507812,
+      "learning_rate": 4.5028884348084166e-05,
+      "loss": 1.4105,
+      "step": 13655
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.2902339696884155,
+      "learning_rate": 4.502536773757758e-05,
+      "loss": 1.306,
+      "step": 13660
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.6430679559707642,
+      "learning_rate": 4.502185002109075e-05,
+      "loss": 1.3452,
+      "step": 13665
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.537097454071045,
+      "learning_rate": 4.5018331198817966e-05,
+      "loss": 1.2132,
+      "step": 13670
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.9888067245483398,
+      "learning_rate": 4.5014811270953546e-05,
+      "loss": 1.1079,
+      "step": 13675
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 2.0669586658477783,
+      "learning_rate": 4.5011290237691905e-05,
+      "loss": 1.3401,
+      "step": 13680
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.0341452360153198,
+      "learning_rate": 4.500776809922751e-05,
+      "loss": 1.2518,
+      "step": 13685
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 1.6001540422439575,
+      "learning_rate": 4.500424485575487e-05,
+      "loss": 1.2202,
+      "step": 13690
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 1.255566120147705,
+      "learning_rate": 4.500072050746859e-05,
+      "loss": 1.187,
+      "step": 13695
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 1.929617166519165,
+      "learning_rate": 4.4997195054563304e-05,
+      "loss": 1.1811,
+      "step": 13700
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 5.457527160644531,
+      "learning_rate": 4.499366849723372e-05,
+      "loss": 1.296,
+      "step": 13705
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 1.2115625143051147,
+      "learning_rate": 4.4990140835674606e-05,
+      "loss": 1.3369,
+      "step": 13710
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.7389108538627625,
+      "learning_rate": 4.498661207008079e-05,
+      "loss": 1.2963,
+      "step": 13715
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 1.6749463081359863,
+      "learning_rate": 4.498308220064717e-05,
+      "loss": 1.1794,
+      "step": 13720
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 2.0370562076568604,
+      "learning_rate": 4.497955122756868e-05,
+      "loss": 1.4314,
+      "step": 13725
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 3.5120744705200195,
+      "learning_rate": 4.4976019151040364e-05,
+      "loss": 1.1369,
+      "step": 13730
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 2.3618757724761963,
+      "learning_rate": 4.497248597125726e-05,
+      "loss": 1.3716,
+      "step": 13735
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 1.495405673980713,
+      "learning_rate": 4.496895168841452e-05,
+      "loss": 1.3534,
+      "step": 13740
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 3.024486541748047,
+      "learning_rate": 4.496541630270733e-05,
+      "loss": 1.2207,
+      "step": 13745
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 2.2674665451049805,
+      "learning_rate": 4.496187981433095e-05,
+      "loss": 1.4337,
+      "step": 13750
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 2.4405548572540283,
+      "learning_rate": 4.49583422234807e-05,
+      "loss": 1.318,
+      "step": 13755
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.9450554847717285,
+      "learning_rate": 4.495480353035196e-05,
+      "loss": 1.3367,
+      "step": 13760
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 2.9744069576263428,
+      "learning_rate": 4.4951263735140156e-05,
+      "loss": 1.2266,
+      "step": 13765
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 2.1863605976104736,
+      "learning_rate": 4.4947722838040795e-05,
+      "loss": 1.2261,
+      "step": 13770
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 1.7596246004104614,
+      "learning_rate": 4.494418083924944e-05,
+      "loss": 1.187,
+      "step": 13775
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 1.8887165784835815,
+      "learning_rate": 4.4940637738961697e-05,
+      "loss": 1.2505,
+      "step": 13780
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 1.4990458488464355,
+      "learning_rate": 4.493709353737327e-05,
+      "loss": 1.2451,
+      "step": 13785
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 1.5014214515686035,
+      "learning_rate": 4.493354823467989e-05,
+      "loss": 1.2606,
+      "step": 13790
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 2.0150039196014404,
+      "learning_rate": 4.4930001831077355e-05,
+      "loss": 1.4081,
+      "step": 13795
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 2.609929323196411,
+      "learning_rate": 4.492645432676154e-05,
+      "loss": 1.2381,
+      "step": 13800
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 1.365830659866333,
+      "learning_rate": 4.4922905721928366e-05,
+      "loss": 1.2332,
+      "step": 13805
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 1.4366599321365356,
+      "learning_rate": 4.491935601677381e-05,
+      "loss": 1.1557,
+      "step": 13810
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 3.553328037261963,
+      "learning_rate": 4.491580521149393e-05,
+      "loss": 1.2962,
+      "step": 13815
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 1.386233925819397,
+      "learning_rate": 4.4912253306284835e-05,
+      "loss": 1.1249,
+      "step": 13820
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 2.232959747314453,
+      "learning_rate": 4.490870030134268e-05,
+      "loss": 1.0412,
+      "step": 13825
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 1.5061743259429932,
+      "learning_rate": 4.49051461968637e-05,
+      "loss": 1.2668,
+      "step": 13830
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 2.83976674079895,
+      "learning_rate": 4.49015909930442e-05,
+      "loss": 1.3338,
+      "step": 13835
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 1.8282452821731567,
+      "learning_rate": 4.48980346900805e-05,
+      "loss": 1.2672,
+      "step": 13840
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 1.090014100074768,
+      "learning_rate": 4.489447728816904e-05,
+      "loss": 1.1899,
+      "step": 13845
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 2.0513787269592285,
+      "learning_rate": 4.489091878750627e-05,
+      "loss": 1.2211,
+      "step": 13850
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 1.3292232751846313,
+      "learning_rate": 4.4887359188288724e-05,
+      "loss": 1.199,
+      "step": 13855
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 2.754366397857666,
+      "learning_rate": 4.4883798490713014e-05,
+      "loss": 1.2695,
+      "step": 13860
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 1.0987385511398315,
+      "learning_rate": 4.488023669497578e-05,
+      "loss": 1.327,
+      "step": 13865
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.9801754951477051,
+      "learning_rate": 4.487667380127373e-05,
+      "loss": 1.291,
+      "step": 13870
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 1.8694130182266235,
+      "learning_rate": 4.4873109809803654e-05,
+      "loss": 1.3247,
+      "step": 13875
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 1.1474727392196655,
+      "learning_rate": 4.486954472076238e-05,
+      "loss": 1.2209,
+      "step": 13880
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 1.853026032447815,
+      "learning_rate": 4.48659785343468e-05,
+      "loss": 1.1524,
+      "step": 13885
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 1.3957430124282837,
+      "learning_rate": 4.4862411250753875e-05,
+      "loss": 1.3743,
+      "step": 13890
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 1.5522494316101074,
+      "learning_rate": 4.485884287018063e-05,
+      "loss": 1.2599,
+      "step": 13895
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 2.7096736431121826,
+      "learning_rate": 4.485527339282412e-05,
+      "loss": 1.3198,
+      "step": 13900
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 1.6966811418533325,
+      "learning_rate": 4.485170281888151e-05,
+      "loss": 1.2198,
+      "step": 13905
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 1.197799563407898,
+      "learning_rate": 4.484813114854999e-05,
+      "loss": 1.3353,
+      "step": 13910
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 1.5476326942443848,
+      "learning_rate": 4.4844558382026814e-05,
+      "loss": 1.1109,
+      "step": 13915
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 2.144993543624878,
+      "learning_rate": 4.484098451950931e-05,
+      "loss": 1.2156,
+      "step": 13920
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 1.5577327013015747,
+      "learning_rate": 4.483740956119485e-05,
+      "loss": 1.3845,
+      "step": 13925
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 1.9885210990905762,
+      "learning_rate": 4.4833833507280884e-05,
+      "loss": 1.2396,
+      "step": 13930
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 1.833871841430664,
+      "learning_rate": 4.483025635796491e-05,
+      "loss": 1.3424,
+      "step": 13935
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 2.1424660682678223,
+      "learning_rate": 4.482667811344448e-05,
+      "loss": 1.2628,
+      "step": 13940
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 2.6087942123413086,
+      "learning_rate": 4.4823098773917235e-05,
+      "loss": 1.1262,
+      "step": 13945
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 1.847556710243225,
+      "learning_rate": 4.4819518339580844e-05,
+      "loss": 1.2522,
+      "step": 13950
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 1.6152390241622925,
+      "learning_rate": 4.4815936810633066e-05,
+      "loss": 1.4579,
+      "step": 13955
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 3.4430344104766846,
+      "learning_rate": 4.4812354187271686e-05,
+      "loss": 1.2838,
+      "step": 13960
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 1.8315435647964478,
+      "learning_rate": 4.4808770469694584e-05,
+      "loss": 1.322,
+      "step": 13965
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 1.6161450147628784,
+      "learning_rate": 4.480518565809967e-05,
+      "loss": 1.4559,
+      "step": 13970
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 1.8581215143203735,
+      "learning_rate": 4.480159975268494e-05,
+      "loss": 1.2889,
+      "step": 13975
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 1.6320557594299316,
+      "learning_rate": 4.479801275364845e-05,
+      "loss": 1.3009,
+      "step": 13980
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 1.0718029737472534,
+      "learning_rate": 4.4794424661188286e-05,
+      "loss": 1.3986,
+      "step": 13985
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 2.1297664642333984,
+      "learning_rate": 4.479083547550263e-05,
+      "loss": 1.1725,
+      "step": 13990
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 2.0025558471679688,
+      "learning_rate": 4.478724519678969e-05,
+      "loss": 1.1128,
+      "step": 13995
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.8468442559242249,
+      "learning_rate": 4.4783653825247776e-05,
+      "loss": 1.1415,
+      "step": 14000
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 1.221956491470337,
+      "learning_rate": 4.478006136107522e-05,
+      "loss": 1.4145,
+      "step": 14005
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 2.629194974899292,
+      "learning_rate": 4.477646780447043e-05,
+      "loss": 1.4677,
+      "step": 14010
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 1.3865878582000732,
+      "learning_rate": 4.477287315563189e-05,
+      "loss": 1.1557,
+      "step": 14015
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 1.6378535032272339,
+      "learning_rate": 4.4769277414758115e-05,
+      "loss": 1.1965,
+      "step": 14020
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 7.970203876495361,
+      "learning_rate": 4.47656805820477e-05,
+      "loss": 1.2789,
+      "step": 14025
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 2.070845365524292,
+      "learning_rate": 4.476208265769929e-05,
+      "loss": 1.2761,
+      "step": 14030
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 4.676011085510254,
+      "learning_rate": 4.475848364191159e-05,
+      "loss": 1.2767,
+      "step": 14035
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 2.338252305984497,
+      "learning_rate": 4.4754883534883384e-05,
+      "loss": 1.3701,
+      "step": 14040
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 1.2467647790908813,
+      "learning_rate": 4.4751282336813494e-05,
+      "loss": 1.1699,
+      "step": 14045
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 1.1198375225067139,
+      "learning_rate": 4.474768004790081e-05,
+      "loss": 1.3193,
+      "step": 14050
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 2.231294870376587,
+      "learning_rate": 4.474407666834428e-05,
+      "loss": 1.3017,
+      "step": 14055
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 1.3994890451431274,
+      "learning_rate": 4.474047219834292e-05,
+      "loss": 1.181,
+      "step": 14060
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 1.440882921218872,
+      "learning_rate": 4.47368666380958e-05,
+      "loss": 1.3528,
+      "step": 14065
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 2.0190446376800537,
+      "learning_rate": 4.4733259987802046e-05,
+      "loss": 1.3357,
+      "step": 14070
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 1.6590640544891357,
+      "learning_rate": 4.4729652247660855e-05,
+      "loss": 1.2267,
+      "step": 14075
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 1.370227336883545,
+      "learning_rate": 4.472604341787149e-05,
+      "loss": 1.2786,
+      "step": 14080
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 1.0957473516464233,
+      "learning_rate": 4.472243349863324e-05,
+      "loss": 1.11,
+      "step": 14085
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 2.918076992034912,
+      "learning_rate": 4.47188224901455e-05,
+      "loss": 1.2905,
+      "step": 14090
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 1.5842289924621582,
+      "learning_rate": 4.4715210392607675e-05,
+      "loss": 1.1318,
+      "step": 14095
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.9747616052627563,
+      "learning_rate": 4.471159720621928e-05,
+      "loss": 1.3338,
+      "step": 14100
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 1.7682857513427734,
+      "learning_rate": 4.470798293117986e-05,
+      "loss": 1.0593,
+      "step": 14105
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 1.311083436012268,
+      "learning_rate": 4.470436756768903e-05,
+      "loss": 1.1753,
+      "step": 14110
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 1.279892086982727,
+      "learning_rate": 4.470075111594646e-05,
+      "loss": 1.2065,
+      "step": 14115
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 1.927426815032959,
+      "learning_rate": 4.4697133576151885e-05,
+      "loss": 1.3022,
+      "step": 14120
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 2.021284580230713,
+      "learning_rate": 4.4693514948505095e-05,
+      "loss": 1.1597,
+      "step": 14125
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 1.2196049690246582,
+      "learning_rate": 4.4689895233205945e-05,
+      "loss": 1.3091,
+      "step": 14130
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 2.0658938884735107,
+      "learning_rate": 4.4686274430454346e-05,
+      "loss": 1.2713,
+      "step": 14135
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 1.2542707920074463,
+      "learning_rate": 4.4682652540450274e-05,
+      "loss": 1.2778,
+      "step": 14140
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 1.4691362380981445,
+      "learning_rate": 4.4679029563393756e-05,
+      "loss": 1.073,
+      "step": 14145
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 2.216343641281128,
+      "learning_rate": 4.46754054994849e-05,
+      "loss": 1.3336,
+      "step": 14150
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 3.187505006790161,
+      "learning_rate": 4.467178034892384e-05,
+      "loss": 1.2517,
+      "step": 14155
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.3655002117156982,
+      "learning_rate": 4.466815411191081e-05,
+      "loss": 1.3117,
+      "step": 14160
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.2147845029830933,
+      "learning_rate": 4.466452678864607e-05,
+      "loss": 1.4121,
+      "step": 14165
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.8723526000976562,
+      "learning_rate": 4.466089837932995e-05,
+      "loss": 1.2157,
+      "step": 14170
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 2.1111645698547363,
+      "learning_rate": 4.465726888416285e-05,
+      "loss": 1.3665,
+      "step": 14175
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 3.004091739654541,
+      "learning_rate": 4.4653638303345225e-05,
+      "loss": 1.4633,
+      "step": 14180
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 2.4839401245117188,
+      "learning_rate": 4.465000663707758e-05,
+      "loss": 1.3408,
+      "step": 14185
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.8565268516540527,
+      "learning_rate": 4.46463738855605e-05,
+      "loss": 1.2731,
+      "step": 14190
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 2.492748975753784,
+      "learning_rate": 4.464274004899461e-05,
+      "loss": 1.3723,
+      "step": 14195
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 2.4697351455688477,
+      "learning_rate": 4.46391051275806e-05,
+      "loss": 1.1248,
+      "step": 14200
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 2.914813280105591,
+      "learning_rate": 4.4635469121519234e-05,
+      "loss": 1.0966,
+      "step": 14205
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.3312268257141113,
+      "learning_rate": 4.463183203101131e-05,
+      "loss": 1.218,
+      "step": 14210
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 3.274707317352295,
+      "learning_rate": 4.4628193856257714e-05,
+      "loss": 1.1181,
+      "step": 14215
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.0050827264785767,
+      "learning_rate": 4.462455459745938e-05,
+      "loss": 1.28,
+      "step": 14220
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 1.2786221504211426,
+      "learning_rate": 4.462091425481728e-05,
+      "loss": 1.1789,
+      "step": 14225
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 1.9664305448532104,
+      "learning_rate": 4.4617272828532495e-05,
+      "loss": 1.2938,
+      "step": 14230
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 1.6089929342269897,
+      "learning_rate": 4.461363031880611e-05,
+      "loss": 1.2964,
+      "step": 14235
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 2.8070948123931885,
+      "learning_rate": 4.460998672583933e-05,
+      "loss": 1.3383,
+      "step": 14240
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 1.145674705505371,
+      "learning_rate": 4.460634204983334e-05,
+      "loss": 1.3326,
+      "step": 14245
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 2.9494822025299072,
+      "learning_rate": 4.4602696290989477e-05,
+      "loss": 1.2622,
+      "step": 14250
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 2.005913734436035,
+      "learning_rate": 4.459904944950907e-05,
+      "loss": 1.267,
+      "step": 14255
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 1.1085389852523804,
+      "learning_rate": 4.4595401525593525e-05,
+      "loss": 1.2181,
+      "step": 14260
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 3.622462272644043,
+      "learning_rate": 4.459175251944433e-05,
+      "loss": 1.3522,
+      "step": 14265
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 2.914292573928833,
+      "learning_rate": 4.458810243126301e-05,
+      "loss": 1.3473,
+      "step": 14270
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 2.464303493499756,
+      "learning_rate": 4.458445126125115e-05,
+      "loss": 1.1967,
+      "step": 14275
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 1.2096035480499268,
+      "learning_rate": 4.45807990096104e-05,
+      "loss": 1.1164,
+      "step": 14280
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 2.403764247894287,
+      "learning_rate": 4.4577145676542474e-05,
+      "loss": 1.1992,
+      "step": 14285
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.3285304307937622,
+      "learning_rate": 4.4573491262249144e-05,
+      "loss": 1.2245,
+      "step": 14290
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.1528428792953491,
+      "learning_rate": 4.4569835766932244e-05,
+      "loss": 1.3815,
+      "step": 14295
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.0769155025482178,
+      "learning_rate": 4.4566179190793646e-05,
+      "loss": 1.1992,
+      "step": 14300
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.2317618131637573,
+      "learning_rate": 4.4562521534035316e-05,
+      "loss": 1.2237,
+      "step": 14305
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.0465807914733887,
+      "learning_rate": 4.455886279685925e-05,
+      "loss": 1.3325,
+      "step": 14310
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 2.072091579437256,
+      "learning_rate": 4.4555202979467526e-05,
+      "loss": 1.3239,
+      "step": 14315
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.5427582263946533,
+      "learning_rate": 4.455154208206227e-05,
+      "loss": 1.284,
+      "step": 14320
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 1.5451370477676392,
+      "learning_rate": 4.454788010484566e-05,
+      "loss": 1.2576,
+      "step": 14325
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 1.7191996574401855,
+      "learning_rate": 4.454421704801996e-05,
+      "loss": 1.3133,
+      "step": 14330
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 1.3877907991409302,
+      "learning_rate": 4.454055291178746e-05,
+      "loss": 1.1534,
+      "step": 14335
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 1.9861388206481934,
+      "learning_rate": 4.453688769635054e-05,
+      "loss": 1.2559,
+      "step": 14340
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 2.1039087772369385,
+      "learning_rate": 4.453322140191162e-05,
+      "loss": 1.3874,
+      "step": 14345
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 1.7028883695602417,
+      "learning_rate": 4.452955402867318e-05,
+      "loss": 1.3288,
+      "step": 14350
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 2.711984634399414,
+      "learning_rate": 4.452588557683777e-05,
+      "loss": 1.2802,
+      "step": 14355
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 1.9355015754699707,
+      "learning_rate": 4.4522216046608004e-05,
+      "loss": 1.2044,
+      "step": 14360
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 1.0914154052734375,
+      "learning_rate": 4.451854543818653e-05,
+      "loss": 1.0817,
+      "step": 14365
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 1.7667779922485352,
+      "learning_rate": 4.451487375177608e-05,
+      "loss": 1.2535,
+      "step": 14370
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 2.6625638008117676,
+      "learning_rate": 4.4511200987579445e-05,
+      "loss": 1.3501,
+      "step": 14375
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 2.3096730709075928,
+      "learning_rate": 4.450752714579946e-05,
+      "loss": 1.2074,
+      "step": 14380
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 1.5009289979934692,
+      "learning_rate": 4.450385222663902e-05,
+      "loss": 1.2544,
+      "step": 14385
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 2.635723352432251,
+      "learning_rate": 4.4500176230301095e-05,
+      "loss": 1.2226,
+      "step": 14390
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 1.7658405303955078,
+      "learning_rate": 4.449649915698871e-05,
+      "loss": 1.2582,
+      "step": 14395
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 3.9313101768493652,
+      "learning_rate": 4.449282100690494e-05,
+      "loss": 1.2111,
+      "step": 14400
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 2.178057909011841,
+      "learning_rate": 4.448914178025293e-05,
+      "loss": 1.4299,
+      "step": 14405
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 1.1012213230133057,
+      "learning_rate": 4.4485461477235865e-05,
+      "loss": 1.0573,
+      "step": 14410
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 1.9754661321640015,
+      "learning_rate": 4.448178009805704e-05,
+      "loss": 1.2845,
+      "step": 14415
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 1.2951463460922241,
+      "learning_rate": 4.4478097642919734e-05,
+      "loss": 1.2188,
+      "step": 14420
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.210935354232788,
+      "learning_rate": 4.447441411202734e-05,
+      "loss": 1.3247,
+      "step": 14425
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.0859736204147339,
+      "learning_rate": 4.447072950558331e-05,
+      "loss": 1.2225,
+      "step": 14430
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.7958767414093018,
+      "learning_rate": 4.446704382379111e-05,
+      "loss": 1.3079,
+      "step": 14435
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.6416263580322266,
+      "learning_rate": 4.4463357066854326e-05,
+      "loss": 1.1547,
+      "step": 14440
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.793069839477539,
+      "learning_rate": 4.445966923497656e-05,
+      "loss": 1.388,
+      "step": 14445
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 2.6959428787231445,
+      "learning_rate": 4.44559803283615e-05,
+      "loss": 1.2552,
+      "step": 14450
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 3.44150710105896,
+      "learning_rate": 4.445229034721285e-05,
+      "loss": 1.2653,
+      "step": 14455
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 2.6374642848968506,
+      "learning_rate": 4.444859929173444e-05,
+      "loss": 1.204,
+      "step": 14460
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 1.654412031173706,
+      "learning_rate": 4.44449071621301e-05,
+      "loss": 1.3707,
+      "step": 14465
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 1.6137704849243164,
+      "learning_rate": 4.444121395860375e-05,
+      "loss": 1.29,
+      "step": 14470
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 1.6746236085891724,
+      "learning_rate": 4.443751968135936e-05,
+      "loss": 1.4198,
+      "step": 14475
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 1.7420883178710938,
+      "learning_rate": 4.4433824330600964e-05,
+      "loss": 1.245,
+      "step": 14480
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 5.652534484863281,
+      "learning_rate": 4.443012790653265e-05,
+      "loss": 1.2699,
+      "step": 14485
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 2.45516037940979,
+      "learning_rate": 4.442643040935856e-05,
+      "loss": 1.1824,
+      "step": 14490
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 2.3738396167755127,
+      "learning_rate": 4.442273183928293e-05,
+      "loss": 1.1621,
+      "step": 14495
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 1.6174094676971436,
+      "learning_rate": 4.4419032196509994e-05,
+      "loss": 1.3374,
+      "step": 14500
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 2.4241867065429688,
+      "learning_rate": 4.44153314812441e-05,
+      "loss": 1.1839,
+      "step": 14505
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 2.8648736476898193,
+      "learning_rate": 4.4411629693689626e-05,
+      "loss": 1.2719,
+      "step": 14510
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 2.2779765129089355,
+      "learning_rate": 4.4407926834051025e-05,
+      "loss": 1.2144,
+      "step": 14515
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 3.223737955093384,
+      "learning_rate": 4.44042229025328e-05,
+      "loss": 1.112,
+      "step": 14520
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.5524216890335083,
+      "learning_rate": 4.440051789933951e-05,
+      "loss": 1.2005,
+      "step": 14525
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 2.5354983806610107,
+      "learning_rate": 4.439681182467579e-05,
+      "loss": 1.0579,
+      "step": 14530
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.221150279045105,
+      "learning_rate": 4.43931046787463e-05,
+      "loss": 1.3449,
+      "step": 14535
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.6341768503189087,
+      "learning_rate": 4.4389396461755804e-05,
+      "loss": 1.2786,
+      "step": 14540
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.0187619924545288,
+      "learning_rate": 4.4385687173909093e-05,
+      "loss": 1.303,
+      "step": 14545
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.5471796989440918,
+      "learning_rate": 4.4381976815411036e-05,
+      "loss": 1.1909,
+      "step": 14550
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.6268233060836792,
+      "learning_rate": 4.437826538646655e-05,
+      "loss": 1.2544,
+      "step": 14555
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 1.3081377744674683,
+      "learning_rate": 4.4374552887280594e-05,
+      "loss": 1.3514,
+      "step": 14560
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 1.7368911504745483,
+      "learning_rate": 4.4370839318058235e-05,
+      "loss": 1.3513,
+      "step": 14565
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.9481498003005981,
+      "learning_rate": 4.4367124679004545e-05,
+      "loss": 1.2849,
+      "step": 14570
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 1.1011961698532104,
+      "learning_rate": 4.4363408970324696e-05,
+      "loss": 1.3394,
+      "step": 14575
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.9509512782096863,
+      "learning_rate": 4.435969219222389e-05,
+      "loss": 1.1828,
+      "step": 14580
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.9883051514625549,
+      "learning_rate": 4.4355974344907414e-05,
+      "loss": 1.1986,
+      "step": 14585
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 1.3182764053344727,
+      "learning_rate": 4.4352255428580595e-05,
+      "loss": 1.4436,
+      "step": 14590
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 1.5108343362808228,
+      "learning_rate": 4.434853544344882e-05,
+      "loss": 1.2572,
+      "step": 14595
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 1.2329615354537964,
+      "learning_rate": 4.434481438971754e-05,
+      "loss": 1.16,
+      "step": 14600
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 2.406236171722412,
+      "learning_rate": 4.4341092267592276e-05,
+      "loss": 1.2099,
+      "step": 14605
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 2.0363478660583496,
+      "learning_rate": 4.433736907727859e-05,
+      "loss": 1.2245,
+      "step": 14610
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 1.3022775650024414,
+      "learning_rate": 4.4333644818982115e-05,
+      "loss": 1.2434,
+      "step": 14615
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.9420449733734131,
+      "learning_rate": 4.4329919492908526e-05,
+      "loss": 1.2814,
+      "step": 14620
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 4.690912246704102,
+      "learning_rate": 4.432619309926357e-05,
+      "loss": 1.2965,
+      "step": 14625
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 2.9743857383728027,
+      "learning_rate": 4.432246563825306e-05,
+      "loss": 1.1934,
+      "step": 14630
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 1.6823005676269531,
+      "learning_rate": 4.431873711008286e-05,
+      "loss": 1.3078,
+      "step": 14635
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 2.596446990966797,
+      "learning_rate": 4.4315007514958896e-05,
+      "loss": 1.1404,
+      "step": 14640
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 2.1992132663726807,
+      "learning_rate": 4.4311276853087144e-05,
+      "loss": 1.0836,
+      "step": 14645
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 1.4632948637008667,
+      "learning_rate": 4.430754512467364e-05,
+      "loss": 1.2743,
+      "step": 14650
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 1.9432785511016846,
+      "learning_rate": 4.430381232992449e-05,
+      "loss": 1.1317,
+      "step": 14655
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 2.028427839279175,
+      "learning_rate": 4.430007846904585e-05,
+      "loss": 1.135,
+      "step": 14660
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 1.5426450967788696,
+      "learning_rate": 4.4296343542243926e-05,
+      "loss": 1.3249,
+      "step": 14665
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 1.5564109086990356,
+      "learning_rate": 4.4292607549725016e-05,
+      "loss": 1.269,
+      "step": 14670
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 1.5921878814697266,
+      "learning_rate": 4.428887049169544e-05,
+      "loss": 1.3068,
+      "step": 14675
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 1.6243040561676025,
+      "learning_rate": 4.4285132368361606e-05,
+      "loss": 1.0717,
+      "step": 14680
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 2.8081164360046387,
+      "learning_rate": 4.428139317992995e-05,
+      "loss": 1.1523,
+      "step": 14685
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 3.132068157196045,
+      "learning_rate": 4.4277652926607e-05,
+      "loss": 1.3471,
+      "step": 14690
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 2.9057159423828125,
+      "learning_rate": 4.42739116085993e-05,
+      "loss": 1.2094,
+      "step": 14695
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.643255591392517,
+      "learning_rate": 4.42701692261135e-05,
+      "loss": 1.5053,
+      "step": 14700
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.549208164215088,
+      "learning_rate": 4.426642577935629e-05,
+      "loss": 1.1451,
+      "step": 14705
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.8293100595474243,
+      "learning_rate": 4.426268126853441e-05,
+      "loss": 1.2295,
+      "step": 14710
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 2.3333680629730225,
+      "learning_rate": 4.425893569385466e-05,
+      "loss": 1.1442,
+      "step": 14715
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.7189589738845825,
+      "learning_rate": 4.425518905552392e-05,
+      "loss": 1.3381,
+      "step": 14720
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 3.104936361312866,
+      "learning_rate": 4.42514413537491e-05,
+      "loss": 1.2112,
+      "step": 14725
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 1.1761854887008667,
+      "learning_rate": 4.424769258873718e-05,
+      "loss": 1.2502,
+      "step": 14730
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 1.4067808389663696,
+      "learning_rate": 4.424394276069521e-05,
+      "loss": 1.3833,
+      "step": 14735
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 1.7447682619094849,
+      "learning_rate": 4.424019186983028e-05,
+      "loss": 1.203,
+      "step": 14740
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 1.739342212677002,
+      "learning_rate": 4.423643991634956e-05,
+      "loss": 1.2031,
+      "step": 14745
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 2.5015289783477783,
+      "learning_rate": 4.423268690046025e-05,
+      "loss": 1.4271,
+      "step": 14750
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 1.010996699333191,
+      "learning_rate": 4.422893282236963e-05,
+      "loss": 1.2328,
+      "step": 14755
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 1.5030901432037354,
+      "learning_rate": 4.422517768228505e-05,
+      "loss": 1.3878,
+      "step": 14760
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 1.2070742845535278,
+      "learning_rate": 4.422142148041388e-05,
+      "loss": 1.3728,
+      "step": 14765
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 1.405918002128601,
+      "learning_rate": 4.421766421696358e-05,
+      "loss": 1.3906,
+      "step": 14770
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 2.8007607460021973,
+      "learning_rate": 4.4213905892141674e-05,
+      "loss": 1.3337,
+      "step": 14775
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 1.6845779418945312,
+      "learning_rate": 4.421014650615571e-05,
+      "loss": 1.3817,
+      "step": 14780
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 2.3008716106414795,
+      "learning_rate": 4.420638605921332e-05,
+      "loss": 1.2791,
+      "step": 14785
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 2.7502546310424805,
+      "learning_rate": 4.4202624551522195e-05,
+      "loss": 1.1827,
+      "step": 14790
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 1.4934040307998657,
+      "learning_rate": 4.419886198329008e-05,
+      "loss": 1.4211,
+      "step": 14795
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 2.8900766372680664,
+      "learning_rate": 4.419509835472476e-05,
+      "loss": 1.3247,
+      "step": 14800
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 1.89850652217865,
+      "learning_rate": 4.4191333666034124e-05,
+      "loss": 1.3627,
+      "step": 14805
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 2.504714250564575,
+      "learning_rate": 4.4187567917426074e-05,
+      "loss": 1.2384,
+      "step": 14810
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 2.1975607872009277,
+      "learning_rate": 4.418380110910859e-05,
+      "loss": 1.1535,
+      "step": 14815
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 2.044182062149048,
+      "learning_rate": 4.4180033241289706e-05,
+      "loss": 1.2568,
+      "step": 14820
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 1.6117463111877441,
+      "learning_rate": 4.417626431417753e-05,
+      "loss": 1.0122,
+      "step": 14825
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 1.8464300632476807,
+      "learning_rate": 4.4172494327980205e-05,
+      "loss": 1.19,
+      "step": 14830
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 2.2936418056488037,
+      "learning_rate": 4.4168723282905954e-05,
+      "loss": 1.0822,
+      "step": 14835
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 1.4737282991409302,
+      "learning_rate": 4.416495117916304e-05,
+      "loss": 1.4572,
+      "step": 14840
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 1.4652832746505737,
+      "learning_rate": 4.416117801695979e-05,
+      "loss": 1.219,
+      "step": 14845
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 1.4503779411315918,
+      "learning_rate": 4.415740379650459e-05,
+      "loss": 1.3568,
+      "step": 14850
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 1.4976866245269775,
+      "learning_rate": 4.41536285180059e-05,
+      "loss": 1.1892,
+      "step": 14855
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 5.945674896240234,
+      "learning_rate": 4.414985218167221e-05,
+      "loss": 1.146,
+      "step": 14860
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 1.585028886795044,
+      "learning_rate": 4.414607478771209e-05,
+      "loss": 1.1714,
+      "step": 14865
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 1.4117640256881714,
+      "learning_rate": 4.4142296336334166e-05,
+      "loss": 1.2991,
+      "step": 14870
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 2.092742443084717,
+      "learning_rate": 4.41385168277471e-05,
+      "loss": 1.1138,
+      "step": 14875
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 2.3509175777435303,
+      "learning_rate": 4.413473626215965e-05,
+      "loss": 1.3,
+      "step": 14880
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 2.7636947631835938,
+      "learning_rate": 4.4130954639780616e-05,
+      "loss": 1.2863,
+      "step": 14885
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 1.4940133094787598,
+      "learning_rate": 4.412717196081883e-05,
+      "loss": 1.2997,
+      "step": 14890
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 2.2368600368499756,
+      "learning_rate": 4.412338822548322e-05,
+      "loss": 1.2766,
+      "step": 14895
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 1.0473464727401733,
+      "learning_rate": 4.411960343398276e-05,
+      "loss": 1.2436,
+      "step": 14900
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 1.2610361576080322,
+      "learning_rate": 4.4115817586526475e-05,
+      "loss": 1.2788,
+      "step": 14905
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 1.074568271636963,
+      "learning_rate": 4.411203068332345e-05,
+      "loss": 1.2546,
+      "step": 14910
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 1.4182261228561401,
+      "learning_rate": 4.4108242724582836e-05,
+      "loss": 1.0938,
+      "step": 14915
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 5.393111705780029,
+      "learning_rate": 4.410445371051385e-05,
+      "loss": 1.2659,
+      "step": 14920
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 1.6223890781402588,
+      "learning_rate": 4.410066364132573e-05,
+      "loss": 1.3116,
+      "step": 14925
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 2.0401763916015625,
+      "learning_rate": 4.409687251722782e-05,
+      "loss": 1.2728,
+      "step": 14930
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 2.3179221153259277,
+      "learning_rate": 4.4093080338429485e-05,
+      "loss": 0.9557,
+      "step": 14935
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.9907518029212952,
+      "learning_rate": 4.4089287105140176e-05,
+      "loss": 1.2843,
+      "step": 14940
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 3.440904378890991,
+      "learning_rate": 4.4085492817569375e-05,
+      "loss": 1.2877,
+      "step": 14945
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 1.7463809251785278,
+      "learning_rate": 4.408169747592665e-05,
+      "loss": 1.2269,
+      "step": 14950
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 1.2096577882766724,
+      "learning_rate": 4.40779010804216e-05,
+      "loss": 1.318,
+      "step": 14955
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 1.5482676029205322,
+      "learning_rate": 4.407410363126391e-05,
+      "loss": 1.2836,
+      "step": 14960
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.8990564346313477,
+      "learning_rate": 4.4070305128663306e-05,
+      "loss": 1.175,
+      "step": 14965
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 1.5660568475723267,
+      "learning_rate": 4.406650557282957e-05,
+      "loss": 1.0142,
+      "step": 14970
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 1.8954380750656128,
+      "learning_rate": 4.406270496397254e-05,
+      "loss": 1.1563,
+      "step": 14975
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 3.845919609069824,
+      "learning_rate": 4.4058903302302136e-05,
+      "loss": 1.1624,
+      "step": 14980
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 2.766300678253174,
+      "learning_rate": 4.4055100588028315e-05,
+      "loss": 1.2274,
+      "step": 14985
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 1.7678008079528809,
+      "learning_rate": 4.405129682136109e-05,
+      "loss": 1.2156,
+      "step": 14990
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 1.1421842575073242,
+      "learning_rate": 4.404749200251055e-05,
+      "loss": 1.2917,
+      "step": 14995
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 2.411649227142334,
+      "learning_rate": 4.4043686131686825e-05,
+      "loss": 1.177,
+      "step": 15000
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.9833216667175293,
+      "learning_rate": 4.403987920910011e-05,
+      "loss": 1.4059,
+      "step": 15005
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 1.068702220916748,
+      "learning_rate": 4.403607123496065e-05,
+      "loss": 1.0853,
+      "step": 15010
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 2.074406385421753,
+      "learning_rate": 4.4032262209478774e-05,
+      "loss": 1.4351,
+      "step": 15015
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 2.2396271228790283,
+      "learning_rate": 4.402845213286483e-05,
+      "loss": 1.3586,
+      "step": 15020
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 2.4395458698272705,
+      "learning_rate": 4.402464100532926e-05,
+      "loss": 1.2868,
+      "step": 15025
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 2.2216920852661133,
+      "learning_rate": 4.402082882708254e-05,
+      "loss": 1.3321,
+      "step": 15030
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 1.8366425037384033,
+      "learning_rate": 4.401701559833521e-05,
+      "loss": 1.3224,
+      "step": 15035
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 2.4641847610473633,
+      "learning_rate": 4.401320131929788e-05,
+      "loss": 1.0611,
+      "step": 15040
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 1.1396476030349731,
+      "learning_rate": 4.400938599018121e-05,
+      "loss": 1.3745,
+      "step": 15045
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 2.272923469543457,
+      "learning_rate": 4.40055696111959e-05,
+      "loss": 1.2228,
+      "step": 15050
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 3.0234322547912598,
+      "learning_rate": 4.400175218255274e-05,
+      "loss": 1.2259,
+      "step": 15055
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 4.0134501457214355,
+      "learning_rate": 4.3997933704462555e-05,
+      "loss": 1.1956,
+      "step": 15060
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 1.733201503753662,
+      "learning_rate": 4.399411417713625e-05,
+      "loss": 1.3304,
+      "step": 15065
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 1.0404208898544312,
+      "learning_rate": 4.3990293600784754e-05,
+      "loss": 1.2417,
+      "step": 15070
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 1.6445467472076416,
+      "learning_rate": 4.398647197561908e-05,
+      "loss": 1.4209,
+      "step": 15075
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 2.307591676712036,
+      "learning_rate": 4.3982649301850296e-05,
+      "loss": 1.4352,
+      "step": 15080
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 1.3652942180633545,
+      "learning_rate": 4.397882557968952e-05,
+      "loss": 1.4074,
+      "step": 15085
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 1.15928053855896,
+      "learning_rate": 4.397500080934794e-05,
+      "loss": 1.175,
+      "step": 15090
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 1.9573633670806885,
+      "learning_rate": 4.397117499103679e-05,
+      "loss": 1.3634,
+      "step": 15095
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 2.113678455352783,
+      "learning_rate": 4.3967348124967365e-05,
+      "loss": 1.1655,
+      "step": 15100
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 2.7751240730285645,
+      "learning_rate": 4.396352021135101e-05,
+      "loss": 1.1952,
+      "step": 15105
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 3.2183516025543213,
+      "learning_rate": 4.395969125039915e-05,
+      "loss": 1.4013,
+      "step": 15110
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 3.216507911682129,
+      "learning_rate": 4.395586124232325e-05,
+      "loss": 1.2403,
+      "step": 15115
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 1.8810077905654907,
+      "learning_rate": 4.395203018733484e-05,
+      "loss": 1.3741,
+      "step": 15120
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 1.4608815908432007,
+      "learning_rate": 4.3948198085645495e-05,
+      "loss": 1.2175,
+      "step": 15125
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 2.5550155639648438,
+      "learning_rate": 4.394436493746687e-05,
+      "loss": 1.3009,
+      "step": 15130
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 1.188637375831604,
+      "learning_rate": 4.394053074301066e-05,
+      "loss": 1.3576,
+      "step": 15135
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 1.8632690906524658,
+      "learning_rate": 4.3936695502488623e-05,
+      "loss": 1.2394,
+      "step": 15140
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 2.738532304763794,
+      "learning_rate": 4.3932859216112584e-05,
+      "loss": 1.2289,
+      "step": 15145
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 1.352799415588379,
+      "learning_rate": 4.39290218840944e-05,
+      "loss": 1.2893,
+      "step": 15150
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 1.2402245998382568,
+      "learning_rate": 4.392518350664602e-05,
+      "loss": 1.2977,
+      "step": 15155
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 1.9382325410842896,
+      "learning_rate": 4.392134408397942e-05,
+      "loss": 1.197,
+      "step": 15160
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 3.8150017261505127,
+      "learning_rate": 4.391750361630666e-05,
+      "loss": 1.3319,
+      "step": 15165
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 1.3974863290786743,
+      "learning_rate": 4.391366210383984e-05,
+      "loss": 1.2548,
+      "step": 15170
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 1.6200858354568481,
+      "learning_rate": 4.390981954679112e-05,
+      "loss": 1.3474,
+      "step": 15175
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 2.464508056640625,
+      "learning_rate": 4.390597594537272e-05,
+      "loss": 1.2525,
+      "step": 15180
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 2.1834940910339355,
+      "learning_rate": 4.390213129979692e-05,
+      "loss": 1.2651,
+      "step": 15185
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 1.6160869598388672,
+      "learning_rate": 4.3898285610276056e-05,
+      "loss": 1.3427,
+      "step": 15190
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 2.14302921295166,
+      "learning_rate": 4.389443887702252e-05,
+      "loss": 1.3329,
+      "step": 15195
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 1.1593176126480103,
+      "learning_rate": 4.3890591100248773e-05,
+      "loss": 1.2283,
+      "step": 15200
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 1.1334714889526367,
+      "learning_rate": 4.388674228016731e-05,
+      "loss": 1.2044,
+      "step": 15205
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 1.1369993686676025,
+      "learning_rate": 4.38828924169907e-05,
+      "loss": 1.1575,
+      "step": 15210
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 1.315269112586975,
+      "learning_rate": 4.387904151093157e-05,
+      "loss": 1.3368,
+      "step": 15215
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 1.833358883857727,
+      "learning_rate": 4.38751895622026e-05,
+      "loss": 1.2118,
+      "step": 15220
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 2.8720498085021973,
+      "learning_rate": 4.387133657101654e-05,
+      "loss": 1.2441,
+      "step": 15225
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 2.276634931564331,
+      "learning_rate": 4.386748253758617e-05,
+      "loss": 1.1128,
+      "step": 15230
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 1.893906831741333,
+      "learning_rate": 4.386362746212435e-05,
+      "loss": 1.2714,
+      "step": 15235
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 2.4002392292022705,
+      "learning_rate": 4.385977134484399e-05,
+      "loss": 1.2763,
+      "step": 15240
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 1.7395859956741333,
+      "learning_rate": 4.385591418595807e-05,
+      "loss": 1.2394,
+      "step": 15245
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 2.103240966796875,
+      "learning_rate": 4.38520559856796e-05,
+      "loss": 1.2259,
+      "step": 15250
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 1.817021369934082,
+      "learning_rate": 4.384819674422168e-05,
+      "loss": 1.4099,
+      "step": 15255
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 3.440264940261841,
+      "learning_rate": 4.3844336461797445e-05,
+      "loss": 1.1517,
+      "step": 15260
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 2.334442615509033,
+      "learning_rate": 4.384047513862009e-05,
+      "loss": 1.1874,
+      "step": 15265
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 1.8950352668762207,
+      "learning_rate": 4.383661277490289e-05,
+      "loss": 1.2903,
+      "step": 15270
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 1.6918821334838867,
+      "learning_rate": 4.3832749370859124e-05,
+      "loss": 1.1923,
+      "step": 15275
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 2.603799819946289,
+      "learning_rate": 4.38288849267022e-05,
+      "loss": 1.3905,
+      "step": 15280
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 1.2318238019943237,
+      "learning_rate": 4.3825019442645534e-05,
+      "loss": 1.2528,
+      "step": 15285
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 2.4402077198028564,
+      "learning_rate": 4.382115291890261e-05,
+      "loss": 1.2203,
+      "step": 15290
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 1.929358720779419,
+      "learning_rate": 4.3817285355686973e-05,
+      "loss": 1.2942,
+      "step": 15295
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 2.6186788082122803,
+      "learning_rate": 4.3813416753212224e-05,
+      "loss": 1.1942,
+      "step": 15300
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 3.274463653564453,
+      "learning_rate": 4.3809547111692024e-05,
+      "loss": 1.3078,
+      "step": 15305
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 3.3684141635894775,
+      "learning_rate": 4.380567643134009e-05,
+      "loss": 1.2054,
+      "step": 15310
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.9591112732887268,
+      "learning_rate": 4.380180471237019e-05,
+      "loss": 1.1912,
+      "step": 15315
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 3.8308017253875732,
+      "learning_rate": 4.379793195499616e-05,
+      "loss": 1.2198,
+      "step": 15320
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 1.5418450832366943,
+      "learning_rate": 4.3794058159431895e-05,
+      "loss": 1.3401,
+      "step": 15325
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 1.2798864841461182,
+      "learning_rate": 4.379018332589132e-05,
+      "loss": 1.0988,
+      "step": 15330
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 1.1847316026687622,
+      "learning_rate": 4.378630745458846e-05,
+      "loss": 1.378,
+      "step": 15335
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 2.026315450668335,
+      "learning_rate": 4.3782430545737377e-05,
+      "loss": 1.267,
+      "step": 15340
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 1.8548667430877686,
+      "learning_rate": 4.3778552599552156e-05,
+      "loss": 1.2185,
+      "step": 15345
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 1.9220659732818604,
+      "learning_rate": 4.377467361624701e-05,
+      "loss": 1.2383,
+      "step": 15350
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 1.5379008054733276,
+      "learning_rate": 4.3770793596036145e-05,
+      "loss": 1.2225,
+      "step": 15355
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.2834293842315674,
+      "learning_rate": 4.3766912539133864e-05,
+      "loss": 1.1713,
+      "step": 15360
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.6671650409698486,
+      "learning_rate": 4.3763030445754514e-05,
+      "loss": 1.3282,
+      "step": 15365
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.0815520286560059,
+      "learning_rate": 4.37591473161125e-05,
+      "loss": 1.3631,
+      "step": 15370
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 2.721876859664917,
+      "learning_rate": 4.375526315042227e-05,
+      "loss": 1.1959,
+      "step": 15375
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.370292067527771,
+      "learning_rate": 4.3751377948898356e-05,
+      "loss": 1.2007,
+      "step": 15380
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 2.0268495082855225,
+      "learning_rate": 4.374749171175533e-05,
+      "loss": 1.243,
+      "step": 15385
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.4112547636032104,
+      "learning_rate": 4.3743604439207817e-05,
+      "loss": 1.1799,
+      "step": 15390
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 1.6107220649719238,
+      "learning_rate": 4.373971613147051e-05,
+      "loss": 1.3369,
+      "step": 15395
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 2.6075730323791504,
+      "learning_rate": 4.373582678875817e-05,
+      "loss": 1.2638,
+      "step": 15400
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 1.0856952667236328,
+      "learning_rate": 4.373193641128559e-05,
+      "loss": 1.2009,
+      "step": 15405
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 2.2727081775665283,
+      "learning_rate": 4.372804499926762e-05,
+      "loss": 1.2948,
+      "step": 15410
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 1.7927814722061157,
+      "learning_rate": 4.3724152552919214e-05,
+      "loss": 1.1401,
+      "step": 15415
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 2.271836996078491,
+      "learning_rate": 4.372025907245532e-05,
+      "loss": 1.2022,
+      "step": 15420
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 2.37410306930542,
+      "learning_rate": 4.371636455809096e-05,
+      "loss": 1.2243,
+      "step": 15425
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 1.7014080286026,
+      "learning_rate": 4.371246901004125e-05,
+      "loss": 1.2851,
+      "step": 15430
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 2.186640977859497,
+      "learning_rate": 4.3708572428521334e-05,
+      "loss": 1.4657,
+      "step": 15435
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 2.204148769378662,
+      "learning_rate": 4.370467481374639e-05,
+      "loss": 1.1876,
+      "step": 15440
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 1.9742987155914307,
+      "learning_rate": 4.3700776165931716e-05,
+      "loss": 1.2819,
+      "step": 15445
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 1.3668123483657837,
+      "learning_rate": 4.369687648529261e-05,
+      "loss": 1.2395,
+      "step": 15450
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 3.831627368927002,
+      "learning_rate": 4.3692975772044444e-05,
+      "loss": 1.1898,
+      "step": 15455
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 1.8598977327346802,
+      "learning_rate": 4.3689074026402665e-05,
+      "loss": 1.2271,
+      "step": 15460
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 1.0298603773117065,
+      "learning_rate": 4.3685171248582744e-05,
+      "loss": 1.2812,
+      "step": 15465
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 2.9451205730438232,
+      "learning_rate": 4.368126743880024e-05,
+      "loss": 1.2084,
+      "step": 15470
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 1.8287848234176636,
+      "learning_rate": 4.367736259727076e-05,
+      "loss": 1.2462,
+      "step": 15475
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 1.9537088871002197,
+      "learning_rate": 4.367345672420995e-05,
+      "loss": 1.2031,
+      "step": 15480
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 2.9025750160217285,
+      "learning_rate": 4.366954981983354e-05,
+      "loss": 1.4563,
+      "step": 15485
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 1.587408185005188,
+      "learning_rate": 4.3665641884357294e-05,
+      "loss": 1.1538,
+      "step": 15490
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 1.1759958267211914,
+      "learning_rate": 4.366173291799705e-05,
+      "loss": 1.2645,
+      "step": 15495
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 4.82040548324585,
+      "learning_rate": 4.3657822920968706e-05,
+      "loss": 1.1618,
+      "step": 15500
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.7620186805725098,
+      "learning_rate": 4.365391189348818e-05,
+      "loss": 1.3226,
+      "step": 15505
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 1.329471230506897,
+      "learning_rate": 4.3649999835771496e-05,
+      "loss": 1.2517,
+      "step": 15510
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 1.040663480758667,
+      "learning_rate": 4.364608674803471e-05,
+      "loss": 1.2898,
+      "step": 15515
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 2.1228902339935303,
+      "learning_rate": 4.3642172630493925e-05,
+      "loss": 1.2581,
+      "step": 15520
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 1.2862385511398315,
+      "learning_rate": 4.3638257483365336e-05,
+      "loss": 1.2414,
+      "step": 15525
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 1.5629061460494995,
+      "learning_rate": 4.363434130686515e-05,
+      "loss": 1.201,
+      "step": 15530
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 1.894118309020996,
+      "learning_rate": 4.3630424101209663e-05,
+      "loss": 1.2576,
+      "step": 15535
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 4.182260513305664,
+      "learning_rate": 4.3626505866615224e-05,
+      "loss": 1.1643,
+      "step": 15540
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 2.6900877952575684,
+      "learning_rate": 4.362258660329822e-05,
+      "loss": 1.1906,
+      "step": 15545
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 9.078900337219238,
+      "learning_rate": 4.361866631147512e-05,
+      "loss": 1.2331,
+      "step": 15550
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 3.6981396675109863,
+      "learning_rate": 4.361474499136243e-05,
+      "loss": 1.3355,
+      "step": 15555
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 1.7250741720199585,
+      "learning_rate": 4.361082264317673e-05,
+      "loss": 1.3161,
+      "step": 15560
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 2.3266186714172363,
+      "learning_rate": 4.360689926713464e-05,
+      "loss": 1.3587,
+      "step": 15565
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 1.8503769636154175,
+      "learning_rate": 4.3602974863452835e-05,
+      "loss": 1.3439,
+      "step": 15570
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 2.670262098312378,
+      "learning_rate": 4.3599049432348074e-05,
+      "loss": 1.2923,
+      "step": 15575
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 1.2793365716934204,
+      "learning_rate": 4.359512297403714e-05,
+      "loss": 1.4055,
+      "step": 15580
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 1.302943468093872,
+      "learning_rate": 4.359119548873689e-05,
+      "loss": 1.2158,
+      "step": 15585
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 1.2726792097091675,
+      "learning_rate": 4.358726697666424e-05,
+      "loss": 1.2166,
+      "step": 15590
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 2.570844888687134,
+      "learning_rate": 4.358333743803616e-05,
+      "loss": 1.1484,
+      "step": 15595
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 2.5889105796813965,
+      "learning_rate": 4.3579406873069664e-05,
+      "loss": 1.31,
+      "step": 15600
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 2.399143695831299,
+      "learning_rate": 4.3575475281981844e-05,
+      "loss": 1.2022,
+      "step": 15605
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.9160411357879639,
+      "learning_rate": 4.357154266498983e-05,
+      "loss": 1.2351,
+      "step": 15610
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 2.3706166744232178,
+      "learning_rate": 4.356760902231082e-05,
+      "loss": 1.3131,
+      "step": 15615
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 3.5267927646636963,
+      "learning_rate": 4.356367435416205e-05,
+      "loss": 1.4429,
+      "step": 15620
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 2.2916717529296875,
+      "learning_rate": 4.3559738660760854e-05,
+      "loss": 1.4017,
+      "step": 15625
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 1.031050205230713,
+      "learning_rate": 4.3555801942324584e-05,
+      "loss": 1.4196,
+      "step": 15630
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 1.4319988489151,
+      "learning_rate": 4.355186419907066e-05,
+      "loss": 1.1985,
+      "step": 15635
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 5.716891765594482,
+      "learning_rate": 4.354792543121656e-05,
+      "loss": 1.2971,
+      "step": 15640
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 2.955399751663208,
+      "learning_rate": 4.3543985638979816e-05,
+      "loss": 1.3152,
+      "step": 15645
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 3.042884349822998,
+      "learning_rate": 4.354004482257802e-05,
+      "loss": 1.2771,
+      "step": 15650
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 1.6653623580932617,
+      "learning_rate": 4.353610298222882e-05,
+      "loss": 1.3533,
+      "step": 15655
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 1.1561282873153687,
+      "learning_rate": 4.353216011814992e-05,
+      "loss": 1.2868,
+      "step": 15660
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 1.7067124843597412,
+      "learning_rate": 4.352821623055908e-05,
+      "loss": 1.2304,
+      "step": 15665
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 1.450724720954895,
+      "learning_rate": 4.352427131967412e-05,
+      "loss": 1.1762,
+      "step": 15670
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 2.304478406906128,
+      "learning_rate": 4.3520325385712904e-05,
+      "loss": 1.1929,
+      "step": 15675
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 1.6809667348861694,
+      "learning_rate": 4.3516378428893366e-05,
+      "loss": 1.238,
+      "step": 15680
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 1.2087801694869995,
+      "learning_rate": 4.35124304494335e-05,
+      "loss": 1.2465,
+      "step": 15685
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 1.7735761404037476,
+      "learning_rate": 4.350848144755134e-05,
+      "loss": 1.2547,
+      "step": 15690
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 2.593214988708496,
+      "learning_rate": 4.3504531423464996e-05,
+      "loss": 1.2919,
+      "step": 15695
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 2.0241143703460693,
+      "learning_rate": 4.350058037739261e-05,
+      "loss": 1.3598,
+      "step": 15700
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 1.4365257024765015,
+      "learning_rate": 4.3496628309552395e-05,
+      "loss": 1.2709,
+      "step": 15705
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 1.55208420753479,
+      "learning_rate": 4.349267522016263e-05,
+      "loss": 1.2687,
+      "step": 15710
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 1.5196239948272705,
+      "learning_rate": 4.348872110944163e-05,
+      "loss": 1.3805,
+      "step": 15715
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 2.587491750717163,
+      "learning_rate": 4.348476597760779e-05,
+      "loss": 1.1504,
+      "step": 15720
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 1.250502586364746,
+      "learning_rate": 4.348080982487953e-05,
+      "loss": 1.351,
+      "step": 15725
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 1.627832055091858,
+      "learning_rate": 4.347685265147536e-05,
+      "loss": 1.2934,
+      "step": 15730
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 1.428882360458374,
+      "learning_rate": 4.347289445761382e-05,
+      "loss": 1.3566,
+      "step": 15735
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 1.8175216913223267,
+      "learning_rate": 4.346893524351352e-05,
+      "loss": 1.3705,
+      "step": 15740
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 2.059349536895752,
+      "learning_rate": 4.3464975009393124e-05,
+      "loss": 1.2061,
+      "step": 15745
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 2.046034336090088,
+      "learning_rate": 4.3461013755471354e-05,
+      "loss": 1.0747,
+      "step": 15750
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 1.8597091436386108,
+      "learning_rate": 4.345705148196698e-05,
+      "loss": 1.2378,
+      "step": 15755
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.176646113395691,
+      "learning_rate": 4.345308818909884e-05,
+      "loss": 1.4726,
+      "step": 15760
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 3.254624605178833,
+      "learning_rate": 4.344912387708582e-05,
+      "loss": 1.2737,
+      "step": 15765
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.0571980476379395,
+      "learning_rate": 4.344515854614686e-05,
+      "loss": 1.1995,
+      "step": 15770
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 2.313420057296753,
+      "learning_rate": 4.3441192196500976e-05,
+      "loss": 1.1803,
+      "step": 15775
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.746813416481018,
+      "learning_rate": 4.3437224828367205e-05,
+      "loss": 1.3353,
+      "step": 15780
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.8490636348724365,
+      "learning_rate": 4.343325644196468e-05,
+      "loss": 1.1052,
+      "step": 15785
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.9969300031661987,
+      "learning_rate": 4.342928703751256e-05,
+      "loss": 1.1735,
+      "step": 15790
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 1.4050122499465942,
+      "learning_rate": 4.342531661523007e-05,
+      "loss": 1.1179,
+      "step": 15795
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 1.8106729984283447,
+      "learning_rate": 4.34213451753365e-05,
+      "loss": 1.4247,
+      "step": 15800
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 1.9320409297943115,
+      "learning_rate": 4.341737271805118e-05,
+      "loss": 1.2531,
+      "step": 15805
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 2.287996292114258,
+      "learning_rate": 4.3413399243593513e-05,
+      "loss": 1.3139,
+      "step": 15810
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.8202207088470459,
+      "learning_rate": 4.3409424752182934e-05,
+      "loss": 1.1417,
+      "step": 15815
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 1.7704706192016602,
+      "learning_rate": 4.340544924403898e-05,
+      "loss": 1.307,
+      "step": 15820
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 2.2683053016662598,
+      "learning_rate": 4.340147271938118e-05,
+      "loss": 1.3663,
+      "step": 15825
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 2.761477470397949,
+      "learning_rate": 4.339749517842917e-05,
+      "loss": 1.3196,
+      "step": 15830
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 2.3123843669891357,
+      "learning_rate": 4.339351662140263e-05,
+      "loss": 1.1369,
+      "step": 15835
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 3.907461166381836,
+      "learning_rate": 4.338953704852128e-05,
+      "loss": 1.2576,
+      "step": 15840
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 1.5930169820785522,
+      "learning_rate": 4.3385556460004925e-05,
+      "loss": 1.2911,
+      "step": 15845
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 3.523927688598633,
+      "learning_rate": 4.338157485607339e-05,
+      "loss": 1.1452,
+      "step": 15850
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 1.7907129526138306,
+      "learning_rate": 4.3377592236946585e-05,
+      "loss": 1.3091,
+      "step": 15855
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 1.3438159227371216,
+      "learning_rate": 4.337360860284446e-05,
+      "loss": 1.2201,
+      "step": 15860
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 3.469675064086914,
+      "learning_rate": 4.336962395398703e-05,
+      "loss": 1.147,
+      "step": 15865
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 1.553625464439392,
+      "learning_rate": 4.3365638290594366e-05,
+      "loss": 1.4353,
+      "step": 15870
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 2.0750131607055664,
+      "learning_rate": 4.336165161288659e-05,
+      "loss": 1.2967,
+      "step": 15875
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 2.1773085594177246,
+      "learning_rate": 4.335766392108387e-05,
+      "loss": 1.2372,
+      "step": 15880
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 2.4969215393066406,
+      "learning_rate": 4.3353675215406465e-05,
+      "loss": 1.2196,
+      "step": 15885
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 1.2031995058059692,
+      "learning_rate": 4.334968549607465e-05,
+      "loss": 1.1849,
+      "step": 15890
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 2.3058178424835205,
+      "learning_rate": 4.3345694763308786e-05,
+      "loss": 1.2714,
+      "step": 15895
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 1.7053008079528809,
+      "learning_rate": 4.334170301732926e-05,
+      "loss": 1.2352,
+      "step": 15900
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 2.5460364818573,
+      "learning_rate": 4.333771025835655e-05,
+      "loss": 1.1808,
+      "step": 15905
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 1.2482750415802002,
+      "learning_rate": 4.333371648661116e-05,
+      "loss": 1.2871,
+      "step": 15910
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 2.138806104660034,
+      "learning_rate": 4.3329721702313665e-05,
+      "loss": 1.3245,
+      "step": 15915
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 1.7979081869125366,
+      "learning_rate": 4.332572590568469e-05,
+      "loss": 1.2463,
+      "step": 15920
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.921890914440155,
+      "learning_rate": 4.332172909694493e-05,
+      "loss": 1.34,
+      "step": 15925
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 2.6406264305114746,
+      "learning_rate": 4.331773127631511e-05,
+      "loss": 1.195,
+      "step": 15930
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 2.4825634956359863,
+      "learning_rate": 4.331373244401603e-05,
+      "loss": 1.144,
+      "step": 15935
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 1.4051223993301392,
+      "learning_rate": 4.330973260026855e-05,
+      "loss": 1.3728,
+      "step": 15940
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 1.1359541416168213,
+      "learning_rate": 4.330573174529356e-05,
+      "loss": 1.3258,
+      "step": 15945
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 1.348656177520752,
+      "learning_rate": 4.330172987931205e-05,
+      "loss": 1.251,
+      "step": 15950
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 2.296485185623169,
+      "learning_rate": 4.3297727002545006e-05,
+      "loss": 1.1964,
+      "step": 15955
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 1.3783894777297974,
+      "learning_rate": 4.329372311521353e-05,
+      "loss": 1.261,
+      "step": 15960
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 3.1686336994171143,
+      "learning_rate": 4.328971821753873e-05,
+      "loss": 1.31,
+      "step": 15965
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 2.1778533458709717,
+      "learning_rate": 4.328571230974181e-05,
+      "loss": 1.1521,
+      "step": 15970
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 1.734237790107727,
+      "learning_rate": 4.328170539204401e-05,
+      "loss": 1.392,
+      "step": 15975
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 2.2675347328186035,
+      "learning_rate": 4.327769746466662e-05,
+      "loss": 1.3108,
+      "step": 15980
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 1.0224782228469849,
+      "learning_rate": 4.3273688527831e-05,
+      "loss": 1.2305,
+      "step": 15985
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 2.3099098205566406,
+      "learning_rate": 4.3269678581758556e-05,
+      "loss": 1.1456,
+      "step": 15990
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 3.697842836380005,
+      "learning_rate": 4.326566762667075e-05,
+      "loss": 1.161,
+      "step": 15995
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 1.7592895030975342,
+      "learning_rate": 4.3261655662789115e-05,
+      "loss": 1.3616,
+      "step": 16000
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 2.622274160385132,
+      "learning_rate": 4.325764269033521e-05,
+      "loss": 1.2472,
+      "step": 16005
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 1.370152473449707,
+      "learning_rate": 4.325362870953069e-05,
+      "loss": 1.2645,
+      "step": 16010
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 1.3675168752670288,
+      "learning_rate": 4.324961372059722e-05,
+      "loss": 1.2626,
+      "step": 16015
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 2.024503231048584,
+      "learning_rate": 4.324559772375656e-05,
+      "loss": 1.3285,
+      "step": 16020
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 1.5145565271377563,
+      "learning_rate": 4.32415807192305e-05,
+      "loss": 1.287,
+      "step": 16025
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 3.959343671798706,
+      "learning_rate": 4.3237562707240897e-05,
+      "loss": 1.2395,
+      "step": 16030
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.6277161836624146,
+      "learning_rate": 4.3233543688009657e-05,
+      "loss": 1.1997,
+      "step": 16035
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.912680983543396,
+      "learning_rate": 4.322952366175876e-05,
+      "loss": 1.2068,
+      "step": 16040
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 3.4466264247894287,
+      "learning_rate": 4.322550262871021e-05,
+      "loss": 1.2512,
+      "step": 16045
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.7129842042922974,
+      "learning_rate": 4.3221480589086104e-05,
+      "loss": 1.1405,
+      "step": 16050
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.4409843683242798,
+      "learning_rate": 4.321745754310856e-05,
+      "loss": 1.3464,
+      "step": 16055
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.8235528469085693,
+      "learning_rate": 4.321343349099978e-05,
+      "loss": 1.2919,
+      "step": 16060
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 1.156563639640808,
+      "learning_rate": 4.3210213525047607e-05,
+      "loss": 1.2472,
+      "step": 16065
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 2.0325233936309814,
+      "learning_rate": 4.320618766246267e-05,
+      "loss": 1.3308,
+      "step": 16070
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 1.3709142208099365,
+      "learning_rate": 4.320216079436892e-05,
+      "loss": 1.3726,
+      "step": 16075
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 2.0729141235351562,
+      "learning_rate": 4.3198132920988746e-05,
+      "loss": 1.2768,
+      "step": 16080
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 2.903198480606079,
+      "learning_rate": 4.3194104042544606e-05,
+      "loss": 1.1628,
+      "step": 16085
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 2.467470645904541,
+      "learning_rate": 4.319007415925901e-05,
+      "loss": 1.2066,
+      "step": 16090
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 1.2951477766036987,
+      "learning_rate": 4.3186043271354534e-05,
+      "loss": 1.3537,
+      "step": 16095
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 1.79331374168396,
+      "learning_rate": 4.318201137905379e-05,
+      "loss": 1.1477,
+      "step": 16100
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 2.5787532329559326,
+      "learning_rate": 4.317797848257945e-05,
+      "loss": 1.2367,
+      "step": 16105
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 2.574690580368042,
+      "learning_rate": 4.3173944582154256e-05,
+      "loss": 1.2734,
+      "step": 16110
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 3.0416769981384277,
+      "learning_rate": 4.316990967800099e-05,
+      "loss": 1.2573,
+      "step": 16115
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 1.79559326171875,
+      "learning_rate": 4.31658737703425e-05,
+      "loss": 1.1909,
+      "step": 16120
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 3.641195774078369,
+      "learning_rate": 4.31618368594017e-05,
+      "loss": 1.2127,
+      "step": 16125
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 2.858097553253174,
+      "learning_rate": 4.315779894540151e-05,
+      "loss": 1.1407,
+      "step": 16130
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 2.4541454315185547,
+      "learning_rate": 4.315376002856496e-05,
+      "loss": 1.2637,
+      "step": 16135
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 4.378109455108643,
+      "learning_rate": 4.3149720109115116e-05,
+      "loss": 1.3395,
+      "step": 16140
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 4.454669952392578,
+      "learning_rate": 4.31456791872751e-05,
+      "loss": 1.2609,
+      "step": 16145
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.9385771751403809,
+      "learning_rate": 4.3141637263268074e-05,
+      "loss": 1.4461,
+      "step": 16150
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 4.431612014770508,
+      "learning_rate": 4.313759433731729e-05,
+      "loss": 1.2072,
+      "step": 16155
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 3.2460215091705322,
+      "learning_rate": 4.313355040964602e-05,
+      "loss": 1.3532,
+      "step": 16160
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 1.6554932594299316,
+      "learning_rate": 4.312950548047761e-05,
+      "loss": 1.4051,
+      "step": 16165
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 1.5239330530166626,
+      "learning_rate": 4.3125459550035454e-05,
+      "loss": 1.2129,
+      "step": 16170
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 1.9607256650924683,
+      "learning_rate": 4.3121412618543014e-05,
+      "loss": 1.1648,
+      "step": 16175
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 1.0055090188980103,
+      "learning_rate": 4.311736468622378e-05,
+      "loss": 1.398,
+      "step": 16180
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 1.6140061616897583,
+      "learning_rate": 4.311331575330134e-05,
+      "loss": 1.3406,
+      "step": 16185
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 8.580751419067383,
+      "learning_rate": 4.310926581999929e-05,
+      "loss": 1.285,
+      "step": 16190
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 2.2242555618286133,
+      "learning_rate": 4.3105214886541315e-05,
+      "loss": 1.1491,
+      "step": 16195
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 4.345371246337891,
+      "learning_rate": 4.310116295315114e-05,
+      "loss": 1.1643,
+      "step": 16200
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 3.590310573577881,
+      "learning_rate": 4.309711002005255e-05,
+      "loss": 1.2797,
+      "step": 16205
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 1.9311707019805908,
+      "learning_rate": 4.309305608746939e-05,
+      "loss": 1.287,
+      "step": 16210
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 2.0158779621124268,
+      "learning_rate": 4.308900115562554e-05,
+      "loss": 1.222,
+      "step": 16215
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.8616341352462769,
+      "learning_rate": 4.308494522474496e-05,
+      "loss": 1.2951,
+      "step": 16220
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 1.497802495956421,
+      "learning_rate": 4.308088829505166e-05,
+      "loss": 1.1485,
+      "step": 16225
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 1.7404403686523438,
+      "learning_rate": 4.307683036676968e-05,
+      "loss": 1.2864,
+      "step": 16230
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 3.3228366374969482,
+      "learning_rate": 4.3072771440123164e-05,
+      "loss": 1.2396,
+      "step": 16235
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 1.6669496297836304,
+      "learning_rate": 4.3068711515336254e-05,
+      "loss": 1.1937,
+      "step": 16240
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 2.0878639221191406,
+      "learning_rate": 4.3064650592633206e-05,
+      "loss": 1.2871,
+      "step": 16245
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 1.9474278688430786,
+      "learning_rate": 4.3060588672238266e-05,
+      "loss": 1.264,
+      "step": 16250
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 1.4628163576126099,
+      "learning_rate": 4.30565257543758e-05,
+      "loss": 1.2895,
+      "step": 16255
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 2.5341408252716064,
+      "learning_rate": 4.3052461839270176e-05,
+      "loss": 1.2429,
+      "step": 16260
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 2.7245945930480957,
+      "learning_rate": 4.3048396927145854e-05,
+      "loss": 1.3481,
+      "step": 16265
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 1.005662202835083,
+      "learning_rate": 4.3044331018227324e-05,
+      "loss": 1.1808,
+      "step": 16270
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.959071695804596,
+      "learning_rate": 4.304026411273915e-05,
+      "loss": 1.2639,
+      "step": 16275
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 1.850095272064209,
+      "learning_rate": 4.303619621090594e-05,
+      "loss": 1.319,
+      "step": 16280
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 2.577181816101074,
+      "learning_rate": 4.3032127312952367e-05,
+      "loss": 1.2413,
+      "step": 16285
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 2.0503063201904297,
+      "learning_rate": 4.302805741910314e-05,
+      "loss": 1.1287,
+      "step": 16290
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 2.769632339477539,
+      "learning_rate": 4.3023986529583046e-05,
+      "loss": 1.2283,
+      "step": 16295
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 2.0074775218963623,
+      "learning_rate": 4.3019914644616904e-05,
+      "loss": 1.1908,
+      "step": 16300
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 1.1852775812149048,
+      "learning_rate": 4.301584176442961e-05,
+      "loss": 1.314,
+      "step": 16305
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 2.0226988792419434,
+      "learning_rate": 4.3011767889246105e-05,
+      "loss": 1.2051,
+      "step": 16310
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 2.0487165451049805,
+      "learning_rate": 4.300769301929138e-05,
+      "loss": 1.1422,
+      "step": 16315
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 2.2297685146331787,
+      "learning_rate": 4.300361715479049e-05,
+      "loss": 1.3844,
+      "step": 16320
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 1.5562113523483276,
+      "learning_rate": 4.2999540295968535e-05,
+      "loss": 1.2164,
+      "step": 16325
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 1.00557541847229,
+      "learning_rate": 4.2995462443050674e-05,
+      "loss": 1.1486,
+      "step": 16330
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 1.7114874124526978,
+      "learning_rate": 4.299138359626213e-05,
+      "loss": 1.3049,
+      "step": 16335
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 1.610742449760437,
+      "learning_rate": 4.2987303755828176e-05,
+      "loss": 1.1456,
+      "step": 16340
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 2.69520902633667,
+      "learning_rate": 4.298322292197413e-05,
+      "loss": 1.2903,
+      "step": 16345
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 1.4834461212158203,
+      "learning_rate": 4.2979141094925365e-05,
+      "loss": 1.1353,
+      "step": 16350
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 2.4924049377441406,
+      "learning_rate": 4.297505827490734e-05,
+      "loss": 1.2812,
+      "step": 16355
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 1.1296648979187012,
+      "learning_rate": 4.297097446214553e-05,
+      "loss": 1.0921,
+      "step": 16360
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.9421242475509644,
+      "learning_rate": 4.296688965686547e-05,
+      "loss": 1.1506,
+      "step": 16365
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 1.1938999891281128,
+      "learning_rate": 4.2962803859292776e-05,
+      "loss": 1.243,
+      "step": 16370
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 2.3676552772521973,
+      "learning_rate": 4.29587170696531e-05,
+      "loss": 1.2123,
+      "step": 16375
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 1.4618945121765137,
+      "learning_rate": 4.295462928817214e-05,
+      "loss": 1.4101,
+      "step": 16380
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 1.811956763267517,
+      "learning_rate": 4.2950540515075664e-05,
+      "loss": 1.2196,
+      "step": 16385
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 2.578814744949341,
+      "learning_rate": 4.294645075058951e-05,
+      "loss": 1.3067,
+      "step": 16390
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 2.0412240028381348,
+      "learning_rate": 4.294235999493952e-05,
+      "loss": 1.123,
+      "step": 16395
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 1.7934796810150146,
+      "learning_rate": 4.293826824835164e-05,
+      "loss": 1.1079,
+      "step": 16400
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 1.1003563404083252,
+      "learning_rate": 4.293417551105186e-05,
+      "loss": 1.2645,
+      "step": 16405
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 2.166717767715454,
+      "learning_rate": 4.29300817832662e-05,
+      "loss": 1.1873,
+      "step": 16410
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 2.064626455307007,
+      "learning_rate": 4.292598706522075e-05,
+      "loss": 1.3719,
+      "step": 16415
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 2.102949857711792,
+      "learning_rate": 4.2921891357141686e-05,
+      "loss": 1.3074,
+      "step": 16420
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 2.0334112644195557,
+      "learning_rate": 4.2917794659255183e-05,
+      "loss": 1.3698,
+      "step": 16425
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 1.0431818962097168,
+      "learning_rate": 4.29136969717875e-05,
+      "loss": 1.2163,
+      "step": 16430
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 1.496396541595459,
+      "learning_rate": 4.290959829496497e-05,
+      "loss": 1.3081,
+      "step": 16435
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 1.6219655275344849,
+      "learning_rate": 4.290549862901393e-05,
+      "loss": 1.2083,
+      "step": 16440
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 1.4473518133163452,
+      "learning_rate": 4.290139797416081e-05,
+      "loss": 1.1601,
+      "step": 16445
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 1.5219734907150269,
+      "learning_rate": 4.28972963306321e-05,
+      "loss": 1.0509,
+      "step": 16450
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 3.6221630573272705,
+      "learning_rate": 4.28931936986543e-05,
+      "loss": 1.2632,
+      "step": 16455
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 1.292728066444397,
+      "learning_rate": 4.2889090078454016e-05,
+      "loss": 1.3043,
+      "step": 16460
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 2.52746844291687,
+      "learning_rate": 4.2884985470257885e-05,
+      "loss": 1.2606,
+      "step": 16465
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 2.0615930557250977,
+      "learning_rate": 4.288087987429261e-05,
+      "loss": 1.4306,
+      "step": 16470
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 2.1873435974121094,
+      "learning_rate": 4.287677329078491e-05,
+      "loss": 1.3791,
+      "step": 16475
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.9649629592895508,
+      "learning_rate": 4.2872665719961605e-05,
+      "loss": 1.2359,
+      "step": 16480
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 2.3767576217651367,
+      "learning_rate": 4.2868557162049546e-05,
+      "loss": 1.3599,
+      "step": 16485
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 4.670454025268555,
+      "learning_rate": 4.286444761727566e-05,
+      "loss": 1.1474,
+      "step": 16490
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 3.1629886627197266,
+      "learning_rate": 4.286033708586689e-05,
+      "loss": 1.2123,
+      "step": 16495
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 1.4645919799804688,
+      "learning_rate": 4.2856225568050266e-05,
+      "loss": 1.4218,
+      "step": 16500
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 1.6503212451934814,
+      "learning_rate": 4.2852113064052874e-05,
+      "loss": 1.2461,
+      "step": 16505
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 1.6822096109390259,
+      "learning_rate": 4.2847999574101826e-05,
+      "loss": 1.127,
+      "step": 16510
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 1.8511792421340942,
+      "learning_rate": 4.284388509842432e-05,
+      "loss": 1.0953,
+      "step": 16515
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 3.1232521533966064,
+      "learning_rate": 4.2839769637247584e-05,
+      "loss": 1.1931,
+      "step": 16520
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 1.0340989828109741,
+      "learning_rate": 4.283565319079892e-05,
+      "loss": 1.2855,
+      "step": 16525
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 1.5845731496810913,
+      "learning_rate": 4.2831535759305664e-05,
+      "loss": 1.2968,
+      "step": 16530
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 1.5348531007766724,
+      "learning_rate": 4.282741734299523e-05,
+      "loss": 1.3528,
+      "step": 16535
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 1.6758315563201904,
+      "learning_rate": 4.2823297942095054e-05,
+      "loss": 1.3494,
+      "step": 16540
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 2.661583185195923,
+      "learning_rate": 4.2819177556832665e-05,
+      "loss": 1.3549,
+      "step": 16545
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 1.3939776420593262,
+      "learning_rate": 4.281505618743562e-05,
+      "loss": 1.25,
+      "step": 16550
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 4.4011616706848145,
+      "learning_rate": 4.281093383413154e-05,
+      "loss": 1.1899,
+      "step": 16555
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 2.488006830215454,
+      "learning_rate": 4.28068104971481e-05,
+      "loss": 1.153,
+      "step": 16560
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 2.225281000137329,
+      "learning_rate": 4.2802686176713026e-05,
+      "loss": 1.2394,
+      "step": 16565
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.9629842042922974,
+      "learning_rate": 4.27985608730541e-05,
+      "loss": 1.1729,
+      "step": 16570
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 2.1733222007751465,
+      "learning_rate": 4.279443458639916e-05,
+      "loss": 1.0447,
+      "step": 16575
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 2.7707877159118652,
+      "learning_rate": 4.279030731697609e-05,
+      "loss": 1.3554,
+      "step": 16580
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 1.399963617324829,
+      "learning_rate": 4.278617906501283e-05,
+      "loss": 1.3338,
+      "step": 16585
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 1.5772522687911987,
+      "learning_rate": 4.2782049830737394e-05,
+      "loss": 1.4114,
+      "step": 16590
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 2.678196907043457,
+      "learning_rate": 4.277791961437784e-05,
+      "loss": 1.2212,
+      "step": 16595
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.8356932997703552,
+      "learning_rate": 4.277378841616225e-05,
+      "loss": 1.1251,
+      "step": 16600
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 2.3631153106689453,
+      "learning_rate": 4.276965623631881e-05,
+      "loss": 1.0944,
+      "step": 16605
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 1.5095417499542236,
+      "learning_rate": 4.276552307507572e-05,
+      "loss": 1.2006,
+      "step": 16610
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 1.5276120901107788,
+      "learning_rate": 4.2761388932661264e-05,
+      "loss": 1.2489,
+      "step": 16615
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 2.6862080097198486,
+      "learning_rate": 4.275725380930375e-05,
+      "loss": 1.2177,
+      "step": 16620
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 3.0933616161346436,
+      "learning_rate": 4.275311770523157e-05,
+      "loss": 1.3473,
+      "step": 16625
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 2.2336065769195557,
+      "learning_rate": 4.2748980620673155e-05,
+      "loss": 1.4323,
+      "step": 16630
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 1.266717791557312,
+      "learning_rate": 4.274484255585699e-05,
+      "loss": 1.2058,
+      "step": 16635
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 3.835813522338867,
+      "learning_rate": 4.274070351101161e-05,
+      "loss": 1.2881,
+      "step": 16640
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 1.388725757598877,
+      "learning_rate": 4.273656348636562e-05,
+      "loss": 1.1769,
+      "step": 16645
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 1.5018970966339111,
+      "learning_rate": 4.273242248214767e-05,
+      "loss": 1.2342,
+      "step": 16650
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 1.3790775537490845,
+      "learning_rate": 4.272828049858645e-05,
+      "loss": 1.2862,
+      "step": 16655
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 1.3459662199020386,
+      "learning_rate": 4.2724137535910727e-05,
+      "loss": 1.2824,
+      "step": 16660
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 2.3956856727600098,
+      "learning_rate": 4.2719993594349316e-05,
+      "loss": 1.2141,
+      "step": 16665
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 4.897758960723877,
+      "learning_rate": 4.271584867413107e-05,
+      "loss": 1.2121,
+      "step": 16670
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 1.78498113155365,
+      "learning_rate": 4.271170277548492e-05,
+      "loss": 1.4079,
+      "step": 16675
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 2.1399378776550293,
+      "learning_rate": 4.270755589863983e-05,
+      "loss": 1.205,
+      "step": 16680
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 2.0815482139587402,
+      "learning_rate": 4.2703408043824845e-05,
+      "loss": 1.2547,
+      "step": 16685
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 2.381340980529785,
+      "learning_rate": 4.2699259211269025e-05,
+      "loss": 1.3422,
+      "step": 16690
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 2.151851177215576,
+      "learning_rate": 4.269510940120152e-05,
+      "loss": 1.2764,
+      "step": 16695
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.3695425987243652,
+      "learning_rate": 4.269095861385152e-05,
+      "loss": 1.2128,
+      "step": 16700
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 2.25834321975708,
+      "learning_rate": 4.268680684944826e-05,
+      "loss": 1.5246,
+      "step": 16705
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.995069980621338,
+      "learning_rate": 4.268265410822104e-05,
+      "loss": 1.3022,
+      "step": 16710
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.8739442825317383,
+      "learning_rate": 4.267850039039922e-05,
+      "loss": 1.1678,
+      "step": 16715
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.2266640663146973,
+      "learning_rate": 4.26743456962122e-05,
+      "loss": 1.1569,
+      "step": 16720
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 2.092092990875244,
+      "learning_rate": 4.2670190025889436e-05,
+      "loss": 1.3194,
+      "step": 16725
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 2.1528186798095703,
+      "learning_rate": 4.266603337966044e-05,
+      "loss": 1.2965,
+      "step": 16730
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 1.078891396522522,
+      "learning_rate": 4.266187575775479e-05,
+      "loss": 1.1757,
+      "step": 16735
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 2.155930995941162,
+      "learning_rate": 4.265771716040209e-05,
+      "loss": 1.1875,
+      "step": 16740
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 1.9117164611816406,
+      "learning_rate": 4.265355758783203e-05,
+      "loss": 1.1627,
+      "step": 16745
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 2.450892210006714,
+      "learning_rate": 4.264939704027434e-05,
+      "loss": 1.3303,
+      "step": 16750
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 1.585326910018921,
+      "learning_rate": 4.2645235517958796e-05,
+      "loss": 0.9849,
+      "step": 16755
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 1.436105728149414,
+      "learning_rate": 4.264107302111523e-05,
+      "loss": 1.2115,
+      "step": 16760
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 3.469156503677368,
+      "learning_rate": 4.2636909549973535e-05,
+      "loss": 1.2543,
+      "step": 16765
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 1.467289924621582,
+      "learning_rate": 4.263274510476366e-05,
+      "loss": 1.2511,
+      "step": 16770
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 1.1534559726715088,
+      "learning_rate": 4.262857968571561e-05,
+      "loss": 1.1398,
+      "step": 16775
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 2.7166788578033447,
+      "learning_rate": 4.262441329305942e-05,
+      "loss": 1.213,
+      "step": 16780
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 3.0506703853607178,
+      "learning_rate": 4.262024592702521e-05,
+      "loss": 1.2207,
+      "step": 16785
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 2.2028605937957764,
+      "learning_rate": 4.2616077587843126e-05,
+      "loss": 1.2165,
+      "step": 16790
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 1.4191491603851318,
+      "learning_rate": 4.26119082757434e-05,
+      "loss": 1.2871,
+      "step": 16795
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 0.948824942111969,
+      "learning_rate": 4.2607737990956276e-05,
+      "loss": 1.1885,
+      "step": 16800
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 1.8765445947647095,
+      "learning_rate": 4.260356673371208e-05,
+      "loss": 1.228,
+      "step": 16805
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 1.0872840881347656,
+      "learning_rate": 4.2599394504241195e-05,
+      "loss": 1.2076,
+      "step": 16810
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 1.3634966611862183,
+      "learning_rate": 4.259522130277405e-05,
+      "loss": 1.2841,
+      "step": 16815
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 2.5034995079040527,
+      "learning_rate": 4.259104712954112e-05,
+      "loss": 1.2005,
+      "step": 16820
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 1.3833112716674805,
+      "learning_rate": 4.2586871984772943e-05,
+      "loss": 1.0861,
+      "step": 16825
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 2.4615163803100586,
+      "learning_rate": 4.25826958687001e-05,
+      "loss": 1.2638,
+      "step": 16830
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 3.113060474395752,
+      "learning_rate": 4.257851878155324e-05,
+      "loss": 1.3006,
+      "step": 16835
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 2.4361579418182373,
+      "learning_rate": 4.257434072356307e-05,
+      "loss": 1.2466,
+      "step": 16840
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 2.1952314376831055,
+      "learning_rate": 4.257016169496032e-05,
+      "loss": 1.2439,
+      "step": 16845
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 1.1961103677749634,
+      "learning_rate": 4.256598169597581e-05,
+      "loss": 1.306,
+      "step": 16850
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 1.3207539319992065,
+      "learning_rate": 4.2561800726840385e-05,
+      "loss": 1.2848,
+      "step": 16855
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 1.850471019744873,
+      "learning_rate": 4.255761878778496e-05,
+      "loss": 1.3269,
+      "step": 16860
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 1.6024876832962036,
+      "learning_rate": 4.255343587904051e-05,
+      "loss": 1.2962,
+      "step": 16865
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 2.1519381999969482,
+      "learning_rate": 4.2549252000838025e-05,
+      "loss": 1.3354,
+      "step": 16870
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 2.02545166015625,
+      "learning_rate": 4.25450671534086e-05,
+      "loss": 1.2417,
+      "step": 16875
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 1.6265103816986084,
+      "learning_rate": 4.2540881336983357e-05,
+      "loss": 1.2181,
+      "step": 16880
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 2.541177988052368,
+      "learning_rate": 4.253669455179347e-05,
+      "loss": 1.2521,
+      "step": 16885
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 1.0835845470428467,
+      "learning_rate": 4.253250679807017e-05,
+      "loss": 1.138,
+      "step": 16890
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 1.4941011667251587,
+      "learning_rate": 4.252831807604475e-05,
+      "loss": 1.2487,
+      "step": 16895
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 1.4814062118530273,
+      "learning_rate": 4.252412838594853e-05,
+      "loss": 1.217,
+      "step": 16900
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 1.4170145988464355,
+      "learning_rate": 4.251993772801292e-05,
+      "loss": 1.2874,
+      "step": 16905
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 1.7767305374145508,
+      "learning_rate": 4.251574610246935e-05,
+      "loss": 1.2459,
+      "step": 16910
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 1.7504329681396484,
+      "learning_rate": 4.251155350954935e-05,
+      "loss": 1.1763,
+      "step": 16915
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 2.1863224506378174,
+      "learning_rate": 4.250735994948444e-05,
+      "loss": 1.1726,
+      "step": 16920
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 2.441636085510254,
+      "learning_rate": 4.250316542250624e-05,
+      "loss": 1.2343,
+      "step": 16925
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 2.9591567516326904,
+      "learning_rate": 4.249896992884641e-05,
+      "loss": 1.2953,
+      "step": 16930
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 1.866086721420288,
+      "learning_rate": 4.2494773468736654e-05,
+      "loss": 1.1143,
+      "step": 16935
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 2.113767147064209,
+      "learning_rate": 4.249057604240875e-05,
+      "loss": 1.2088,
+      "step": 16940
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 1.1835196018218994,
+      "learning_rate": 4.2486377650094505e-05,
+      "loss": 1.1204,
+      "step": 16945
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 2.831669807434082,
+      "learning_rate": 4.24821782920258e-05,
+      "loss": 1.0658,
+      "step": 16950
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 2.2646048069000244,
+      "learning_rate": 4.247797796843456e-05,
+      "loss": 1.3129,
+      "step": 16955
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 1.5404542684555054,
+      "learning_rate": 4.247377667955277e-05,
+      "loss": 1.1941,
+      "step": 16960
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 1.8242815732955933,
+      "learning_rate": 4.246957442561245e-05,
+      "loss": 1.194,
+      "step": 16965
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 1.6594616174697876,
+      "learning_rate": 4.246537120684569e-05,
+      "loss": 1.217,
+      "step": 16970
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 2.1032214164733887,
+      "learning_rate": 4.2462007937313256e-05,
+      "loss": 1.1512,
+      "step": 16975
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 1.7993156909942627,
+      "learning_rate": 4.245780298244394e-05,
+      "loss": 1.3622,
+      "step": 16980
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 2.3220622539520264,
+      "learning_rate": 4.245359706339832e-05,
+      "loss": 1.1386,
+      "step": 16985
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 3.0997767448425293,
+      "learning_rate": 4.2449390180408666e-05,
+      "loss": 1.0817,
+      "step": 16990
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 1.8498963117599487,
+      "learning_rate": 4.2445182333707334e-05,
+      "loss": 1.1428,
+      "step": 16995
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 2.013101816177368,
+      "learning_rate": 4.244097352352672e-05,
+      "loss": 1.0904,
+      "step": 17000
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 1.1816306114196777,
+      "learning_rate": 4.243676375009926e-05,
+      "loss": 1.1456,
+      "step": 17005
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 2.447376012802124,
+      "learning_rate": 4.243255301365746e-05,
+      "loss": 1.2639,
+      "step": 17010
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 7.374262809753418,
+      "learning_rate": 4.2428341314433884e-05,
+      "loss": 1.3451,
+      "step": 17015
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 1.150455117225647,
+      "learning_rate": 4.2424128652661135e-05,
+      "loss": 1.4079,
+      "step": 17020
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 3.0257742404937744,
+      "learning_rate": 4.241991502857187e-05,
+      "loss": 1.0875,
+      "step": 17025
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 2.3345866203308105,
+      "learning_rate": 4.24157004423988e-05,
+      "loss": 1.0676,
+      "step": 17030
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 2.740541696548462,
+      "learning_rate": 4.24114848943747e-05,
+      "loss": 1.2396,
+      "step": 17035
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 1.1719865798950195,
+      "learning_rate": 4.240726838473239e-05,
+      "loss": 1.2915,
+      "step": 17040
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 2.3873939514160156,
+      "learning_rate": 4.240305091370473e-05,
+      "loss": 1.0754,
+      "step": 17045
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 2.359999179840088,
+      "learning_rate": 4.239883248152467e-05,
+      "loss": 1.0515,
+      "step": 17050
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 2.249455451965332,
+      "learning_rate": 4.2394613088425176e-05,
+      "loss": 1.1345,
+      "step": 17055
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 3.461733818054199,
+      "learning_rate": 4.239039273463927e-05,
+      "loss": 1.3695,
+      "step": 17060
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 1.5373154878616333,
+      "learning_rate": 4.238617142040007e-05,
+      "loss": 1.3273,
+      "step": 17065
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 6.110920429229736,
+      "learning_rate": 4.238194914594068e-05,
+      "loss": 1.3379,
+      "step": 17070
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 1.5171079635620117,
+      "learning_rate": 4.237772591149431e-05,
+      "loss": 1.2593,
+      "step": 17075
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 1.9108515977859497,
+      "learning_rate": 4.23735017172942e-05,
+      "loss": 1.4072,
+      "step": 17080
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 3.4119062423706055,
+      "learning_rate": 4.2369276563573645e-05,
+      "loss": 1.3117,
+      "step": 17085
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 1.446298599243164,
+      "learning_rate": 4.2365050450566e-05,
+      "loss": 1.2207,
+      "step": 17090
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 3.0925724506378174,
+      "learning_rate": 4.2360823378504674e-05,
+      "loss": 1.1398,
+      "step": 17095
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 3.060494899749756,
+      "learning_rate": 4.235659534762312e-05,
+      "loss": 1.2885,
+      "step": 17100
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 2.6073570251464844,
+      "learning_rate": 4.235236635815484e-05,
+      "loss": 1.2522,
+      "step": 17105
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.9218636751174927,
+      "learning_rate": 4.23481364103334e-05,
+      "loss": 1.253,
+      "step": 17110
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.8496900200843811,
+      "learning_rate": 4.234390550439243e-05,
+      "loss": 1.3157,
+      "step": 17115
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 2.110830783843994,
+      "learning_rate": 4.233967364056558e-05,
+      "loss": 1.2034,
+      "step": 17120
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 1.333644986152649,
+      "learning_rate": 4.233544081908658e-05,
+      "loss": 1.0453,
+      "step": 17125
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 3.8100926876068115,
+      "learning_rate": 4.233120704018921e-05,
+      "loss": 1.1498,
+      "step": 17130
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 1.855100154876709,
+      "learning_rate": 4.2326972304107284e-05,
+      "loss": 1.1742,
+      "step": 17135
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 1.8866065740585327,
+      "learning_rate": 4.232273661107468e-05,
+      "loss": 1.2759,
+      "step": 17140
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 1.6143709421157837,
+      "learning_rate": 4.231849996132535e-05,
+      "loss": 1.1766,
+      "step": 17145
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 1.1877189874649048,
+      "learning_rate": 4.231426235509326e-05,
+      "loss": 1.3377,
+      "step": 17150
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 3.203878879547119,
+      "learning_rate": 4.2310023792612466e-05,
+      "loss": 1.2728,
+      "step": 17155
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 2.035184860229492,
+      "learning_rate": 4.230578427411705e-05,
+      "loss": 1.2227,
+      "step": 17160
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 2.1178765296936035,
+      "learning_rate": 4.230154379984115e-05,
+      "loss": 1.0745,
+      "step": 17165
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 1.0364670753479004,
+      "learning_rate": 4.229730237001897e-05,
+      "loss": 1.1975,
+      "step": 17170
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 1.6296310424804688,
+      "learning_rate": 4.2293059984884765e-05,
+      "loss": 1.301,
+      "step": 17175
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 2.1096887588500977,
+      "learning_rate": 4.228881664467282e-05,
+      "loss": 1.2428,
+      "step": 17180
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 3.0210041999816895,
+      "learning_rate": 4.228457234961752e-05,
+      "loss": 1.2328,
+      "step": 17185
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 1.9511592388153076,
+      "learning_rate": 4.2280327099953245e-05,
+      "loss": 1.084,
+      "step": 17190
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 2.9204232692718506,
+      "learning_rate": 4.227608089591447e-05,
+      "loss": 1.2676,
+      "step": 17195
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 1.7769725322723389,
+      "learning_rate": 4.22718337377357e-05,
+      "loss": 1.2958,
+      "step": 17200
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 1.5279228687286377,
+      "learning_rate": 4.2267585625651516e-05,
+      "loss": 1.2537,
+      "step": 17205
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 2.0030903816223145,
+      "learning_rate": 4.2263336559896514e-05,
+      "loss": 1.1446,
+      "step": 17210
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 1.8052384853363037,
+      "learning_rate": 4.2259086540705384e-05,
+      "loss": 1.3123,
+      "step": 17215
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 1.5273395776748657,
+      "learning_rate": 4.225483556831284e-05,
+      "loss": 1.3197,
+      "step": 17220
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 7.197212219238281,
+      "learning_rate": 4.225058364295367e-05,
+      "loss": 1.2388,
+      "step": 17225
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 2.869080066680908,
+      "learning_rate": 4.22463307648627e-05,
+      "loss": 1.1988,
+      "step": 17230
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 1.3211283683776855,
+      "learning_rate": 4.22420769342748e-05,
+      "loss": 1.3071,
+      "step": 17235
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 0.8859264850616455,
+      "learning_rate": 4.223782215142491e-05,
+      "loss": 1.2127,
+      "step": 17240
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 2.70227313041687,
+      "learning_rate": 4.2233566416548024e-05,
+      "loss": 1.2588,
+      "step": 17245
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 1.2390819787979126,
+      "learning_rate": 4.2229309729879174e-05,
+      "loss": 1.3169,
+      "step": 17250
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 1.9545568227767944,
+      "learning_rate": 4.222505209165346e-05,
+      "loss": 1.1279,
+      "step": 17255
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 2.1074938774108887,
+      "learning_rate": 4.2220793502106035e-05,
+      "loss": 1.093,
+      "step": 17260
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 2.4729721546173096,
+      "learning_rate": 4.2216533961472074e-05,
+      "loss": 1.2353,
+      "step": 17265
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 2.2002952098846436,
+      "learning_rate": 4.221227346998684e-05,
+      "loss": 1.3197,
+      "step": 17270
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 2.468754768371582,
+      "learning_rate": 4.220801202788563e-05,
+      "loss": 1.2085,
+      "step": 17275
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 1.546784520149231,
+      "learning_rate": 4.22037496354038e-05,
+      "loss": 1.0504,
+      "step": 17280
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 2.2986133098602295,
+      "learning_rate": 4.219948629277677e-05,
+      "loss": 1.0963,
+      "step": 17285
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 1.0062131881713867,
+      "learning_rate": 4.219522200023999e-05,
+      "loss": 1.1797,
+      "step": 17290
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 1.0225803852081299,
+      "learning_rate": 4.219095675802896e-05,
+      "loss": 1.1858,
+      "step": 17295
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 1.347981572151184,
+      "learning_rate": 4.218669056637926e-05,
+      "loss": 1.2715,
+      "step": 17300
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 2.4075846672058105,
+      "learning_rate": 4.218242342552651e-05,
+      "loss": 1.0566,
+      "step": 17305
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 4.251238822937012,
+      "learning_rate": 4.217815533570637e-05,
+      "loss": 1.1106,
+      "step": 17310
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 2.3025827407836914,
+      "learning_rate": 4.2173886297154575e-05,
+      "loss": 1.2947,
+      "step": 17315
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 2.7054014205932617,
+      "learning_rate": 4.216961631010689e-05,
+      "loss": 1.1512,
+      "step": 17320
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 2.6047892570495605,
+      "learning_rate": 4.2165345374799134e-05,
+      "loss": 1.1665,
+      "step": 17325
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 2.1984410285949707,
+      "learning_rate": 4.2161073491467196e-05,
+      "loss": 1.2841,
+      "step": 17330
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 1.6015806198120117,
+      "learning_rate": 4.215680066034701e-05,
+      "loss": 1.3405,
+      "step": 17335
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 5.569428443908691,
+      "learning_rate": 4.215252688167456e-05,
+      "loss": 1.0902,
+      "step": 17340
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 2.5024139881134033,
+      "learning_rate": 4.2148252155685875e-05,
+      "loss": 1.174,
+      "step": 17345
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 1.9789233207702637,
+      "learning_rate": 4.2143976482617055e-05,
+      "loss": 1.2465,
+      "step": 17350
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 2.1717369556427,
+      "learning_rate": 4.213969986270423e-05,
+      "loss": 1.1812,
+      "step": 17355
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 3.2507376670837402,
+      "learning_rate": 4.21354222961836e-05,
+      "loss": 1.2907,
+      "step": 17360
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 1.7060433626174927,
+      "learning_rate": 4.21311437832914e-05,
+      "loss": 1.2651,
+      "step": 17365
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 5.012210369110107,
+      "learning_rate": 4.212686432426394e-05,
+      "loss": 1.1516,
+      "step": 17370
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 5.409265995025635,
+      "learning_rate": 4.2122583919337566e-05,
+      "loss": 1.1225,
+      "step": 17375
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 2.1262056827545166,
+      "learning_rate": 4.211830256874868e-05,
+      "loss": 1.2868,
+      "step": 17380
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 1.5993378162384033,
+      "learning_rate": 4.211402027273373e-05,
+      "loss": 1.333,
+      "step": 17385
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 1.085625171661377,
+      "learning_rate": 4.2109737031529245e-05,
+      "loss": 1.2228,
+      "step": 17390
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 1.2686785459518433,
+      "learning_rate": 4.2105452845371754e-05,
+      "loss": 1.3888,
+      "step": 17395
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 1.7684992551803589,
+      "learning_rate": 4.210116771449789e-05,
+      "loss": 1.1147,
+      "step": 17400
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 4.321854114532471,
+      "learning_rate": 4.209688163914431e-05,
+      "loss": 1.2644,
+      "step": 17405
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 1.0918437242507935,
+      "learning_rate": 4.209259461954772e-05,
+      "loss": 0.9742,
+      "step": 17410
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 1.0674161911010742,
+      "learning_rate": 4.20883066559449e-05,
+      "loss": 1.3029,
+      "step": 17415
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 10.988029479980469,
+      "learning_rate": 4.208401774857267e-05,
+      "loss": 1.4338,
+      "step": 17420
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 1.3051105737686157,
+      "learning_rate": 4.2079727897667896e-05,
+      "loss": 1.1686,
+      "step": 17425
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 1.0172728300094604,
+      "learning_rate": 4.2075437103467495e-05,
+      "loss": 1.1133,
+      "step": 17430
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 1.7416491508483887,
+      "learning_rate": 4.207114536620846e-05,
+      "loss": 1.1215,
+      "step": 17435
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 1.3212878704071045,
+      "learning_rate": 4.206685268612781e-05,
+      "loss": 1.1689,
+      "step": 17440
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 3.1066105365753174,
+      "learning_rate": 4.206255906346262e-05,
+      "loss": 1.0547,
+      "step": 17445
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 4.053175449371338,
+      "learning_rate": 4.205826449845005e-05,
+      "loss": 1.2893,
+      "step": 17450
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 1.900181531906128,
+      "learning_rate": 4.205396899132724e-05,
+      "loss": 1.2453,
+      "step": 17455
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 0.7705186009407043,
+      "learning_rate": 4.2049672542331454e-05,
+      "loss": 1.2095,
+      "step": 17460
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 1.5405861139297485,
+      "learning_rate": 4.2045375151699976e-05,
+      "loss": 1.2622,
+      "step": 17465
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 3.1085848808288574,
+      "learning_rate": 4.204107681967015e-05,
+      "loss": 1.1181,
+      "step": 17470
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 1.0741022825241089,
+      "learning_rate": 4.203677754647936e-05,
+      "loss": 1.1253,
+      "step": 17475
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 1.2874560356140137,
+      "learning_rate": 4.2032477332365054e-05,
+      "loss": 1.1252,
+      "step": 17480
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 2.286153554916382,
+      "learning_rate": 4.2028176177564736e-05,
+      "loss": 1.1286,
+      "step": 17485
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 1.706329107284546,
+      "learning_rate": 4.202387408231595e-05,
+      "loss": 1.1123,
+      "step": 17490
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 2.1467063426971436,
+      "learning_rate": 4.2019571046856284e-05,
+      "loss": 1.2309,
+      "step": 17495
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 2.3436472415924072,
+      "learning_rate": 4.2015267071423404e-05,
+      "loss": 1.1822,
+      "step": 17500
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 1.5204914808273315,
+      "learning_rate": 4.2010962156255004e-05,
+      "loss": 1.1763,
+      "step": 17505
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 1.3198552131652832,
+      "learning_rate": 4.200665630158885e-05,
+      "loss": 1.174,
+      "step": 17510
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 1.4208104610443115,
+      "learning_rate": 4.200234950766275e-05,
+      "loss": 1.1142,
+      "step": 17515
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 2.4655168056488037,
+      "learning_rate": 4.199804177471456e-05,
+      "loss": 0.9846,
+      "step": 17520
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 1.046281099319458,
+      "learning_rate": 4.199373310298219e-05,
+      "loss": 1.2307,
+      "step": 17525
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 1.329183578491211,
+      "learning_rate": 4.19894234927036e-05,
+      "loss": 1.1682,
+      "step": 17530
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 1.4482097625732422,
+      "learning_rate": 4.198511294411681e-05,
+      "loss": 1.2648,
+      "step": 17535
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 1.9892488718032837,
+      "learning_rate": 4.1980801457459895e-05,
+      "loss": 1.0417,
+      "step": 17540
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 2.176584243774414,
+      "learning_rate": 4.197648903297096e-05,
+      "loss": 1.2264,
+      "step": 17545
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 1.82011878490448,
+      "learning_rate": 4.197217567088818e-05,
+      "loss": 1.1264,
+      "step": 17550
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 1.9624288082122803,
+      "learning_rate": 4.196786137144979e-05,
+      "loss": 1.2737,
+      "step": 17555
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 1.3855583667755127,
+      "learning_rate": 4.196354613489404e-05,
+      "loss": 1.3331,
+      "step": 17560
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 1.9667317867279053,
+      "learning_rate": 4.195922996145928e-05,
+      "loss": 1.1719,
+      "step": 17565
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 1.6437666416168213,
+      "learning_rate": 4.1954912851383864e-05,
+      "loss": 1.1407,
+      "step": 17570
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 1.1012728214263916,
+      "learning_rate": 4.1950594804906246e-05,
+      "loss": 1.3084,
+      "step": 17575
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 1.9337821006774902,
+      "learning_rate": 4.1946275822264904e-05,
+      "loss": 1.1738,
+      "step": 17580
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 3.0107462406158447,
+      "learning_rate": 4.194195590369835e-05,
+      "loss": 1.2292,
+      "step": 17585
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 1.6357334852218628,
+      "learning_rate": 4.193763504944518e-05,
+      "loss": 1.2465,
+      "step": 17590
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 3.3010284900665283,
+      "learning_rate": 4.193331325974403e-05,
+      "loss": 1.2547,
+      "step": 17595
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 1.5490888357162476,
+      "learning_rate": 4.19289905348336e-05,
+      "loss": 1.2406,
+      "step": 17600
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 1.6628612279891968,
+      "learning_rate": 4.192466687495262e-05,
+      "loss": 1.0981,
+      "step": 17605
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 1.1924035549163818,
+      "learning_rate": 4.192034228033987e-05,
+      "loss": 1.1726,
+      "step": 17610
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 2.363252639770508,
+      "learning_rate": 4.191601675123422e-05,
+      "loss": 1.2418,
+      "step": 17615
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 2.983738899230957,
+      "learning_rate": 4.1911690287874535e-05,
+      "loss": 1.1598,
+      "step": 17620
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 2.698246955871582,
+      "learning_rate": 4.190736289049977e-05,
+      "loss": 1.256,
+      "step": 17625
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 2.4981305599212646,
+      "learning_rate": 4.190303455934894e-05,
+      "loss": 1.2023,
+      "step": 17630
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 1.513262152671814,
+      "learning_rate": 4.189870529466107e-05,
+      "loss": 1.1823,
+      "step": 17635
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 2.1456382274627686,
+      "learning_rate": 4.1894375096675274e-05,
+      "loss": 1.007,
+      "step": 17640
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 2.4811346530914307,
+      "learning_rate": 4.189004396563071e-05,
+      "loss": 1.2597,
+      "step": 17645
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 2.1642322540283203,
+      "learning_rate": 4.1885711901766564e-05,
+      "loss": 1.2648,
+      "step": 17650
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 2.7846784591674805,
+      "learning_rate": 4.188137890532211e-05,
+      "loss": 1.0887,
+      "step": 17655
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 2.358762502670288,
+      "learning_rate": 4.187704497653665e-05,
+      "loss": 1.2641,
+      "step": 17660
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 1.4334055185317993,
+      "learning_rate": 4.1872710115649525e-05,
+      "loss": 1.2008,
+      "step": 17665
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 1.1440168619155884,
+      "learning_rate": 4.1868374322900163e-05,
+      "loss": 1.2595,
+      "step": 17670
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 1.5361130237579346,
+      "learning_rate": 4.186403759852802e-05,
+      "loss": 1.2388,
+      "step": 17675
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 2.4247677326202393,
+      "learning_rate": 4.185969994277262e-05,
+      "loss": 1.0073,
+      "step": 17680
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 1.3010989427566528,
+      "learning_rate": 4.1855361355873506e-05,
+      "loss": 1.3015,
+      "step": 17685
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 2.8107826709747314,
+      "learning_rate": 4.185102183807031e-05,
+      "loss": 1.167,
+      "step": 17690
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 1.6690555810928345,
+      "learning_rate": 4.1846681389602686e-05,
+      "loss": 1.2343,
+      "step": 17695
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 1.7709966897964478,
+      "learning_rate": 4.1842340010710366e-05,
+      "loss": 1.3241,
+      "step": 17700
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 4.506913661956787,
+      "learning_rate": 4.1837997701633115e-05,
+      "loss": 1.3076,
+      "step": 17705
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 2.404618501663208,
+      "learning_rate": 4.183365446261075e-05,
+      "loss": 1.1784,
+      "step": 17710
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 1.8350248336791992,
+      "learning_rate": 4.182931029388315e-05,
+      "loss": 1.2567,
+      "step": 17715
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 1.5399670600891113,
+      "learning_rate": 4.182496519569023e-05,
+      "loss": 1.1838,
+      "step": 17720
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 0.7773048281669617,
+      "learning_rate": 4.1820619168271975e-05,
+      "loss": 1.2516,
+      "step": 17725
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 2.276472330093384,
+      "learning_rate": 4.181627221186841e-05,
+      "loss": 1.37,
+      "step": 17730
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 2.218336582183838,
+      "learning_rate": 4.18119243267196e-05,
+      "loss": 1.3398,
+      "step": 17735
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 1.3375170230865479,
+      "learning_rate": 4.180757551306569e-05,
+      "loss": 1.3,
+      "step": 17740
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 1.2494932413101196,
+      "learning_rate": 4.180322577114686e-05,
+      "loss": 1.2727,
+      "step": 17745
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 1.8490458726882935,
+      "learning_rate": 4.1798875101203336e-05,
+      "loss": 1.3535,
+      "step": 17750
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 3.2048110961914062,
+      "learning_rate": 4.179452350347539e-05,
+      "loss": 1.1024,
+      "step": 17755
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 2.3998045921325684,
+      "learning_rate": 4.179017097820338e-05,
+      "loss": 1.1105,
+      "step": 17760
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 3.1060426235198975,
+      "learning_rate": 4.178581752562767e-05,
+      "loss": 1.223,
+      "step": 17765
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 2.371044874191284,
+      "learning_rate": 4.178146314598872e-05,
+      "loss": 1.2038,
+      "step": 17770
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 1.7323418855667114,
+      "learning_rate": 4.177710783952699e-05,
+      "loss": 1.2121,
+      "step": 17775
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 2.284435272216797,
+      "learning_rate": 4.177275160648304e-05,
+      "loss": 1.3172,
+      "step": 17780
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 3.9668426513671875,
+      "learning_rate": 4.1768394447097446e-05,
+      "loss": 1.2761,
+      "step": 17785
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 0.9461215734481812,
+      "learning_rate": 4.176403636161086e-05,
+      "loss": 1.2843,
+      "step": 17790
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 2.7739803791046143,
+      "learning_rate": 4.1759677350263976e-05,
+      "loss": 1.4001,
+      "step": 17795
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 1.284010648727417,
+      "learning_rate": 4.1755317413297526e-05,
+      "loss": 1.2278,
+      "step": 17800
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 2.6691951751708984,
+      "learning_rate": 4.175095655095232e-05,
+      "loss": 1.1456,
+      "step": 17805
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 2.771559476852417,
+      "learning_rate": 4.174659476346919e-05,
+      "loss": 1.1955,
+      "step": 17810
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 1.4969909191131592,
+      "learning_rate": 4.174223205108904e-05,
+      "loss": 1.2585,
+      "step": 17815
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 1.8227806091308594,
+      "learning_rate": 4.1737868414052817e-05,
+      "loss": 1.3033,
+      "step": 17820
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 1.4669899940490723,
+      "learning_rate": 4.1733503852601516e-05,
+      "loss": 1.2654,
+      "step": 17825
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 3.621481418609619,
+      "learning_rate": 4.172913836697619e-05,
+      "loss": 1.2247,
+      "step": 17830
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 1.8669898509979248,
+      "learning_rate": 4.172477195741794e-05,
+      "loss": 1.1747,
+      "step": 17835
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 1.145015001296997,
+      "learning_rate": 4.1720404624167925e-05,
+      "loss": 1.3372,
+      "step": 17840
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 1.5588964223861694,
+      "learning_rate": 4.171603636746734e-05,
+      "loss": 1.1368,
+      "step": 17845
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 3.020059823989868,
+      "learning_rate": 4.171166718755744e-05,
+      "loss": 1.1614,
+      "step": 17850
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 2.658262252807617,
+      "learning_rate": 4.170729708467953e-05,
+      "loss": 1.1811,
+      "step": 17855
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 1.9605822563171387,
+      "learning_rate": 4.170292605907498e-05,
+      "loss": 1.1067,
+      "step": 17860
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 2.259817600250244,
+      "learning_rate": 4.169855411098517e-05,
+      "loss": 1.3053,
+      "step": 17865
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 1.7595691680908203,
+      "learning_rate": 4.169418124065159e-05,
+      "loss": 1.2888,
+      "step": 17870
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 1.319011926651001,
+      "learning_rate": 4.168980744831572e-05,
+      "loss": 1.0001,
+      "step": 17875
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 1.3110008239746094,
+      "learning_rate": 4.168543273421913e-05,
+      "loss": 1.3487,
+      "step": 17880
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 2.2493362426757812,
+      "learning_rate": 4.168105709860344e-05,
+      "loss": 1.2849,
+      "step": 17885
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 1.7619404792785645,
+      "learning_rate": 4.167668054171031e-05,
+      "loss": 1.237,
+      "step": 17890
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 1.684557318687439,
+      "learning_rate": 4.167230306378144e-05,
+      "loss": 1.2893,
+      "step": 17895
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 3.6283462047576904,
+      "learning_rate": 4.1667924665058605e-05,
+      "loss": 1.3225,
+      "step": 17900
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 1.8916960954666138,
+      "learning_rate": 4.166354534578362e-05,
+      "loss": 1.2069,
+      "step": 17905
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 1.7808947563171387,
+      "learning_rate": 4.165916510619834e-05,
+      "loss": 1.2129,
+      "step": 17910
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 2.4525959491729736,
+      "learning_rate": 4.1654783946544695e-05,
+      "loss": 1.2451,
+      "step": 17915
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 1.7354241609573364,
+      "learning_rate": 4.165040186706464e-05,
+      "loss": 1.2769,
+      "step": 17920
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 2.6670055389404297,
+      "learning_rate": 4.1646018868000194e-05,
+      "loss": 1.3005,
+      "step": 17925
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 2.470824956893921,
+      "learning_rate": 4.164163494959342e-05,
+      "loss": 1.1619,
+      "step": 17930
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 2.5797417163848877,
+      "learning_rate": 4.1637250112086466e-05,
+      "loss": 1.3173,
+      "step": 17935
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 2.280491828918457,
+      "learning_rate": 4.163286435572147e-05,
+      "loss": 1.323,
+      "step": 17940
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 1.8656688928604126,
+      "learning_rate": 4.162847768074067e-05,
+      "loss": 1.3305,
+      "step": 17945
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 1.572505235671997,
+      "learning_rate": 4.162409008738632e-05,
+      "loss": 1.1925,
+      "step": 17950
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 1.0871065855026245,
+      "learning_rate": 4.161970157590077e-05,
+      "loss": 1.2474,
+      "step": 17955
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 1.81941819190979,
+      "learning_rate": 4.161531214652637e-05,
+      "loss": 1.1525,
+      "step": 17960
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 1.6600680351257324,
+      "learning_rate": 4.161092179950555e-05,
+      "loss": 1.1595,
+      "step": 17965
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 1.9169402122497559,
+      "learning_rate": 4.160653053508079e-05,
+      "loss": 1.2836,
+      "step": 17970
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 2.0961391925811768,
+      "learning_rate": 4.16021383534946e-05,
+      "loss": 1.1986,
+      "step": 17975
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 1.4828705787658691,
+      "learning_rate": 4.159774525498957e-05,
+      "loss": 1.215,
+      "step": 17980
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 1.643730640411377,
+      "learning_rate": 4.159335123980833e-05,
+      "loss": 1.1822,
+      "step": 17985
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 2.4399468898773193,
+      "learning_rate": 4.158895630819354e-05,
+      "loss": 1.3621,
+      "step": 17990
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 1.5563558340072632,
+      "learning_rate": 4.158456046038794e-05,
+      "loss": 1.1871,
+      "step": 17995
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 2.194761037826538,
+      "learning_rate": 4.15801636966343e-05,
+      "loss": 1.269,
+      "step": 18000
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 3.4397332668304443,
+      "learning_rate": 4.157576601717546e-05,
+      "loss": 1.2749,
+      "step": 18005
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 1.9305585622787476,
+      "learning_rate": 4.1571367422254296e-05,
+      "loss": 1.2515,
+      "step": 18010
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 2.5230295658111572,
+      "learning_rate": 4.156696791211372e-05,
+      "loss": 1.3452,
+      "step": 18015
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 4.581267833709717,
+      "learning_rate": 4.156256748699673e-05,
+      "loss": 1.0429,
+      "step": 18020
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 2.436643123626709,
+      "learning_rate": 4.155816614714636e-05,
+      "loss": 1.2162,
+      "step": 18025
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 3.484376907348633,
+      "learning_rate": 4.155376389280569e-05,
+      "loss": 1.1535,
+      "step": 18030
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 0.869539201259613,
+      "learning_rate": 4.1549360724217835e-05,
+      "loss": 1.238,
+      "step": 18035
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 2.1661620140075684,
+      "learning_rate": 4.1544956641625996e-05,
+      "loss": 1.3041,
+      "step": 18040
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 2.650646209716797,
+      "learning_rate": 4.15405516452734e-05,
+      "loss": 1.1902,
+      "step": 18045
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 3.9446349143981934,
+      "learning_rate": 4.153614573540332e-05,
+      "loss": 1.337,
+      "step": 18050
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 1.5067466497421265,
+      "learning_rate": 4.153173891225911e-05,
+      "loss": 1.2288,
+      "step": 18055
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 3.6524410247802734,
+      "learning_rate": 4.152733117608413e-05,
+      "loss": 1.2194,
+      "step": 18060
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 1.016907811164856,
+      "learning_rate": 4.1522922527121846e-05,
+      "loss": 1.1394,
+      "step": 18065
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 2.3441882133483887,
+      "learning_rate": 4.151851296561572e-05,
+      "loss": 1.1889,
+      "step": 18070
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 2.799960136413574,
+      "learning_rate": 4.1514102491809286e-05,
+      "loss": 1.1934,
+      "step": 18075
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 3.4592952728271484,
+      "learning_rate": 4.1509691105946145e-05,
+      "loss": 1.2256,
+      "step": 18080
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 1.311344027519226,
+      "learning_rate": 4.150527880826992e-05,
+      "loss": 1.1273,
+      "step": 18085
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 2.329275131225586,
+      "learning_rate": 4.15008655990243e-05,
+      "loss": 1.3712,
+      "step": 18090
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 4.58236026763916,
+      "learning_rate": 4.149645147845303e-05,
+      "loss": 1.141,
+      "step": 18095
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 1.4890109300613403,
+      "learning_rate": 4.149203644679989e-05,
+      "loss": 1.1462,
+      "step": 18100
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 3.4665701389312744,
+      "learning_rate": 4.148762050430872e-05,
+      "loss": 1.2364,
+      "step": 18105
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 2.3981378078460693,
+      "learning_rate": 4.148320365122341e-05,
+      "loss": 1.2632,
+      "step": 18110
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 2.2879130840301514,
+      "learning_rate": 4.147878588778789e-05,
+      "loss": 1.1986,
+      "step": 18115
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 1.7750098705291748,
+      "learning_rate": 4.1474367214246156e-05,
+      "loss": 1.168,
+      "step": 18120
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 1.1014642715454102,
+      "learning_rate": 4.146994763084225e-05,
+      "loss": 1.2932,
+      "step": 18125
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 2.6048409938812256,
+      "learning_rate": 4.1465527137820255e-05,
+      "loss": 1.2311,
+      "step": 18130
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 1.7655155658721924,
+      "learning_rate": 4.146110573542431e-05,
+      "loss": 1.1891,
+      "step": 18135
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 2.196241617202759,
+      "learning_rate": 4.14566834238986e-05,
+      "loss": 1.1351,
+      "step": 18140
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 1.649719476699829,
+      "learning_rate": 4.145226020348737e-05,
+      "loss": 1.2341,
+      "step": 18145
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 1.7216572761535645,
+      "learning_rate": 4.1447836074434916e-05,
+      "loss": 1.2227,
+      "step": 18150
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 2.0533690452575684,
+      "learning_rate": 4.144341103698557e-05,
+      "loss": 1.2416,
+      "step": 18155
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 3.26836895942688,
+      "learning_rate": 4.143898509138373e-05,
+      "loss": 1.1773,
+      "step": 18160
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 1.7304611206054688,
+      "learning_rate": 4.1434558237873824e-05,
+      "loss": 1.3271,
+      "step": 18165
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 1.9267404079437256,
+      "learning_rate": 4.143013047670035e-05,
+      "loss": 1.2153,
+      "step": 18170
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 1.807952642440796,
+      "learning_rate": 4.1425701808107855e-05,
+      "loss": 1.2352,
+      "step": 18175
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 1.5293511152267456,
+      "learning_rate": 4.142127223234091e-05,
+      "loss": 1.2766,
+      "step": 18180
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 2.029411554336548,
+      "learning_rate": 4.1416841749644174e-05,
+      "loss": 1.0328,
+      "step": 18185
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 1.7925702333450317,
+      "learning_rate": 4.1412410360262334e-05,
+      "loss": 1.366,
+      "step": 18190
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.8310097455978394,
+      "learning_rate": 4.140797806444013e-05,
+      "loss": 1.212,
+      "step": 18195
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 3.394538402557373,
+      "learning_rate": 4.140354486242235e-05,
+      "loss": 1.3675,
+      "step": 18200
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 4.078660488128662,
+      "learning_rate": 4.139911075445384e-05,
+      "loss": 1.1619,
+      "step": 18205
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 1.416306495666504,
+      "learning_rate": 4.1394675740779485e-05,
+      "loss": 1.0888,
+      "step": 18210
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 3.147977828979492,
+      "learning_rate": 4.139023982164424e-05,
+      "loss": 1.1083,
+      "step": 18215
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 3.413358688354492,
+      "learning_rate": 4.138580299729308e-05,
+      "loss": 1.1791,
+      "step": 18220
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 1.7212944030761719,
+      "learning_rate": 4.138136526797105e-05,
+      "loss": 1.2017,
+      "step": 18225
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 2.027756690979004,
+      "learning_rate": 4.137692663392325e-05,
+      "loss": 1.1783,
+      "step": 18230
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 0.9645988345146179,
+      "learning_rate": 4.137248709539481e-05,
+      "loss": 1.1465,
+      "step": 18235
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 2.282986879348755,
+      "learning_rate": 4.1368046652630924e-05,
+      "loss": 1.2574,
+      "step": 18240
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 2.9821066856384277,
+      "learning_rate": 4.136360530587684e-05,
+      "loss": 1.2142,
+      "step": 18245
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 2.920485258102417,
+      "learning_rate": 4.135916305537784e-05,
+      "loss": 1.1583,
+      "step": 18250
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 2.7918076515197754,
+      "learning_rate": 4.135471990137927e-05,
+      "loss": 1.1705,
+      "step": 18255
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 1.458203911781311,
+      "learning_rate": 4.135027584412653e-05,
+      "loss": 1.2112,
+      "step": 18260
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 3.7946949005126953,
+      "learning_rate": 4.134583088386504e-05,
+      "loss": 1.2094,
+      "step": 18265
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 2.0881757736206055,
+      "learning_rate": 4.134138502084029e-05,
+      "loss": 1.3534,
+      "step": 18270
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 3.244400978088379,
+      "learning_rate": 4.133693825529785e-05,
+      "loss": 1.3165,
+      "step": 18275
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 1.4288554191589355,
+      "learning_rate": 4.1332490587483286e-05,
+      "loss": 1.2475,
+      "step": 18280
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 3.9546115398406982,
+      "learning_rate": 4.132804201764224e-05,
+      "loss": 1.2295,
+      "step": 18285
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 2.304332971572876,
+      "learning_rate": 4.132448251247545e-05,
+      "loss": 1.3196,
+      "step": 18290
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 2.4342777729034424,
+      "learning_rate": 4.132003231960591e-05,
+      "loss": 1.1165,
+      "step": 18295
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 2.890040397644043,
+      "learning_rate": 4.131558122539796e-05,
+      "loss": 1.1342,
+      "step": 18300
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 1.576117992401123,
+      "learning_rate": 4.131112923009741e-05,
+      "loss": 1.2944,
+      "step": 18305
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 2.7480485439300537,
+      "learning_rate": 4.130667633395015e-05,
+      "loss": 1.3669,
+      "step": 18310
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 2.0813324451446533,
+      "learning_rate": 4.1302222537202104e-05,
+      "loss": 1.1589,
+      "step": 18315
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 5.030991077423096,
+      "learning_rate": 4.129776784009926e-05,
+      "loss": 1.3765,
+      "step": 18320
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 4.91030740737915,
+      "learning_rate": 4.129331224288763e-05,
+      "loss": 1.0896,
+      "step": 18325
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 6.223639965057373,
+      "learning_rate": 4.1288855745813303e-05,
+      "loss": 1.2247,
+      "step": 18330
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 3.3995022773742676,
+      "learning_rate": 4.128439834912241e-05,
+      "loss": 1.4152,
+      "step": 18335
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 1.390557885169983,
+      "learning_rate": 4.127994005306112e-05,
+      "loss": 1.321,
+      "step": 18340
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 1.9957035779953003,
+      "learning_rate": 4.127548085787566e-05,
+      "loss": 1.2822,
+      "step": 18345
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 3.40853214263916,
+      "learning_rate": 4.127102076381231e-05,
+      "loss": 1.159,
+      "step": 18350
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 1.8385875225067139,
+      "learning_rate": 4.12665597711174e-05,
+      "loss": 1.2222,
+      "step": 18355
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 2.742072582244873,
+      "learning_rate": 4.12620978800373e-05,
+      "loss": 1.1475,
+      "step": 18360
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 2.955695629119873,
+      "learning_rate": 4.125763509081844e-05,
+      "loss": 1.2988,
+      "step": 18365
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 2.316948652267456,
+      "learning_rate": 4.125317140370729e-05,
+      "loss": 1.1109,
+      "step": 18370
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 1.7137174606323242,
+      "learning_rate": 4.1248706818950376e-05,
+      "loss": 1.2146,
+      "step": 18375
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 3.3981897830963135,
+      "learning_rate": 4.124424133679428e-05,
+      "loss": 1.291,
+      "step": 18380
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 1.657471776008606,
+      "learning_rate": 4.123977495748561e-05,
+      "loss": 1.1274,
+      "step": 18385
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 1.293222427368164,
+      "learning_rate": 4.123530768127105e-05,
+      "loss": 1.0961,
+      "step": 18390
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 2.9801487922668457,
+      "learning_rate": 4.123083950839733e-05,
+      "loss": 1.1632,
+      "step": 18395
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 2.470231294631958,
+      "learning_rate": 4.122637043911122e-05,
+      "loss": 1.0898,
+      "step": 18400
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 1.0047528743743896,
+      "learning_rate": 4.122190047365952e-05,
+      "loss": 1.174,
+      "step": 18405
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 5.736430644989014,
+      "learning_rate": 4.121742961228913e-05,
+      "loss": 1.1775,
+      "step": 18410
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 1.3789207935333252,
+      "learning_rate": 4.121295785524696e-05,
+      "loss": 1.2466,
+      "step": 18415
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 1.9205785989761353,
+      "learning_rate": 4.120848520277998e-05,
+      "loss": 1.0881,
+      "step": 18420
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 2.249664783477783,
+      "learning_rate": 4.12040116551352e-05,
+      "loss": 1.3297,
+      "step": 18425
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 1.7185654640197754,
+      "learning_rate": 4.1199537212559705e-05,
+      "loss": 1.1238,
+      "step": 18430
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 3.920592784881592,
+      "learning_rate": 4.119506187530061e-05,
+      "loss": 1.1606,
+      "step": 18435
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 1.8103728294372559,
+      "learning_rate": 4.119058564360509e-05,
+      "loss": 1.3541,
+      "step": 18440
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 2.7962160110473633,
+      "learning_rate": 4.1186108517720344e-05,
+      "loss": 1.1774,
+      "step": 18445
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 4.4177350997924805,
+      "learning_rate": 4.1181630497893645e-05,
+      "loss": 1.1272,
+      "step": 18450
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 1.938101887702942,
+      "learning_rate": 4.117715158437232e-05,
+      "loss": 1.1472,
+      "step": 18455
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 1.0241434574127197,
+      "learning_rate": 4.117267177740373e-05,
+      "loss": 1.2322,
+      "step": 18460
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 1.9051687717437744,
+      "learning_rate": 4.116819107723529e-05,
+      "loss": 1.1203,
+      "step": 18465
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 0.9832600951194763,
+      "learning_rate": 4.1163709484114456e-05,
+      "loss": 1.1336,
+      "step": 18470
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 1.4782296419143677,
+      "learning_rate": 4.1159226998288754e-05,
+      "loss": 1.3825,
+      "step": 18475
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 2.1868507862091064,
+      "learning_rate": 4.1154743620005734e-05,
+      "loss": 1.3936,
+      "step": 18480
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 1.0652060508728027,
+      "learning_rate": 4.115025934951302e-05,
+      "loss": 1.0603,
+      "step": 18485
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 1.29878568649292,
+      "learning_rate": 4.1145774187058265e-05,
+      "loss": 1.1407,
+      "step": 18490
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 1.3626949787139893,
+      "learning_rate": 4.114128813288919e-05,
+      "loss": 1.1889,
+      "step": 18495
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 1.5440853834152222,
+      "learning_rate": 4.113680118725355e-05,
+      "loss": 1.333,
+      "step": 18500
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 4.131174564361572,
+      "learning_rate": 4.1132313350399155e-05,
+      "loss": 1.0847,
+      "step": 18505
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 2.0806546211242676,
+      "learning_rate": 4.112782462257386e-05,
+      "loss": 1.3398,
+      "step": 18510
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 2.146475076675415,
+      "learning_rate": 4.112333500402558e-05,
+      "loss": 1.1576,
+      "step": 18515
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 2.3793232440948486,
+      "learning_rate": 4.111884449500225e-05,
+      "loss": 1.2433,
+      "step": 18520
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 1.008089303970337,
+      "learning_rate": 4.11143530957519e-05,
+      "loss": 1.2024,
+      "step": 18525
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 2.3187623023986816,
+      "learning_rate": 4.110986080652259e-05,
+      "loss": 1.1646,
+      "step": 18530
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 2.2866110801696777,
+      "learning_rate": 4.1105367627562405e-05,
+      "loss": 1.2074,
+      "step": 18535
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 3.7783994674682617,
+      "learning_rate": 4.110087355911951e-05,
+      "loss": 1.1768,
+      "step": 18540
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 3.170124053955078,
+      "learning_rate": 4.1096378601442095e-05,
+      "loss": 1.0867,
+      "step": 18545
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 1.4972915649414062,
+      "learning_rate": 4.109188275477843e-05,
+      "loss": 1.0619,
+      "step": 18550
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 3.528488874435425,
+      "learning_rate": 4.1087386019376804e-05,
+      "loss": 1.3686,
+      "step": 18555
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 2.188445806503296,
+      "learning_rate": 4.108288839548557e-05,
+      "loss": 1.2527,
+      "step": 18560
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 1.9165037870407104,
+      "learning_rate": 4.107838988335313e-05,
+      "loss": 1.3966,
+      "step": 18565
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 3.255675792694092,
+      "learning_rate": 4.1073890483227925e-05,
+      "loss": 1.1216,
+      "step": 18570
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 1.5940176248550415,
+      "learning_rate": 4.106939019535846e-05,
+      "loss": 1.2242,
+      "step": 18575
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 0.7701283097267151,
+      "learning_rate": 4.106488901999328e-05,
+      "loss": 1.3358,
+      "step": 18580
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 1.4922525882720947,
+      "learning_rate": 4.106038695738097e-05,
+      "loss": 1.1585,
+      "step": 18585
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 3.2928543090820312,
+      "learning_rate": 4.105588400777018e-05,
+      "loss": 1.1905,
+      "step": 18590
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 2.7126290798187256,
+      "learning_rate": 4.1051380171409616e-05,
+      "loss": 1.2003,
+      "step": 18595
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 3.020399332046509,
+      "learning_rate": 4.104687544854801e-05,
+      "loss": 1.1377,
+      "step": 18600
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 0.8808854222297668,
+      "learning_rate": 4.104236983943415e-05,
+      "loss": 1.2103,
+      "step": 18605
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 4.727932929992676,
+      "learning_rate": 4.1037863344316875e-05,
+      "loss": 1.1696,
+      "step": 18610
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 6.011281967163086,
+      "learning_rate": 4.103335596344508e-05,
+      "loss": 1.199,
+      "step": 18615
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 1.4836233854293823,
+      "learning_rate": 4.10288476970677e-05,
+      "loss": 1.1668,
+      "step": 18620
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 1.927235722541809,
+      "learning_rate": 4.1024338545433724e-05,
+      "loss": 1.2543,
+      "step": 18625
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 3.635470390319824,
+      "learning_rate": 4.101982850879218e-05,
+      "loss": 1.0165,
+      "step": 18630
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 2.8242850303649902,
+      "learning_rate": 4.101531758739217e-05,
+      "loss": 1.3089,
+      "step": 18635
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 2.512054920196533,
+      "learning_rate": 4.101080578148281e-05,
+      "loss": 1.1262,
+      "step": 18640
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 2.9997546672821045,
+      "learning_rate": 4.10062930913133e-05,
+      "loss": 1.2634,
+      "step": 18645
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 2.239625930786133,
+      "learning_rate": 4.1001779517132846e-05,
+      "loss": 1.1789,
+      "step": 18650
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 1.2925533056259155,
+      "learning_rate": 4.099726505919075e-05,
+      "loss": 1.3636,
+      "step": 18655
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 2.2485687732696533,
+      "learning_rate": 4.099274971773632e-05,
+      "loss": 1.1971,
+      "step": 18660
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 1.4676896333694458,
+      "learning_rate": 4.098823349301896e-05,
+      "loss": 1.0614,
+      "step": 18665
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 1.4249932765960693,
+      "learning_rate": 4.0983716385288083e-05,
+      "loss": 0.9559,
+      "step": 18670
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 1.0708461999893188,
+      "learning_rate": 4.097919839479316e-05,
+      "loss": 1.2269,
+      "step": 18675
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 2.775709390640259,
+      "learning_rate": 4.097467952178372e-05,
+      "loss": 1.1256,
+      "step": 18680
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 3.5211985111236572,
+      "learning_rate": 4.097015976650934e-05,
+      "loss": 1.3445,
+      "step": 18685
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 4.323126316070557,
+      "learning_rate": 4.0965639129219626e-05,
+      "loss": 1.3169,
+      "step": 18690
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 1.8097007274627686,
+      "learning_rate": 4.096111761016426e-05,
+      "loss": 1.219,
+      "step": 18695
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 1.6146286725997925,
+      "learning_rate": 4.095659520959297e-05,
+      "loss": 1.3955,
+      "step": 18700
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 2.256901741027832,
+      "learning_rate": 4.09520719277555e-05,
+      "loss": 0.9899,
+      "step": 18705
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 2.670027494430542,
+      "learning_rate": 4.094754776490168e-05,
+      "loss": 1.2336,
+      "step": 18710
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 1.1742836236953735,
+      "learning_rate": 4.094302272128138e-05,
+      "loss": 1.3383,
+      "step": 18715
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 1.9772650003433228,
+      "learning_rate": 4.0938496797144506e-05,
+      "loss": 1.2239,
+      "step": 18720
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 1.8488539457321167,
+      "learning_rate": 4.093396999274102e-05,
+      "loss": 1.0871,
+      "step": 18725
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 3.4919984340667725,
+      "learning_rate": 4.092944230832093e-05,
+      "loss": 1.2676,
+      "step": 18730
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 1.2979923486709595,
+      "learning_rate": 4.092491374413431e-05,
+      "loss": 1.2437,
+      "step": 18735
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 1.2796027660369873,
+      "learning_rate": 4.092038430043125e-05,
+      "loss": 1.1508,
+      "step": 18740
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 1.175561547279358,
+      "learning_rate": 4.09158539774619e-05,
+      "loss": 1.1637,
+      "step": 18745
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 3.259338140487671,
+      "learning_rate": 4.0911322775476494e-05,
+      "loss": 1.3101,
+      "step": 18750
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 1.183301568031311,
+      "learning_rate": 4.0906790694725275e-05,
+      "loss": 1.1849,
+      "step": 18755
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 3.8314900398254395,
+      "learning_rate": 4.090225773545853e-05,
+      "loss": 1.2224,
+      "step": 18760
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 1.5235013961791992,
+      "learning_rate": 4.089772389792662e-05,
+      "loss": 1.3048,
+      "step": 18765
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 2.4287803173065186,
+      "learning_rate": 4.089318918237994e-05,
+      "loss": 1.1388,
+      "step": 18770
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 1.4137362241744995,
+      "learning_rate": 4.0888653589068946e-05,
+      "loss": 1.2584,
+      "step": 18775
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 1.061883807182312,
+      "learning_rate": 4.0884117118244136e-05,
+      "loss": 1.3847,
+      "step": 18780
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 3.9546823501586914,
+      "learning_rate": 4.087957977015604e-05,
+      "loss": 1.2478,
+      "step": 18785
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 1.4414864778518677,
+      "learning_rate": 4.087504154505526e-05,
+      "loss": 1.3395,
+      "step": 18790
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 1.690242052078247,
+      "learning_rate": 4.0870502443192446e-05,
+      "loss": 1.0451,
+      "step": 18795
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 2.187337636947632,
+      "learning_rate": 4.086596246481826e-05,
+      "loss": 1.1884,
+      "step": 18800
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 2.2674190998077393,
+      "learning_rate": 4.086142161018347e-05,
+      "loss": 1.323,
+      "step": 18805
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 2.5989344120025635,
+      "learning_rate": 4.0856879879538854e-05,
+      "loss": 1.1205,
+      "step": 18810
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 1.1521856784820557,
+      "learning_rate": 4.085233727313524e-05,
+      "loss": 1.2544,
+      "step": 18815
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 3.605713367462158,
+      "learning_rate": 4.084779379122352e-05,
+      "loss": 1.2099,
+      "step": 18820
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 1.6035972833633423,
+      "learning_rate": 4.0843249434054624e-05,
+      "loss": 1.3516,
+      "step": 18825
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 1.7690247297286987,
+      "learning_rate": 4.083870420187953e-05,
+      "loss": 1.2547,
+      "step": 18830
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 1.6774495840072632,
+      "learning_rate": 4.083415809494926e-05,
+      "loss": 1.1669,
+      "step": 18835
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 2.1360936164855957,
+      "learning_rate": 4.08296111135149e-05,
+      "loss": 1.1833,
+      "step": 18840
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 3.100994110107422,
+      "learning_rate": 4.082506325782757e-05,
+      "loss": 1.1586,
+      "step": 18845
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 2.5224337577819824,
+      "learning_rate": 4.0820514528138444e-05,
+      "loss": 1.2535,
+      "step": 18850
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 1.2927287817001343,
+      "learning_rate": 4.0815964924698745e-05,
+      "loss": 1.1492,
+      "step": 18855
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 1.2006562948226929,
+      "learning_rate": 4.0811414447759755e-05,
+      "loss": 1.1727,
+      "step": 18860
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 2.254915475845337,
+      "learning_rate": 4.080686309757277e-05,
+      "loss": 1.2347,
+      "step": 18865
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 1.6122604608535767,
+      "learning_rate": 4.080231087438916e-05,
+      "loss": 1.0974,
+      "step": 18870
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 2.5616037845611572,
+      "learning_rate": 4.079775777846036e-05,
+      "loss": 1.1608,
+      "step": 18875
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 4.505390644073486,
+      "learning_rate": 4.07932038100378e-05,
+      "loss": 1.1997,
+      "step": 18880
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 3.1718831062316895,
+      "learning_rate": 4.0788648969373035e-05,
+      "loss": 1.2127,
+      "step": 18885
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 2.7245662212371826,
+      "learning_rate": 4.078409325671758e-05,
+      "loss": 1.2764,
+      "step": 18890
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 0.9645751118659973,
+      "learning_rate": 4.0779536672323074e-05,
+      "loss": 1.2709,
+      "step": 18895
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 2.0981924533843994,
+      "learning_rate": 4.077497921644115e-05,
+      "loss": 1.2451,
+      "step": 18900
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 1.6078838109970093,
+      "learning_rate": 4.077042088932352e-05,
+      "loss": 1.1987,
+      "step": 18905
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 3.138634443283081,
+      "learning_rate": 4.0765861691221943e-05,
+      "loss": 1.2224,
+      "step": 18910
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 1.9491277933120728,
+      "learning_rate": 4.0761301622388204e-05,
+      "loss": 1.0712,
+      "step": 18915
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 1.6134378910064697,
+      "learning_rate": 4.075674068307417e-05,
+      "loss": 1.2271,
+      "step": 18920
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 4.156822681427002,
+      "learning_rate": 4.075217887353172e-05,
+      "loss": 1.2066,
+      "step": 18925
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 2.2299163341522217,
+      "learning_rate": 4.074761619401281e-05,
+      "loss": 1.2968,
+      "step": 18930
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 1.9773372411727905,
+      "learning_rate": 4.0743052644769416e-05,
+      "loss": 1.1542,
+      "step": 18935
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 1.7787487506866455,
+      "learning_rate": 4.0738488226053595e-05,
+      "loss": 1.2944,
+      "step": 18940
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 1.2517333030700684,
+      "learning_rate": 4.073392293811742e-05,
+      "loss": 1.2636,
+      "step": 18945
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 2.801917314529419,
+      "learning_rate": 4.072935678121305e-05,
+      "loss": 1.2396,
+      "step": 18950
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 2.199772357940674,
+      "learning_rate": 4.0724789755592654e-05,
+      "loss": 1.3366,
+      "step": 18955
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 1.5385900735855103,
+      "learning_rate": 4.072022186150846e-05,
+      "loss": 1.1959,
+      "step": 18960
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 1.959558367729187,
+      "learning_rate": 4.0715653099212744e-05,
+      "loss": 1.1745,
+      "step": 18965
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 4.36882209777832,
+      "learning_rate": 4.071108346895786e-05,
+      "loss": 1.1226,
+      "step": 18970
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 3.006443738937378,
+      "learning_rate": 4.0706512970996145e-05,
+      "loss": 1.1088,
+      "step": 18975
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 1.4092053174972534,
+      "learning_rate": 4.070194160558006e-05,
+      "loss": 1.2145,
+      "step": 18980
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 2.4782750606536865,
+      "learning_rate": 4.069736937296206e-05,
+      "loss": 1.2773,
+      "step": 18985
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 2.681567430496216,
+      "learning_rate": 4.069279627339466e-05,
+      "loss": 1.2458,
+      "step": 18990
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 2.1898365020751953,
+      "learning_rate": 4.068822230713044e-05,
+      "loss": 1.2287,
+      "step": 18995
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 1.1778751611709595,
+      "learning_rate": 4.068364747442201e-05,
+      "loss": 1.3985,
+      "step": 19000
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 2.775921106338501,
+      "learning_rate": 4.0679071775522024e-05,
+      "loss": 1.3502,
+      "step": 19005
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 0.9329620003700256,
+      "learning_rate": 4.0674495210683214e-05,
+      "loss": 1.0845,
+      "step": 19010
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 2.163052797317505,
+      "learning_rate": 4.0669917780158315e-05,
+      "loss": 1.1712,
+      "step": 19015
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 1.6568050384521484,
+      "learning_rate": 4.066533948420015e-05,
+      "loss": 1.3567,
+      "step": 19020
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 1.2599117755889893,
+      "learning_rate": 4.0660760323061564e-05,
+      "loss": 1.4662,
+      "step": 19025
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 1.5358624458312988,
+      "learning_rate": 4.065618029699547e-05,
+      "loss": 1.372,
+      "step": 19030
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 2.070814609527588,
+      "learning_rate": 4.06515994062548e-05,
+      "loss": 1.2393,
+      "step": 19035
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 1.2835952043533325,
+      "learning_rate": 4.0647017651092575e-05,
+      "loss": 1.1488,
+      "step": 19040
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 3.1609346866607666,
+      "learning_rate": 4.064243503176183e-05,
+      "loss": 1.2701,
+      "step": 19045
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 4.8114471435546875,
+      "learning_rate": 4.063785154851565e-05,
+      "loss": 1.1842,
+      "step": 19050
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 2.764596462249756,
+      "learning_rate": 4.0633267201607197e-05,
+      "loss": 1.1671,
+      "step": 19055
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 1.5256291627883911,
+      "learning_rate": 4.062868199128964e-05,
+      "loss": 1.2943,
+      "step": 19060
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 3.720947504043579,
+      "learning_rate": 4.062409591781622e-05,
+      "loss": 1.2156,
+      "step": 19065
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 1.2810076475143433,
+      "learning_rate": 4.061950898144021e-05,
+      "loss": 1.1736,
+      "step": 19070
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 1.708155632019043,
+      "learning_rate": 4.061492118241497e-05,
+      "loss": 1.2937,
+      "step": 19075
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 2.4375946521759033,
+      "learning_rate": 4.0610332520993866e-05,
+      "loss": 1.2401,
+      "step": 19080
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 2.5409374237060547,
+      "learning_rate": 4.060574299743032e-05,
+      "loss": 1.1797,
+      "step": 19085
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 1.2663114070892334,
+      "learning_rate": 4.0601152611977797e-05,
+      "loss": 1.3379,
+      "step": 19090
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 1.4234334230422974,
+      "learning_rate": 4.059656136488985e-05,
+      "loss": 1.2776,
+      "step": 19095
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 2.477080821990967,
+      "learning_rate": 4.059196925642002e-05,
+      "loss": 1.1275,
+      "step": 19100
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 2.8818910121917725,
+      "learning_rate": 4.058737628682193e-05,
+      "loss": 1.2485,
+      "step": 19105
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 1.8685482740402222,
+      "learning_rate": 4.0582782456349264e-05,
+      "loss": 1.1152,
+      "step": 19110
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 2.1116809844970703,
+      "learning_rate": 4.057818776525571e-05,
+      "loss": 1.2397,
+      "step": 19115
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 3.2058467864990234,
+      "learning_rate": 4.0573592213795045e-05,
+      "loss": 1.3506,
+      "step": 19120
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 2.4281482696533203,
+      "learning_rate": 4.0568995802221066e-05,
+      "loss": 1.0978,
+      "step": 19125
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 3.1970787048339844,
+      "learning_rate": 4.056439853078763e-05,
+      "loss": 1.1549,
+      "step": 19130
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 4.877747535705566,
+      "learning_rate": 4.0559800399748645e-05,
+      "loss": 1.186,
+      "step": 19135
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 1.5900225639343262,
+      "learning_rate": 4.055520140935806e-05,
+      "loss": 1.2458,
+      "step": 19140
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 2.55224871635437,
+      "learning_rate": 4.055060155986986e-05,
+      "loss": 1.2647,
+      "step": 19145
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 1.5223407745361328,
+      "learning_rate": 4.054600085153811e-05,
+      "loss": 1.2311,
+      "step": 19150
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 1.986936330795288,
+      "learning_rate": 4.054139928461689e-05,
+      "loss": 1.1517,
+      "step": 19155
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 2.5802133083343506,
+      "learning_rate": 4.0536796859360336e-05,
+      "loss": 1.2503,
+      "step": 19160
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 1.5326321125030518,
+      "learning_rate": 4.053219357602265e-05,
+      "loss": 1.2379,
+      "step": 19165
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 2.684556484222412,
+      "learning_rate": 4.0527589434858046e-05,
+      "loss": 1.2704,
+      "step": 19170
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 1.2292155027389526,
+      "learning_rate": 4.0522984436120826e-05,
+      "loss": 1.4196,
+      "step": 19175
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 3.3510990142822266,
+      "learning_rate": 4.051837858006531e-05,
+      "loss": 1.0128,
+      "step": 19180
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 2.779120922088623,
+      "learning_rate": 4.051377186694588e-05,
+      "loss": 1.3174,
+      "step": 19185
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 2.3590750694274902,
+      "learning_rate": 4.0509164297016944e-05,
+      "loss": 1.2303,
+      "step": 19190
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 1.3253639936447144,
+      "learning_rate": 4.050455587053299e-05,
+      "loss": 1.3627,
+      "step": 19195
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 2.0595991611480713,
+      "learning_rate": 4.049994658774853e-05,
+      "loss": 1.2418,
+      "step": 19200
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 1.3671808242797852,
+      "learning_rate": 4.0495336448918135e-05,
+      "loss": 1.2585,
+      "step": 19205
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 1.1783967018127441,
+      "learning_rate": 4.0490725454296414e-05,
+      "loss": 1.1087,
+      "step": 19210
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 1.43436598777771,
+      "learning_rate": 4.048611360413803e-05,
+      "loss": 1.0994,
+      "step": 19215
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 2.8289105892181396,
+      "learning_rate": 4.048150089869768e-05,
+      "loss": 1.2857,
+      "step": 19220
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 2.906419277191162,
+      "learning_rate": 4.047688733823013e-05,
+      "loss": 1.3255,
+      "step": 19225
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 1.00028395652771,
+      "learning_rate": 4.0472272922990185e-05,
+      "loss": 1.3454,
+      "step": 19230
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 1.3520787954330444,
+      "learning_rate": 4.046765765323269e-05,
+      "loss": 1.2744,
+      "step": 19235
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 3.6939780712127686,
+      "learning_rate": 4.046304152921253e-05,
+      "loss": 1.277,
+      "step": 19240
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 1.3129464387893677,
+      "learning_rate": 4.045842455118467e-05,
+      "loss": 1.2569,
+      "step": 19245
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 1.3802201747894287,
+      "learning_rate": 4.045380671940409e-05,
+      "loss": 1.3112,
+      "step": 19250
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 2.282783031463623,
+      "learning_rate": 4.0449188034125825e-05,
+      "loss": 1.1128,
+      "step": 19255
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 2.5896713733673096,
+      "learning_rate": 4.044456849560496e-05,
+      "loss": 1.2365,
+      "step": 19260
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 1.469196081161499,
+      "learning_rate": 4.043994810409664e-05,
+      "loss": 1.2037,
+      "step": 19265
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 2.223722457885742,
+      "learning_rate": 4.043532685985602e-05,
+      "loss": 1.377,
+      "step": 19270
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 1.6373085975646973,
+      "learning_rate": 4.043070476313835e-05,
+      "loss": 1.1027,
+      "step": 19275
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 1.401114821434021,
+      "learning_rate": 4.0426081814198905e-05,
+      "loss": 1.148,
+      "step": 19280
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 1.4653375148773193,
+      "learning_rate": 4.042145801329298e-05,
+      "loss": 1.3143,
+      "step": 19285
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 1.659494161605835,
+      "learning_rate": 4.0416833360675966e-05,
+      "loss": 1.2026,
+      "step": 19290
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 2.820892810821533,
+      "learning_rate": 4.0412207856603266e-05,
+      "loss": 1.1208,
+      "step": 19295
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 2.835115909576416,
+      "learning_rate": 4.040758150133035e-05,
+      "loss": 1.2334,
+      "step": 19300
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 1.002402901649475,
+      "learning_rate": 4.040295429511273e-05,
+      "loss": 1.2697,
+      "step": 19305
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 1.9748258590698242,
+      "learning_rate": 4.0398326238205946e-05,
+      "loss": 1.3538,
+      "step": 19310
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 1.5948344469070435,
+      "learning_rate": 4.039369733086561e-05,
+      "loss": 1.2056,
+      "step": 19315
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 5.312704563140869,
+      "learning_rate": 4.038906757334737e-05,
+      "loss": 1.2161,
+      "step": 19320
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 2.936213970184326,
+      "learning_rate": 4.0384436965906924e-05,
+      "loss": 1.1106,
+      "step": 19325
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 2.379530429840088,
+      "learning_rate": 4.037980550880002e-05,
+      "loss": 1.2153,
+      "step": 19330
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 3.4053995609283447,
+      "learning_rate": 4.0375173202282444e-05,
+      "loss": 1.1903,
+      "step": 19335
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 2.1918978691101074,
+      "learning_rate": 4.0370540046610026e-05,
+      "loss": 1.1617,
+      "step": 19340
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 1.1753573417663574,
+      "learning_rate": 4.036590604203867e-05,
+      "loss": 1.141,
+      "step": 19345
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 2.638122081756592,
+      "learning_rate": 4.036127118882429e-05,
+      "loss": 1.0728,
+      "step": 19350
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 1.511003017425537,
+      "learning_rate": 4.035663548722287e-05,
+      "loss": 1.3012,
+      "step": 19355
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 1.878172516822815,
+      "learning_rate": 4.035199893749043e-05,
+      "loss": 1.2585,
+      "step": 19360
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 1.5874683856964111,
+      "learning_rate": 4.0347361539883045e-05,
+      "loss": 1.1257,
+      "step": 19365
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 2.5368497371673584,
+      "learning_rate": 4.034272329465684e-05,
+      "loss": 1.277,
+      "step": 19370
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 1.9644641876220703,
+      "learning_rate": 4.033808420206798e-05,
+      "loss": 1.267,
+      "step": 19375
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 3.0246572494506836,
+      "learning_rate": 4.0333444262372666e-05,
+      "loss": 1.1498,
+      "step": 19380
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 2.288479804992676,
+      "learning_rate": 4.032880347582716e-05,
+      "loss": 1.2349,
+      "step": 19385
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 1.7660423517227173,
+      "learning_rate": 4.032416184268778e-05,
+      "loss": 1.2463,
+      "step": 19390
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 2.469409465789795,
+      "learning_rate": 4.031951936321086e-05,
+      "loss": 1.2732,
+      "step": 19395
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 2.5984413623809814,
+      "learning_rate": 4.0314876037652814e-05,
+      "loss": 1.1642,
+      "step": 19400
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 4.0700225830078125,
+      "learning_rate": 4.0310231866270086e-05,
+      "loss": 1.1763,
+      "step": 19405
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 1.4601423740386963,
+      "learning_rate": 4.030558684931917e-05,
+      "loss": 1.1787,
+      "step": 19410
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 2.451631546020508,
+      "learning_rate": 4.0300940987056596e-05,
+      "loss": 1.0628,
+      "step": 19415
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 2.339339017868042,
+      "learning_rate": 4.029629427973895e-05,
+      "loss": 1.1614,
+      "step": 19420
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 2.694387674331665,
+      "learning_rate": 4.0291646727622875e-05,
+      "loss": 1.1025,
+      "step": 19425
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 3.171658754348755,
+      "learning_rate": 4.0286998330965056e-05,
+      "loss": 1.2235,
+      "step": 19430
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 3.675020694732666,
+      "learning_rate": 4.02823490900222e-05,
+      "loss": 1.2227,
+      "step": 19435
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 1.9126026630401611,
+      "learning_rate": 4.027769900505109e-05,
+      "loss": 1.0964,
+      "step": 19440
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 2.5049381256103516,
+      "learning_rate": 4.027304807630854e-05,
+      "loss": 1.3127,
+      "step": 19445
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 1.2870092391967773,
+      "learning_rate": 4.0268396304051426e-05,
+      "loss": 1.3275,
+      "step": 19450
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 2.4361467361450195,
+      "learning_rate": 4.026374368853665e-05,
+      "loss": 1.0307,
+      "step": 19455
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 2.4936602115631104,
+      "learning_rate": 4.025909023002118e-05,
+      "loss": 1.176,
+      "step": 19460
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 3.3425686359405518,
+      "learning_rate": 4.025443592876201e-05,
+      "loss": 1.0798,
+      "step": 19465
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 2.0376148223876953,
+      "learning_rate": 4.024978078501621e-05,
+      "loss": 1.1973,
+      "step": 19470
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 4.597989559173584,
+      "learning_rate": 4.0245124799040864e-05,
+      "loss": 1.3608,
+      "step": 19475
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 1.8406076431274414,
+      "learning_rate": 4.024046797109312e-05,
+      "loss": 1.2414,
+      "step": 19480
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 1.4122562408447266,
+      "learning_rate": 4.023581030143018e-05,
+      "loss": 1.2824,
+      "step": 19485
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 1.3209871053695679,
+      "learning_rate": 4.023115179030926e-05,
+      "loss": 1.2982,
+      "step": 19490
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 2.4503726959228516,
+      "learning_rate": 4.0226492437987676e-05,
+      "loss": 1.1107,
+      "step": 19495
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 1.1615526676177979,
+      "learning_rate": 4.022183224472272e-05,
+      "loss": 1.1436,
+      "step": 19500
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 2.6597816944122314,
+      "learning_rate": 4.021717121077181e-05,
+      "loss": 1.189,
+      "step": 19505
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 2.0554018020629883,
+      "learning_rate": 4.0212509336392345e-05,
+      "loss": 1.1992,
+      "step": 19510
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 2.7470314502716064,
+      "learning_rate": 4.02078466218418e-05,
+      "loss": 1.378,
+      "step": 19515
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 1.204832673072815,
+      "learning_rate": 4.020318306737769e-05,
+      "loss": 1.1576,
+      "step": 19520
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 1.8212392330169678,
+      "learning_rate": 4.019851867325759e-05,
+      "loss": 1.3113,
+      "step": 19525
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 2.319030523300171,
+      "learning_rate": 4.01938534397391e-05,
+      "loss": 1.313,
+      "step": 19530
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 1.1338316202163696,
+      "learning_rate": 4.018918736707988e-05,
+      "loss": 1.1278,
+      "step": 19535
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 1.2906646728515625,
+      "learning_rate": 4.018452045553762e-05,
+      "loss": 1.2246,
+      "step": 19540
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 3.1656301021575928,
+      "learning_rate": 4.017985270537009e-05,
+      "loss": 1.2851,
+      "step": 19545
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 1.0056490898132324,
+      "learning_rate": 4.017518411683507e-05,
+      "loss": 1.1831,
+      "step": 19550
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 1.4432793855667114,
+      "learning_rate": 4.01705146901904e-05,
+      "loss": 1.2026,
+      "step": 19555
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 2.5495402812957764,
+      "learning_rate": 4.016584442569398e-05,
+      "loss": 1.1733,
+      "step": 19560
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 4.6060099601745605,
+      "learning_rate": 4.016117332360373e-05,
+      "loss": 1.1664,
+      "step": 19565
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 1.2806930541992188,
+      "learning_rate": 4.015650138417764e-05,
+      "loss": 1.2543,
+      "step": 19570
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 1.6852878332138062,
+      "learning_rate": 4.015182860767373e-05,
+      "loss": 1.1064,
+      "step": 19575
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 1.5390417575836182,
+      "learning_rate": 4.014715499435008e-05,
+      "loss": 1.2544,
+      "step": 19580
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 2.724890947341919,
+      "learning_rate": 4.01424805444648e-05,
+      "loss": 1.1592,
+      "step": 19585
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 5.490366458892822,
+      "learning_rate": 4.013780525827606e-05,
+      "loss": 1.151,
+      "step": 19590
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 2.0615406036376953,
+      "learning_rate": 4.0133129136042066e-05,
+      "loss": 1.2411,
+      "step": 19595
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 5.847554683685303,
+      "learning_rate": 4.012845217802109e-05,
+      "loss": 1.2015,
+      "step": 19600
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 1.873935341835022,
+      "learning_rate": 4.0123774384471425e-05,
+      "loss": 1.2761,
+      "step": 19605
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 2.046846389770508,
+      "learning_rate": 4.0119095755651414e-05,
+      "loss": 1.2653,
+      "step": 19610
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 2.1344308853149414,
+      "learning_rate": 4.011441629181946e-05,
+      "loss": 1.3688,
+      "step": 19615
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 1.3133376836776733,
+      "learning_rate": 4.010973599323401e-05,
+      "loss": 1.3191,
+      "step": 19620
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 1.7745246887207031,
+      "learning_rate": 4.010505486015354e-05,
+      "loss": 1.1906,
+      "step": 19625
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 2.6860477924346924,
+      "learning_rate": 4.010037289283659e-05,
+      "loss": 1.1099,
+      "step": 19630
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 2.37652850151062,
+      "learning_rate": 4.009569009154175e-05,
+      "loss": 1.0423,
+      "step": 19635
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 1.3366721868515015,
+      "learning_rate": 4.0091006456527634e-05,
+      "loss": 1.3428,
+      "step": 19640
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 5.54878568649292,
+      "learning_rate": 4.008632198805292e-05,
+      "loss": 1.1666,
+      "step": 19645
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 0.9139274954795837,
+      "learning_rate": 4.008163668637632e-05,
+      "loss": 1.0742,
+      "step": 19650
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 1.8782241344451904,
+      "learning_rate": 4.00769505517566e-05,
+      "loss": 1.2777,
+      "step": 19655
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 1.8892041444778442,
+      "learning_rate": 4.0072263584452576e-05,
+      "loss": 1.3023,
+      "step": 19660
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 1.7364705801010132,
+      "learning_rate": 4.0067575784723104e-05,
+      "loss": 1.2377,
+      "step": 19665
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 0.6901882886886597,
+      "learning_rate": 4.0062887152827075e-05,
+      "loss": 1.2162,
+      "step": 19670
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 1.6840230226516724,
+      "learning_rate": 4.005819768902346e-05,
+      "loss": 1.3165,
+      "step": 19675
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 1.284379243850708,
+      "learning_rate": 4.005350739357122e-05,
+      "loss": 1.2768,
+      "step": 19680
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 1.735634446144104,
+      "learning_rate": 4.004881626672943e-05,
+      "loss": 1.1276,
+      "step": 19685
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 1.7231804132461548,
+      "learning_rate": 4.004412430875715e-05,
+      "loss": 1.2146,
+      "step": 19690
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 3.919288396835327,
+      "learning_rate": 4.0039431519913525e-05,
+      "loss": 1.1341,
+      "step": 19695
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 1.3154077529907227,
+      "learning_rate": 4.0034737900457734e-05,
+      "loss": 1.209,
+      "step": 19700
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 2.288764715194702,
+      "learning_rate": 4.003004345064899e-05,
+      "loss": 1.2333,
+      "step": 19705
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 1.616042137145996,
+      "learning_rate": 4.002534817074657e-05,
+      "loss": 1.2067,
+      "step": 19710
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 2.097806692123413,
+      "learning_rate": 4.002065206100979e-05,
+      "loss": 1.2761,
+      "step": 19715
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 4.625036716461182,
+      "learning_rate": 4.001595512169801e-05,
+      "loss": 1.1133,
+      "step": 19720
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 1.4965091943740845,
+      "learning_rate": 4.001125735307063e-05,
+      "loss": 1.2998,
+      "step": 19725
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 1.4583796262741089,
+      "learning_rate": 4.000655875538712e-05,
+      "loss": 1.1086,
+      "step": 19730
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 2.9411721229553223,
+      "learning_rate": 4.000185932890697e-05,
+      "loss": 1.2033,
+      "step": 19735
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 1.419301986694336,
+      "learning_rate": 3.999715907388971e-05,
+      "loss": 1.2032,
+      "step": 19740
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 1.0543030500411987,
+      "learning_rate": 3.999245799059496e-05,
+      "loss": 1.3008,
+      "step": 19745
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 4.684767723083496,
+      "learning_rate": 3.998775607928232e-05,
+      "loss": 1.3266,
+      "step": 19750
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 3.58650541305542,
+      "learning_rate": 3.99830533402115e-05,
+      "loss": 1.2931,
+      "step": 19755
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 1.2763997316360474,
+      "learning_rate": 3.997834977364222e-05,
+      "loss": 1.0974,
+      "step": 19760
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 2.0618629455566406,
+      "learning_rate": 3.9973645379834255e-05,
+      "loss": 1.1197,
+      "step": 19765
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 1.2482560873031616,
+      "learning_rate": 3.9968940159047416e-05,
+      "loss": 1.1536,
+      "step": 19770
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 1.5299739837646484,
+      "learning_rate": 3.9964234111541567e-05,
+      "loss": 1.2666,
+      "step": 19775
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 6.283170223236084,
+      "learning_rate": 3.9959527237576624e-05,
+      "loss": 1.2071,
+      "step": 19780
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 1.4811429977416992,
+      "learning_rate": 3.995481953741254e-05,
+      "loss": 1.1695,
+      "step": 19785
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 2.082146406173706,
+      "learning_rate": 3.995011101130932e-05,
+      "loss": 1.2389,
+      "step": 19790
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 1.7142033576965332,
+      "learning_rate": 3.994540165952701e-05,
+      "loss": 1.3164,
+      "step": 19795
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 1.3246126174926758,
+      "learning_rate": 3.99406914823257e-05,
+      "loss": 1.33,
+      "step": 19800
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 3.4477603435516357,
+      "learning_rate": 3.993598047996553e-05,
+      "loss": 1.1553,
+      "step": 19805
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 1.8627071380615234,
+      "learning_rate": 3.9931268652706676e-05,
+      "loss": 1.1348,
+      "step": 19810
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 3.652338981628418,
+      "learning_rate": 3.992655600080938e-05,
+      "loss": 1.03,
+      "step": 19815
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 2.0452940464019775,
+      "learning_rate": 3.992184252453392e-05,
+      "loss": 1.0227,
+      "step": 19820
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 2.3490209579467773,
+      "learning_rate": 3.99171282241406e-05,
+      "loss": 1.2441,
+      "step": 19825
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 1.5200634002685547,
+      "learning_rate": 3.991241309988979e-05,
+      "loss": 1.2541,
+      "step": 19830
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 2.570288896560669,
+      "learning_rate": 3.9907697152041915e-05,
+      "loss": 1.3228,
+      "step": 19835
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 1.5330473184585571,
+      "learning_rate": 3.9902980380857414e-05,
+      "loss": 1.2349,
+      "step": 19840
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 1.0376076698303223,
+      "learning_rate": 3.9898262786596794e-05,
+      "loss": 1.1803,
+      "step": 19845
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 1.4914852380752563,
+      "learning_rate": 3.989354436952061e-05,
+      "loss": 1.1023,
+      "step": 19850
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 1.438751220703125,
+      "learning_rate": 3.988882512988945e-05,
+      "loss": 1.2842,
+      "step": 19855
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 2.238565683364868,
+      "learning_rate": 3.988410506796396e-05,
+      "loss": 1.2185,
+      "step": 19860
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 5.389926910400391,
+      "learning_rate": 3.987938418400482e-05,
+      "loss": 1.1369,
+      "step": 19865
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 4.195040702819824,
+      "learning_rate": 3.987466247827275e-05,
+      "loss": 1.1121,
+      "step": 19870
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 1.3536615371704102,
+      "learning_rate": 3.986993995102853e-05,
+      "loss": 1.3349,
+      "step": 19875
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 1.4799764156341553,
+      "learning_rate": 3.9865216602532994e-05,
+      "loss": 1.2105,
+      "step": 19880
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 1.5953404903411865,
+      "learning_rate": 3.986049243304699e-05,
+      "loss": 1.1709,
+      "step": 19885
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 3.760248899459839,
+      "learning_rate": 3.9855767442831436e-05,
+      "loss": 1.1154,
+      "step": 19890
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 4.010507106781006,
+      "learning_rate": 3.985104163214729e-05,
+      "loss": 1.3683,
+      "step": 19895
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 2.708240032196045,
+      "learning_rate": 3.984631500125555e-05,
+      "loss": 1.233,
+      "step": 19900
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 1.024338960647583,
+      "learning_rate": 3.984158755041726e-05,
+      "loss": 1.1541,
+      "step": 19905
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 3.15925931930542,
+      "learning_rate": 3.9836859279893526e-05,
+      "loss": 1.2759,
+      "step": 19910
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 3.913562536239624,
+      "learning_rate": 3.9832130189945475e-05,
+      "loss": 1.1862,
+      "step": 19915
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 2.1706655025482178,
+      "learning_rate": 3.982740028083428e-05,
+      "loss": 1.252,
+      "step": 19920
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 1.2419073581695557,
+      "learning_rate": 3.982266955282119e-05,
+      "loss": 1.3196,
+      "step": 19925
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 1.833250880241394,
+      "learning_rate": 3.9817938006167465e-05,
+      "loss": 1.1608,
+      "step": 19930
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 1.1308797597885132,
+      "learning_rate": 3.9813205641134424e-05,
+      "loss": 1.2362,
+      "step": 19935
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 1.6005351543426514,
+      "learning_rate": 3.980847245798344e-05,
+      "loss": 1.2911,
+      "step": 19940
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 1.2938203811645508,
+      "learning_rate": 3.9803738456975905e-05,
+      "loss": 1.342,
+      "step": 19945
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 5.012538433074951,
+      "learning_rate": 3.9799003638373283e-05,
+      "loss": 1.1956,
+      "step": 19950
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 2.233893394470215,
+      "learning_rate": 3.979426800243708e-05,
+      "loss": 1.2192,
+      "step": 19955
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 4.2855448722839355,
+      "learning_rate": 3.978953154942883e-05,
+      "loss": 1.2098,
+      "step": 19960
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 2.717545747756958,
+      "learning_rate": 3.978479427961012e-05,
+      "loss": 1.265,
+      "step": 19965
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 2.452308177947998,
+      "learning_rate": 3.97800561932426e-05,
+      "loss": 1.2003,
+      "step": 19970
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 3.6436851024627686,
+      "learning_rate": 3.977531729058793e-05,
+      "loss": 1.3046,
+      "step": 19975
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 2.9817821979522705,
+      "learning_rate": 3.977057757190785e-05,
+      "loss": 1.0906,
+      "step": 19980
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 2.8185782432556152,
+      "learning_rate": 3.9765837037464124e-05,
+      "loss": 1.3706,
+      "step": 19985
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 3.414907217025757,
+      "learning_rate": 3.9761095687518565e-05,
+      "loss": 1.1195,
+      "step": 19990
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 2.366452217102051,
+      "learning_rate": 3.9756353522333034e-05,
+      "loss": 1.3436,
+      "step": 19995
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 2.4633522033691406,
+      "learning_rate": 3.975161054216944e-05,
+      "loss": 1.2036,
+      "step": 20000
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 2.6752207279205322,
+      "learning_rate": 3.9746866747289726e-05,
+      "loss": 1.3225,
+      "step": 20005
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 1.7009152173995972,
+      "learning_rate": 3.9742122137955884e-05,
+      "loss": 1.2566,
+      "step": 20010
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 2.860818386077881,
+      "learning_rate": 3.9737376714429964e-05,
+      "loss": 1.2588,
+      "step": 20015
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 4.960500717163086,
+      "learning_rate": 3.973263047697405e-05,
+      "loss": 1.1748,
+      "step": 20020
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 1.3426363468170166,
+      "learning_rate": 3.972788342585027e-05,
+      "loss": 1.2358,
+      "step": 20025
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 5.134890556335449,
+      "learning_rate": 3.972313556132079e-05,
+      "loss": 1.2629,
+      "step": 20030
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 3.40563702583313,
+      "learning_rate": 3.971838688364784e-05,
+      "loss": 1.186,
+      "step": 20035
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 2.111565113067627,
+      "learning_rate": 3.9713637393093686e-05,
+      "loss": 1.2938,
+      "step": 20040
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.0406635999679565,
+      "learning_rate": 3.970888708992063e-05,
+      "loss": 1.1778,
+      "step": 20045
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 2.193850517272949,
+      "learning_rate": 3.9704135974391026e-05,
+      "loss": 1.206,
+      "step": 20050
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.8618187308311462,
+      "learning_rate": 3.969938404676728e-05,
+      "loss": 1.2366,
+      "step": 20055
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.8126907348632812,
+      "learning_rate": 3.969463130731183e-05,
+      "loss": 0.9741,
+      "step": 20060
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.3693639039993286,
+      "learning_rate": 3.968987775628717e-05,
+      "loss": 1.0962,
+      "step": 20065
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.7154806852340698,
+      "learning_rate": 3.9685123393955824e-05,
+      "loss": 1.4039,
+      "step": 20070
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 1.4863255023956299,
+      "learning_rate": 3.968036822058038e-05,
+      "loss": 1.1968,
+      "step": 20075
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 9.683221817016602,
+      "learning_rate": 3.9675612236423466e-05,
+      "loss": 1.1326,
+      "step": 20080
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 2.053426504135132,
+      "learning_rate": 3.9670855441747737e-05,
+      "loss": 1.2913,
+      "step": 20085
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 1.7866415977478027,
+      "learning_rate": 3.9666097836815915e-05,
+      "loss": 1.2354,
+      "step": 20090
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 1.6911275386810303,
+      "learning_rate": 3.9661339421890746e-05,
+      "loss": 1.0844,
+      "step": 20095
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 1.2909938097000122,
+      "learning_rate": 3.965658019723505e-05,
+      "loss": 1.2505,
+      "step": 20100
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 2.46484375,
+      "learning_rate": 3.965182016311165e-05,
+      "loss": 0.985,
+      "step": 20105
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 1.2409467697143555,
+      "learning_rate": 3.964705931978346e-05,
+      "loss": 1.314,
+      "step": 20110
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 1.6112093925476074,
+      "learning_rate": 3.964229766751342e-05,
+      "loss": 1.1812,
+      "step": 20115
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 2.3558475971221924,
+      "learning_rate": 3.9637535206564485e-05,
+      "loss": 1.2568,
+      "step": 20120
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 2.557868242263794,
+      "learning_rate": 3.96327719371997e-05,
+      "loss": 1.0185,
+      "step": 20125
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 1.7992500066757202,
+      "learning_rate": 3.962800785968213e-05,
+      "loss": 1.115,
+      "step": 20130
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 5.036953926086426,
+      "learning_rate": 3.962324297427489e-05,
+      "loss": 1.0766,
+      "step": 20135
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 1.705336332321167,
+      "learning_rate": 3.9618477281241146e-05,
+      "loss": 1.1394,
+      "step": 20140
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 1.967340111732483,
+      "learning_rate": 3.9613710780844096e-05,
+      "loss": 1.2392,
+      "step": 20145
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 2.6726629734039307,
+      "learning_rate": 3.9608943473346985e-05,
+      "loss": 1.2141,
+      "step": 20150
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 1.7967979907989502,
+      "learning_rate": 3.960417535901311e-05,
+      "loss": 1.0865,
+      "step": 20155
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 2.194570779800415,
+      "learning_rate": 3.959940643810581e-05,
+      "loss": 1.0641,
+      "step": 20160
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 3.644632339477539,
+      "learning_rate": 3.9594636710888475e-05,
+      "loss": 1.2237,
+      "step": 20165
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 1.7848505973815918,
+      "learning_rate": 3.9589866177624515e-05,
+      "loss": 1.1259,
+      "step": 20170
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 1.647910714149475,
+      "learning_rate": 3.958509483857742e-05,
+      "loss": 1.3501,
+      "step": 20175
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 2.3628077507019043,
+      "learning_rate": 3.95803226940107e-05,
+      "loss": 1.2317,
+      "step": 20180
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 1.816735863685608,
+      "learning_rate": 3.95755497441879e-05,
+      "loss": 0.9811,
+      "step": 20185
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 1.5639019012451172,
+      "learning_rate": 3.957077598937264e-05,
+      "loss": 1.1717,
+      "step": 20190
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 1.4008920192718506,
+      "learning_rate": 3.956600142982858e-05,
+      "loss": 1.1988,
+      "step": 20195
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 3.0697457790374756,
+      "learning_rate": 3.956122606581939e-05,
+      "loss": 1.0945,
+      "step": 20200
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 1.7359205484390259,
+      "learning_rate": 3.9556449897608824e-05,
+      "loss": 1.1508,
+      "step": 20205
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 2.08030104637146,
+      "learning_rate": 3.955167292546066e-05,
+      "loss": 1.0043,
+      "step": 20210
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 2.151505470275879,
+      "learning_rate": 3.9546895149638737e-05,
+      "loss": 1.3255,
+      "step": 20215
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 2.6911120414733887,
+      "learning_rate": 3.954211657040691e-05,
+      "loss": 1.0752,
+      "step": 20220
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 3.3755829334259033,
+      "learning_rate": 3.953733718802909e-05,
+      "loss": 1.3089,
+      "step": 20225
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 1.3556479215621948,
+      "learning_rate": 3.953255700276925e-05,
+      "loss": 1.1838,
+      "step": 20230
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 1.6404752731323242,
+      "learning_rate": 3.95277760148914e-05,
+      "loss": 1.1368,
+      "step": 20235
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 1.5696030855178833,
+      "learning_rate": 3.9522994224659576e-05,
+      "loss": 1.0713,
+      "step": 20240
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 0.9167082905769348,
+      "learning_rate": 3.951821163233788e-05,
+      "loss": 1.0272,
+      "step": 20245
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 2.055103063583374,
+      "learning_rate": 3.951342823819044e-05,
+      "loss": 1.1271,
+      "step": 20250
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 3.4541079998016357,
+      "learning_rate": 3.950864404248145e-05,
+      "loss": 1.3,
+      "step": 20255
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 3.939716100692749,
+      "learning_rate": 3.950385904547513e-05,
+      "loss": 1.2012,
+      "step": 20260
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 1.5123727321624756,
+      "learning_rate": 3.9499073247435755e-05,
+      "loss": 1.2772,
+      "step": 20265
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 1.4127657413482666,
+      "learning_rate": 3.949428664862762e-05,
+      "loss": 1.3369,
+      "step": 20270
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 2.0082712173461914,
+      "learning_rate": 3.948949924931512e-05,
+      "loss": 1.2377,
+      "step": 20275
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 1.3078731298446655,
+      "learning_rate": 3.9484711049762625e-05,
+      "loss": 1.097,
+      "step": 20280
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 1.3073629140853882,
+      "learning_rate": 3.94799220502346e-05,
+      "loss": 1.0726,
+      "step": 20285
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 1.1648201942443848,
+      "learning_rate": 3.9475132250995525e-05,
+      "loss": 1.2021,
+      "step": 20290
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 4.134536266326904,
+      "learning_rate": 3.947034165230995e-05,
+      "loss": 1.1795,
+      "step": 20295
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 1.0476043224334717,
+      "learning_rate": 3.946555025444244e-05,
+      "loss": 1.1613,
+      "step": 20300
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 2.554640769958496,
+      "learning_rate": 3.9460758057657626e-05,
+      "loss": 1.1222,
+      "step": 20305
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 3.514789342880249,
+      "learning_rate": 3.945596506222018e-05,
+      "loss": 1.0743,
+      "step": 20310
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 1.703673243522644,
+      "learning_rate": 3.945117126839481e-05,
+      "loss": 1.1762,
+      "step": 20315
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 2.6589813232421875,
+      "learning_rate": 3.944637667644627e-05,
+      "loss": 1.1546,
+      "step": 20320
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 3.949831008911133,
+      "learning_rate": 3.9441581286639365e-05,
+      "loss": 1.2165,
+      "step": 20325
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 2.2509963512420654,
+      "learning_rate": 3.943678509923893e-05,
+      "loss": 1.2908,
+      "step": 20330
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 2.6351144313812256,
+      "learning_rate": 3.9431988114509854e-05,
+      "loss": 1.2144,
+      "step": 20335
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 1.2921767234802246,
+      "learning_rate": 3.942719033271709e-05,
+      "loss": 1.2182,
+      "step": 20340
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 1.6154694557189941,
+      "learning_rate": 3.9422391754125596e-05,
+      "loss": 1.1248,
+      "step": 20345
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 1.5173944234848022,
+      "learning_rate": 3.94175923790004e-05,
+      "loss": 1.1687,
+      "step": 20350
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 1.5915690660476685,
+      "learning_rate": 3.941279220760655e-05,
+      "loss": 1.3048,
+      "step": 20355
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 2.2365562915802,
+      "learning_rate": 3.940799124020918e-05,
+      "loss": 1.1774,
+      "step": 20360
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 2.3418896198272705,
+      "learning_rate": 3.9403189477073424e-05,
+      "loss": 1.387,
+      "step": 20365
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 1.1092318296432495,
+      "learning_rate": 3.939838691846449e-05,
+      "loss": 1.2785,
+      "step": 20370
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 1.0265756845474243,
+      "learning_rate": 3.939358356464761e-05,
+      "loss": 1.139,
+      "step": 20375
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 1.068686842918396,
+      "learning_rate": 3.9388779415888075e-05,
+      "loss": 1.1121,
+      "step": 20380
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 1.821524977684021,
+      "learning_rate": 3.93839744724512e-05,
+      "loss": 1.008,
+      "step": 20385
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 2.26947021484375,
+      "learning_rate": 3.937916873460237e-05,
+      "loss": 1.2466,
+      "step": 20390
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 2.6822071075439453,
+      "learning_rate": 3.9374362202607e-05,
+      "loss": 1.0863,
+      "step": 20395
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 2.3898754119873047,
+      "learning_rate": 3.936955487673054e-05,
+      "loss": 1.1818,
+      "step": 20400
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 2.3284709453582764,
+      "learning_rate": 3.9364746757238515e-05,
+      "loss": 1.2534,
+      "step": 20405
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 2.5323164463043213,
+      "learning_rate": 3.935993784439644e-05,
+      "loss": 1.1993,
+      "step": 20410
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 1.2077038288116455,
+      "learning_rate": 3.935512813846994e-05,
+      "loss": 1.0831,
+      "step": 20415
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 1.9154731035232544,
+      "learning_rate": 3.935031763972462e-05,
+      "loss": 1.1491,
+      "step": 20420
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 1.4190819263458252,
+      "learning_rate": 3.9345506348426184e-05,
+      "loss": 1.0991,
+      "step": 20425
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 1.8110177516937256,
+      "learning_rate": 3.934069426484034e-05,
+      "loss": 1.3591,
+      "step": 20430
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 2.452357530593872,
+      "learning_rate": 3.9335881389232854e-05,
+      "loss": 1.1955,
+      "step": 20435
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 3.6811835765838623,
+      "learning_rate": 3.9331067721869555e-05,
+      "loss": 1.042,
+      "step": 20440
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 1.9267451763153076,
+      "learning_rate": 3.932625326301627e-05,
+      "loss": 1.0402,
+      "step": 20445
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 1.685020089149475,
+      "learning_rate": 3.9321438012938906e-05,
+      "loss": 1.1954,
+      "step": 20450
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 2.5964527130126953,
+      "learning_rate": 3.931662197190341e-05,
+      "loss": 1.069,
+      "step": 20455
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 6.929611682891846,
+      "learning_rate": 3.931180514017576e-05,
+      "loss": 1.1414,
+      "step": 20460
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 1.4876823425292969,
+      "learning_rate": 3.9306987518022e-05,
+      "loss": 1.293,
+      "step": 20465
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 1.7589577436447144,
+      "learning_rate": 3.930216910570818e-05,
+      "loss": 1.0805,
+      "step": 20470
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 3.688732862472534,
+      "learning_rate": 3.929734990350043e-05,
+      "loss": 1.2528,
+      "step": 20475
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 1.360373854637146,
+      "learning_rate": 3.929252991166491e-05,
+      "loss": 1.3237,
+      "step": 20480
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 6.932022571563721,
+      "learning_rate": 3.928770913046781e-05,
+      "loss": 1.1989,
+      "step": 20485
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 2.913423538208008,
+      "learning_rate": 3.928288756017539e-05,
+      "loss": 1.1123,
+      "step": 20490
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 1.7784669399261475,
+      "learning_rate": 3.927806520105394e-05,
+      "loss": 1.1983,
+      "step": 20495
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 1.5603896379470825,
+      "learning_rate": 3.927324205336979e-05,
+      "loss": 1.2451,
+      "step": 20500
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 3.020375967025757,
+      "learning_rate": 3.9268418117389314e-05,
+      "loss": 1.1557,
+      "step": 20505
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 2.243635416030884,
+      "learning_rate": 3.926359339337894e-05,
+      "loss": 1.1213,
+      "step": 20510
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 3.9490649700164795,
+      "learning_rate": 3.9258767881605126e-05,
+      "loss": 1.1739,
+      "step": 20515
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 2.237060308456421,
+      "learning_rate": 3.925394158233438e-05,
+      "loss": 1.2768,
+      "step": 20520
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 1.9132599830627441,
+      "learning_rate": 3.924911449583326e-05,
+      "loss": 1.0558,
+      "step": 20525
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 1.5452789068222046,
+      "learning_rate": 3.924428662236836e-05,
+      "loss": 1.3715,
+      "step": 20530
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 4.26951265335083,
+      "learning_rate": 3.923945796220632e-05,
+      "loss": 1.2972,
+      "step": 20535
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 8.05988883972168,
+      "learning_rate": 3.9234628515613806e-05,
+      "loss": 1.1404,
+      "step": 20540
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 2.191908359527588,
+      "learning_rate": 3.922979828285757e-05,
+      "loss": 1.1919,
+      "step": 20545
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 1.6250762939453125,
+      "learning_rate": 3.922496726420435e-05,
+      "loss": 1.1798,
+      "step": 20550
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 2.0135927200317383,
+      "learning_rate": 3.9220135459920984e-05,
+      "loss": 1.1462,
+      "step": 20555
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 2.279970407485962,
+      "learning_rate": 3.921530287027431e-05,
+      "loss": 1.2313,
+      "step": 20560
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 1.4251229763031006,
+      "learning_rate": 3.921046949553124e-05,
+      "loss": 1.2208,
+      "step": 20565
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 2.408446788787842,
+      "learning_rate": 3.920563533595871e-05,
+      "loss": 0.9747,
+      "step": 20570
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 4.922577381134033,
+      "learning_rate": 3.9200800391823705e-05,
+      "loss": 1.1785,
+      "step": 20575
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 1.8073464632034302,
+      "learning_rate": 3.919596466339326e-05,
+      "loss": 1.0418,
+      "step": 20580
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 1.2471339702606201,
+      "learning_rate": 3.9191128150934435e-05,
+      "loss": 1.1297,
+      "step": 20585
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 3.711719036102295,
+      "learning_rate": 3.918629085471436e-05,
+      "loss": 1.1975,
+      "step": 20590
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 4.0079498291015625,
+      "learning_rate": 3.918145277500018e-05,
+      "loss": 1.2807,
+      "step": 20595
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 1.5881859064102173,
+      "learning_rate": 3.9176613912059114e-05,
+      "loss": 1.1369,
+      "step": 20600
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 1.388296127319336,
+      "learning_rate": 3.917177426615839e-05,
+      "loss": 1.1525,
+      "step": 20605
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 1.6135921478271484,
+      "learning_rate": 3.916790198588648e-05,
+      "loss": 1.2703,
+      "step": 20610
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 1.2012652158737183,
+      "learning_rate": 3.916306093133198e-05,
+      "loss": 1.2268,
+      "step": 20615
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 1.4433081150054932,
+      "learning_rate": 3.915821909456635e-05,
+      "loss": 1.2495,
+      "step": 20620
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 1.3392014503479004,
+      "learning_rate": 3.9153376475856995e-05,
+      "loss": 1.2917,
+      "step": 20625
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 1.8551833629608154,
+      "learning_rate": 3.9148533075471364e-05,
+      "loss": 1.1313,
+      "step": 20630
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 2.474085569381714,
+      "learning_rate": 3.914368889367697e-05,
+      "loss": 1.123,
+      "step": 20635
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 1.6884278059005737,
+      "learning_rate": 3.9138843930741334e-05,
+      "loss": 1.0056,
+      "step": 20640
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 9.735162734985352,
+      "learning_rate": 3.9133998186932036e-05,
+      "loss": 1.1052,
+      "step": 20645
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 3.7487850189208984,
+      "learning_rate": 3.912915166251672e-05,
+      "loss": 1.2539,
+      "step": 20650
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 2.0177783966064453,
+      "learning_rate": 3.912430435776304e-05,
+      "loss": 1.0794,
+      "step": 20655
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 1.6745036840438843,
+      "learning_rate": 3.911945627293871e-05,
+      "loss": 1.1716,
+      "step": 20660
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 3.979721784591675,
+      "learning_rate": 3.9114607408311486e-05,
+      "loss": 1.1044,
+      "step": 20665
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 3.073629140853882,
+      "learning_rate": 3.9109757764149166e-05,
+      "loss": 1.0468,
+      "step": 20670
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 1.9026330709457397,
+      "learning_rate": 3.91049073407196e-05,
+      "loss": 1.1539,
+      "step": 20675
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 2.6179862022399902,
+      "learning_rate": 3.910005613829065e-05,
+      "loss": 1.142,
+      "step": 20680
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 1.8899461030960083,
+      "learning_rate": 3.909520415713027e-05,
+      "loss": 1.2276,
+      "step": 20685
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 5.7648444175720215,
+      "learning_rate": 3.909035139750641e-05,
+      "loss": 1.166,
+      "step": 20690
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 6.039294719696045,
+      "learning_rate": 3.908549785968708e-05,
+      "loss": 1.1015,
+      "step": 20695
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 1.34919273853302,
+      "learning_rate": 3.908064354394035e-05,
+      "loss": 1.215,
+      "step": 20700
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 2.3863883018493652,
+      "learning_rate": 3.907578845053432e-05,
+      "loss": 1.3305,
+      "step": 20705
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 1.9854397773742676,
+      "learning_rate": 3.907093257973712e-05,
+      "loss": 1.2554,
+      "step": 20710
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 2.0534327030181885,
+      "learning_rate": 3.9066075931816934e-05,
+      "loss": 1.132,
+      "step": 20715
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 4.463151454925537,
+      "learning_rate": 3.9061218507042e-05,
+      "loss": 1.1111,
+      "step": 20720
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 1.3392293453216553,
+      "learning_rate": 3.905636030568058e-05,
+      "loss": 1.0829,
+      "step": 20725
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 2.5353002548217773,
+      "learning_rate": 3.905150132800099e-05,
+      "loss": 1.1461,
+      "step": 20730
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 4.353736400604248,
+      "learning_rate": 3.90466415742716e-05,
+      "loss": 1.187,
+      "step": 20735
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 4.582551956176758,
+      "learning_rate": 3.904178104476078e-05,
+      "loss": 1.132,
+      "step": 20740
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 1.9312151670455933,
+      "learning_rate": 3.9036919739737e-05,
+      "loss": 1.1476,
+      "step": 20745
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 2.270596981048584,
+      "learning_rate": 3.9032057659468734e-05,
+      "loss": 1.0766,
+      "step": 20750
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 1.6759518384933472,
+      "learning_rate": 3.902719480422451e-05,
+      "loss": 1.0496,
+      "step": 20755
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 1.878208875656128,
+      "learning_rate": 3.902233117427289e-05,
+      "loss": 1.3187,
+      "step": 20760
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 1.0941290855407715,
+      "learning_rate": 3.9017466769882494e-05,
+      "loss": 1.1848,
+      "step": 20765
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 4.294778347015381,
+      "learning_rate": 3.901260159132198e-05,
+      "loss": 1.2239,
+      "step": 20770
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 3.6383235454559326,
+      "learning_rate": 3.900773563886004e-05,
+      "loss": 1.1626,
+      "step": 20775
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 2.137620449066162,
+      "learning_rate": 3.900286891276543e-05,
+      "loss": 1.273,
+      "step": 20780
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 3.62593150138855,
+      "learning_rate": 3.8998001413306926e-05,
+      "loss": 1.1256,
+      "step": 20785
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 3.164325475692749,
+      "learning_rate": 3.899313314075335e-05,
+      "loss": 1.1586,
+      "step": 20790
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 3.6429853439331055,
+      "learning_rate": 3.898826409537358e-05,
+      "loss": 1.1242,
+      "step": 20795
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 2.646456241607666,
+      "learning_rate": 3.898339427743652e-05,
+      "loss": 1.2348,
+      "step": 20800
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 2.920931100845337,
+      "learning_rate": 3.897852368721113e-05,
+      "loss": 1.1977,
+      "step": 20805
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 3.684735059738159,
+      "learning_rate": 3.8973652324966404e-05,
+      "loss": 1.2288,
+      "step": 20810
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 2.1889755725860596,
+      "learning_rate": 3.896878019097139e-05,
+      "loss": 1.1471,
+      "step": 20815
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 3.232788324356079,
+      "learning_rate": 3.896390728549516e-05,
+      "loss": 1.1925,
+      "step": 20820
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 1.4268032312393188,
+      "learning_rate": 3.895903360880685e-05,
+      "loss": 1.0222,
+      "step": 20825
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 2.271416187286377,
+      "learning_rate": 3.895415916117562e-05,
+      "loss": 1.0918,
+      "step": 20830
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 4.371371269226074,
+      "learning_rate": 3.894928394287068e-05,
+      "loss": 1.2376,
+      "step": 20835
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 1.4696122407913208,
+      "learning_rate": 3.89444079541613e-05,
+      "loss": 1.3061,
+      "step": 20840
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 2.775634527206421,
+      "learning_rate": 3.893953119531676e-05,
+      "loss": 1.069,
+      "step": 20845
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 1.066505789756775,
+      "learning_rate": 3.893465366660639e-05,
+      "loss": 1.2153,
+      "step": 20850
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 1.7460033893585205,
+      "learning_rate": 3.8929775368299595e-05,
+      "loss": 1.0575,
+      "step": 20855
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 1.4495829343795776,
+      "learning_rate": 3.892489630066578e-05,
+      "loss": 1.1096,
+      "step": 20860
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 2.495727062225342,
+      "learning_rate": 3.892001646397441e-05,
+      "loss": 1.1207,
+      "step": 20865
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 2.2231380939483643,
+      "learning_rate": 3.891513585849501e-05,
+      "loss": 1.1967,
+      "step": 20870
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 7.131553649902344,
+      "learning_rate": 3.891025448449711e-05,
+      "loss": 1.1515,
+      "step": 20875
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 1.6254703998565674,
+      "learning_rate": 3.890537234225033e-05,
+      "loss": 1.2101,
+      "step": 20880
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 2.034390687942505,
+      "learning_rate": 3.890048943202428e-05,
+      "loss": 1.0613,
+      "step": 20885
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 2.165940999984741,
+      "learning_rate": 3.8895605754088646e-05,
+      "loss": 1.0818,
+      "step": 20890
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 1.1313718557357788,
+      "learning_rate": 3.889072130871315e-05,
+      "loss": 1.0127,
+      "step": 20895
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 3.206895589828491,
+      "learning_rate": 3.888583609616755e-05,
+      "loss": 1.2237,
+      "step": 20900
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 4.657969951629639,
+      "learning_rate": 3.888095011672167e-05,
+      "loss": 1.1528,
+      "step": 20905
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 1.6976109743118286,
+      "learning_rate": 3.887606337064534e-05,
+      "loss": 1.1705,
+      "step": 20910
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 1.7403982877731323,
+      "learning_rate": 3.887117585820844e-05,
+      "loss": 1.1797,
+      "step": 20915
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 3.2801010608673096,
+      "learning_rate": 3.8866287579680925e-05,
+      "loss": 1.1335,
+      "step": 20920
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 2.5484225749969482,
+      "learning_rate": 3.886139853533276e-05,
+      "loss": 1.2226,
+      "step": 20925
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 2.7119576930999756,
+      "learning_rate": 3.8856508725433966e-05,
+      "loss": 1.3315,
+      "step": 20930
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 4.890354156494141,
+      "learning_rate": 3.8851618150254594e-05,
+      "loss": 1.1151,
+      "step": 20935
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 2.2012619972229004,
+      "learning_rate": 3.884672681006475e-05,
+      "loss": 1.1209,
+      "step": 20940
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 2.108077049255371,
+      "learning_rate": 3.884183470513457e-05,
+      "loss": 1.2317,
+      "step": 20945
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 2.5535576343536377,
+      "learning_rate": 3.883694183573426e-05,
+      "loss": 1.253,
+      "step": 20950
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 3.921041965484619,
+      "learning_rate": 3.883204820213403e-05,
+      "loss": 1.2734,
+      "step": 20955
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 2.2731990814208984,
+      "learning_rate": 3.882715380460416e-05,
+      "loss": 1.1832,
+      "step": 20960
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 2.7740683555603027,
+      "learning_rate": 3.882225864341494e-05,
+      "loss": 1.2319,
+      "step": 20965
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 1.0402427911758423,
+      "learning_rate": 3.8817362718836755e-05,
+      "loss": 1.3091,
+      "step": 20970
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 3.553283214569092,
+      "learning_rate": 3.8812466031139996e-05,
+      "loss": 1.0222,
+      "step": 20975
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 2.163384437561035,
+      "learning_rate": 3.8807568580595085e-05,
+      "loss": 1.1629,
+      "step": 20980
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 5.054582118988037,
+      "learning_rate": 3.8802670367472517e-05,
+      "loss": 1.1109,
+      "step": 20985
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 3.024733543395996,
+      "learning_rate": 3.879777139204281e-05,
+      "loss": 1.1245,
+      "step": 20990
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 2.468595027923584,
+      "learning_rate": 3.879287165457654e-05,
+      "loss": 1.3344,
+      "step": 20995
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 5.894863605499268,
+      "learning_rate": 3.878797115534429e-05,
+      "loss": 1.3523,
+      "step": 21000
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 1.7448843717575073,
+      "learning_rate": 3.878306989461673e-05,
+      "loss": 1.1512,
+      "step": 21005
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 2.0726964473724365,
+      "learning_rate": 3.8778167872664554e-05,
+      "loss": 1.1512,
+      "step": 21010
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 2.647890090942383,
+      "learning_rate": 3.8773265089758483e-05,
+      "loss": 1.1089,
+      "step": 21015
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 2.005052328109741,
+      "learning_rate": 3.87683615461693e-05,
+      "loss": 1.2923,
+      "step": 21020
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 2.600606918334961,
+      "learning_rate": 3.8763457242167816e-05,
+      "loss": 1.1266,
+      "step": 21025
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 1.3668138980865479,
+      "learning_rate": 3.87585521780249e-05,
+      "loss": 1.1652,
+      "step": 21030
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 2.5960161685943604,
+      "learning_rate": 3.8753646354011444e-05,
+      "loss": 1.1451,
+      "step": 21035
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 2.091399669647217,
+      "learning_rate": 3.87487397703984e-05,
+      "loss": 1.0908,
+      "step": 21040
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 1.7568382024765015,
+      "learning_rate": 3.8743832427456736e-05,
+      "loss": 1.1526,
+      "step": 21045
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 1.715632677078247,
+      "learning_rate": 3.873892432545751e-05,
+      "loss": 1.2262,
+      "step": 21050
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 2.9848005771636963,
+      "learning_rate": 3.873401546467177e-05,
+      "loss": 1.2044,
+      "step": 21055
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 1.247778058052063,
+      "learning_rate": 3.872910584537063e-05,
+      "loss": 1.1523,
+      "step": 21060
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 3.150620460510254,
+      "learning_rate": 3.872419546782524e-05,
+      "loss": 1.0884,
+      "step": 21065
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 1.8623440265655518,
+      "learning_rate": 3.8719284332306804e-05,
+      "loss": 1.2578,
+      "step": 21070
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 1.1926594972610474,
+      "learning_rate": 3.871437243908655e-05,
+      "loss": 1.1856,
+      "step": 21075
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 2.9586501121520996,
+      "learning_rate": 3.870945978843578e-05,
+      "loss": 1.0101,
+      "step": 21080
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 1.617187261581421,
+      "learning_rate": 3.8704546380625776e-05,
+      "loss": 1.1834,
+      "step": 21085
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 1.3432159423828125,
+      "learning_rate": 3.869963221592793e-05,
+      "loss": 1.2137,
+      "step": 21090
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 1.8698570728302002,
+      "learning_rate": 3.8694717294613625e-05,
+      "loss": 1.2116,
+      "step": 21095
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 2.4080424308776855,
+      "learning_rate": 3.868980161695433e-05,
+      "loss": 1.2621,
+      "step": 21100
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 4.020505905151367,
+      "learning_rate": 3.868488518322152e-05,
+      "loss": 1.106,
+      "step": 21105
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 1.8761104345321655,
+      "learning_rate": 3.8679967993686726e-05,
+      "loss": 1.1593,
+      "step": 21110
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 1.0828648805618286,
+      "learning_rate": 3.867505004862152e-05,
+      "loss": 1.1695,
+      "step": 21115
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 2.3236336708068848,
+      "learning_rate": 3.8670131348297514e-05,
+      "loss": 1.2771,
+      "step": 21120
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 2.3681468963623047,
+      "learning_rate": 3.8665211892986355e-05,
+      "loss": 1.3653,
+      "step": 21125
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 3.011075973510742,
+      "learning_rate": 3.8660291682959753e-05,
+      "loss": 1.1936,
+      "step": 21130
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 1.638193964958191,
+      "learning_rate": 3.865537071848944e-05,
+      "loss": 1.2263,
+      "step": 21135
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.8209080696105957,
+      "learning_rate": 3.86504489998472e-05,
+      "loss": 1.1663,
+      "step": 21140
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 2.627833843231201,
+      "learning_rate": 3.864552652730485e-05,
+      "loss": 1.0747,
+      "step": 21145
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 1.2440749406814575,
+      "learning_rate": 3.8640603301134245e-05,
+      "loss": 1.211,
+      "step": 21150
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 1.7105375528335571,
+      "learning_rate": 3.863567932160731e-05,
+      "loss": 1.177,
+      "step": 21155
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 1.2131013870239258,
+      "learning_rate": 3.863075458899598e-05,
+      "loss": 1.0843,
+      "step": 21160
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 5.162259101867676,
+      "learning_rate": 3.862582910357223e-05,
+      "loss": 1.2929,
+      "step": 21165
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 3.677018165588379,
+      "learning_rate": 3.862090286560811e-05,
+      "loss": 1.0206,
+      "step": 21170
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 1.5079269409179688,
+      "learning_rate": 3.861597587537568e-05,
+      "loss": 1.2481,
+      "step": 21175
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 2.0663700103759766,
+      "learning_rate": 3.861104813314705e-05,
+      "loss": 1.0672,
+      "step": 21180
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 2.495488166809082,
+      "learning_rate": 3.8606119639194394e-05,
+      "loss": 1.1044,
+      "step": 21185
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 2.68648099899292,
+      "learning_rate": 3.8601190393789885e-05,
+      "loss": 1.1525,
+      "step": 21190
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 1.7816789150238037,
+      "learning_rate": 3.8596260397205766e-05,
+      "loss": 1.2182,
+      "step": 21195
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 1.921111822128296,
+      "learning_rate": 3.859132964971432e-05,
+      "loss": 1.1855,
+      "step": 21200
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 0.9423299431800842,
+      "learning_rate": 3.8586398151587864e-05,
+      "loss": 1.2339,
+      "step": 21205
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 1.9230526685714722,
+      "learning_rate": 3.858146590309877e-05,
+      "loss": 1.1414,
+      "step": 21210
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 1.7740641832351685,
+      "learning_rate": 3.857653290451941e-05,
+      "loss": 1.3103,
+      "step": 21215
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 2.815051555633545,
+      "learning_rate": 3.857159915612227e-05,
+      "loss": 1.3497,
+      "step": 21220
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 2.79278302192688,
+      "learning_rate": 3.856666465817981e-05,
+      "loss": 1.129,
+      "step": 21225
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 1.6344680786132812,
+      "learning_rate": 3.8561729410964556e-05,
+      "loss": 1.1296,
+      "step": 21230
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 2.4964382648468018,
+      "learning_rate": 3.855679341474909e-05,
+      "loss": 1.2656,
+      "step": 21235
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 2.5263328552246094,
+      "learning_rate": 3.855185666980602e-05,
+      "loss": 1.3019,
+      "step": 21240
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 3.0418660640716553,
+      "learning_rate": 3.854691917640798e-05,
+      "loss": 1.2104,
+      "step": 21245
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 2.407768726348877,
+      "learning_rate": 3.854198093482768e-05,
+      "loss": 1.0718,
+      "step": 21250
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 3.624297857284546,
+      "learning_rate": 3.853704194533785e-05,
+      "loss": 1.2319,
+      "step": 21255
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 3.435147285461426,
+      "learning_rate": 3.8532102208211265e-05,
+      "loss": 1.1739,
+      "step": 21260
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 2.9471611976623535,
+      "learning_rate": 3.852716172372074e-05,
+      "loss": 1.1199,
+      "step": 21265
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 3.1769185066223145,
+      "learning_rate": 3.8522220492139136e-05,
+      "loss": 1.1471,
+      "step": 21270
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 3.634047269821167,
+      "learning_rate": 3.8517278513739345e-05,
+      "loss": 1.1882,
+      "step": 21275
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 3.191857099533081,
+      "learning_rate": 3.851233578879432e-05,
+      "loss": 1.1523,
+      "step": 21280
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 3.0446479320526123,
+      "learning_rate": 3.850739231757702e-05,
+      "loss": 1.1287,
+      "step": 21285
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 1.9125072956085205,
+      "learning_rate": 3.850244810036049e-05,
+      "loss": 1.0729,
+      "step": 21290
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 6.34651517868042,
+      "learning_rate": 3.849750313741779e-05,
+      "loss": 1.1506,
+      "step": 21295
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 1.6267149448394775,
+      "learning_rate": 3.8492557429022026e-05,
+      "loss": 1.0437,
+      "step": 21300
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 2.244823694229126,
+      "learning_rate": 3.848761097544633e-05,
+      "loss": 1.1806,
+      "step": 21305
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 1.7018314599990845,
+      "learning_rate": 3.8482663776963904e-05,
+      "loss": 1.1546,
+      "step": 21310
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 1.9691760540008545,
+      "learning_rate": 3.847771583384797e-05,
+      "loss": 1.089,
+      "step": 21315
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 2.011826992034912,
+      "learning_rate": 3.8472767146371805e-05,
+      "loss": 1.0998,
+      "step": 21320
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 1.590854525566101,
+      "learning_rate": 3.846781771480871e-05,
+      "loss": 1.2689,
+      "step": 21325
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 3.3987576961517334,
+      "learning_rate": 3.846286753943205e-05,
+      "loss": 1.1582,
+      "step": 21330
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 1.704407811164856,
+      "learning_rate": 3.84579166205152e-05,
+      "loss": 1.234,
+      "step": 21335
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 1.0222856998443604,
+      "learning_rate": 3.845296495833161e-05,
+      "loss": 1.2021,
+      "step": 21340
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 2.390643835067749,
+      "learning_rate": 3.844801255315474e-05,
+      "loss": 1.3623,
+      "step": 21345
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 1.4073227643966675,
+      "learning_rate": 3.844305940525812e-05,
+      "loss": 1.2204,
+      "step": 21350
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 3.146881341934204,
+      "learning_rate": 3.84381055149153e-05,
+      "loss": 1.3272,
+      "step": 21355
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 1.6763017177581787,
+      "learning_rate": 3.843315088239988e-05,
+      "loss": 1.1796,
+      "step": 21360
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 2.7298102378845215,
+      "learning_rate": 3.8428195507985505e-05,
+      "loss": 1.1095,
+      "step": 21365
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 2.3021061420440674,
+      "learning_rate": 3.842323939194584e-05,
+      "loss": 1.0336,
+      "step": 21370
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 2.4729456901550293,
+      "learning_rate": 3.841828253455463e-05,
+      "loss": 1.3022,
+      "step": 21375
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 2.751861572265625,
+      "learning_rate": 3.841332493608561e-05,
+      "loss": 1.3114,
+      "step": 21380
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 3.295260190963745,
+      "learning_rate": 3.840836659681261e-05,
+      "loss": 1.1114,
+      "step": 21385
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 4.862136363983154,
+      "learning_rate": 3.840340751700945e-05,
+      "loss": 1.1214,
+      "step": 21390
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 1.8735803365707397,
+      "learning_rate": 3.8398447696950036e-05,
+      "loss": 1.1672,
+      "step": 21395
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 1.558856725692749,
+      "learning_rate": 3.839348713690827e-05,
+      "loss": 1.2898,
+      "step": 21400
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 7.87240743637085,
+      "learning_rate": 3.838852583715814e-05,
+      "loss": 1.1136,
+      "step": 21405
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 3.504246473312378,
+      "learning_rate": 3.8383563797973634e-05,
+      "loss": 1.0798,
+      "step": 21410
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 1.1940808296203613,
+      "learning_rate": 3.837860101962882e-05,
+      "loss": 1.1356,
+      "step": 21415
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 1.6067959070205688,
+      "learning_rate": 3.8373637502397775e-05,
+      "loss": 1.1774,
+      "step": 21420
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 5.702824115753174,
+      "learning_rate": 3.836867324655463e-05,
+      "loss": 1.1839,
+      "step": 21425
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 1.5713484287261963,
+      "learning_rate": 3.8363708252373563e-05,
+      "loss": 1.1561,
+      "step": 21430
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 2.0264546871185303,
+      "learning_rate": 3.835874252012878e-05,
+      "loss": 1.2943,
+      "step": 21435
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 1.9369525909423828,
+      "learning_rate": 3.8353776050094524e-05,
+      "loss": 1.0951,
+      "step": 21440
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 1.6942447423934937,
+      "learning_rate": 3.8348808842545106e-05,
+      "loss": 1.0946,
+      "step": 21445
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 1.3976869583129883,
+      "learning_rate": 3.8343840897754845e-05,
+      "loss": 1.2452,
+      "step": 21450
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 3.386245012283325,
+      "learning_rate": 3.833887221599812e-05,
+      "loss": 1.3216,
+      "step": 21455
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 1.7361987829208374,
+      "learning_rate": 3.833390279754935e-05,
+      "loss": 1.2033,
+      "step": 21460
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 4.732967853546143,
+      "learning_rate": 3.832893264268299e-05,
+      "loss": 1.202,
+      "step": 21465
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 11.390531539916992,
+      "learning_rate": 3.8323961751673545e-05,
+      "loss": 1.1177,
+      "step": 21470
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 1.5871869325637817,
+      "learning_rate": 3.831899012479553e-05,
+      "loss": 1.1747,
+      "step": 21475
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 2.522251605987549,
+      "learning_rate": 3.8314017762323526e-05,
+      "loss": 1.0981,
+      "step": 21480
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 2.3659377098083496,
+      "learning_rate": 3.830904466453218e-05,
+      "loss": 1.1489,
+      "step": 21485
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 2.8032987117767334,
+      "learning_rate": 3.830407083169612e-05,
+      "loss": 1.1038,
+      "step": 21490
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 3.808497428894043,
+      "learning_rate": 3.829909626409006e-05,
+      "loss": 1.1027,
+      "step": 21495
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 2.006352663040161,
+      "learning_rate": 3.829412096198873e-05,
+      "loss": 1.3041,
+      "step": 21500
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 2.3574745655059814,
+      "learning_rate": 3.828914492566693e-05,
+      "loss": 1.1786,
+      "step": 21505
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 2.2047548294067383,
+      "learning_rate": 3.828416815539946e-05,
+      "loss": 1.2032,
+      "step": 21510
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 3.6130099296569824,
+      "learning_rate": 3.8279190651461195e-05,
+      "loss": 1.0937,
+      "step": 21515
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 2.250490427017212,
+      "learning_rate": 3.827421241412703e-05,
+      "loss": 1.3764,
+      "step": 21520
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 2.148017168045044,
+      "learning_rate": 3.8269233443671914e-05,
+      "loss": 1.2094,
+      "step": 21525
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 3.4997782707214355,
+      "learning_rate": 3.826425374037083e-05,
+      "loss": 1.1462,
+      "step": 21530
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 1.9769136905670166,
+      "learning_rate": 3.825927330449879e-05,
+      "loss": 1.0678,
+      "step": 21535
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 1.7285351753234863,
+      "learning_rate": 3.825429213633087e-05,
+      "loss": 1.3424,
+      "step": 21540
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 2.1386027336120605,
+      "learning_rate": 3.8249310236142175e-05,
+      "loss": 1.0702,
+      "step": 21545
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 4.701780319213867,
+      "learning_rate": 3.8244327604207856e-05,
+      "loss": 1.2914,
+      "step": 21550
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.979941725730896,
+      "learning_rate": 3.8239344240803077e-05,
+      "loss": 1.313,
+      "step": 21555
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 2.4116852283477783,
+      "learning_rate": 3.823436014620308e-05,
+      "loss": 1.1429,
+      "step": 21560
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 2.062809705734253,
+      "learning_rate": 3.822937532068314e-05,
+      "loss": 1.064,
+      "step": 21565
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.9960473775863647,
+      "learning_rate": 3.822438976451854e-05,
+      "loss": 1.3403,
+      "step": 21570
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.9557723999023438,
+      "learning_rate": 3.821940347798464e-05,
+      "loss": 1.2886,
+      "step": 21575
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 1.1672139167785645,
+      "learning_rate": 3.8214416461356825e-05,
+      "loss": 1.0831,
+      "step": 21580
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 1.0356839895248413,
+      "learning_rate": 3.8209428714910536e-05,
+      "loss": 1.2411,
+      "step": 21585
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 2.4075944423675537,
+      "learning_rate": 3.820444023892122e-05,
+      "loss": 1.3309,
+      "step": 21590
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 3.5480706691741943,
+      "learning_rate": 3.8199451033664395e-05,
+      "loss": 1.214,
+      "step": 21595
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 14.751591682434082,
+      "learning_rate": 3.8194461099415615e-05,
+      "loss": 1.1567,
+      "step": 21600
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 2.6197705268859863,
+      "learning_rate": 3.818947043645046e-05,
+      "loss": 1.3019,
+      "step": 21605
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 3.7465360164642334,
+      "learning_rate": 3.818447904504456e-05,
+      "loss": 1.0903,
+      "step": 21610
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 1.5686691999435425,
+      "learning_rate": 3.817948692547358e-05,
+      "loss": 1.1012,
+      "step": 21615
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 1.792001485824585,
+      "learning_rate": 3.8174494078013254e-05,
+      "loss": 1.2688,
+      "step": 21620
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 1.9300975799560547,
+      "learning_rate": 3.8169500502939305e-05,
+      "loss": 1.0274,
+      "step": 21625
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 2.6718950271606445,
+      "learning_rate": 3.816450620052754e-05,
+      "loss": 1.2466,
+      "step": 21630
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 1.7628939151763916,
+      "learning_rate": 3.8159511171053767e-05,
+      "loss": 1.1826,
+      "step": 21635
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 3.9780194759368896,
+      "learning_rate": 3.8154515414793874e-05,
+      "loss": 1.0958,
+      "step": 21640
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 1.881233811378479,
+      "learning_rate": 3.814951893202377e-05,
+      "loss": 1.0608,
+      "step": 21645
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 2.578082323074341,
+      "learning_rate": 3.814452172301941e-05,
+      "loss": 1.1508,
+      "step": 21650
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 4.1634721755981445,
+      "learning_rate": 3.813952378805677e-05,
+      "loss": 1.1515,
+      "step": 21655
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 2.9157748222351074,
+      "learning_rate": 3.8134525127411896e-05,
+      "loss": 1.2368,
+      "step": 21660
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 4.483683109283447,
+      "learning_rate": 3.812952574136085e-05,
+      "loss": 1.0416,
+      "step": 21665
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 2.9984755516052246,
+      "learning_rate": 3.812452563017974e-05,
+      "loss": 1.0572,
+      "step": 21670
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 1.295201301574707,
+      "learning_rate": 3.8119524794144724e-05,
+      "loss": 1.0067,
+      "step": 21675
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 2.1043081283569336,
+      "learning_rate": 3.811452323353199e-05,
+      "loss": 1.2859,
+      "step": 21680
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 1.3787553310394287,
+      "learning_rate": 3.810952094861777e-05,
+      "loss": 1.1761,
+      "step": 21685
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 1.4187425374984741,
+      "learning_rate": 3.810451793967834e-05,
+      "loss": 1.2269,
+      "step": 21690
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 4.034739017486572,
+      "learning_rate": 3.809951420699e-05,
+      "loss": 1.0818,
+      "step": 21695
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 1.0711634159088135,
+      "learning_rate": 3.809450975082911e-05,
+      "loss": 1.1358,
+      "step": 21700
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 7.229355335235596,
+      "learning_rate": 3.808950457147205e-05,
+      "loss": 1.3021,
+      "step": 21705
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 1.897261619567871,
+      "learning_rate": 3.808449866919527e-05,
+      "loss": 1.2492,
+      "step": 21710
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 2.2863614559173584,
+      "learning_rate": 3.807949204427522e-05,
+      "loss": 1.0721,
+      "step": 21715
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 2.1453750133514404,
+      "learning_rate": 3.807448469698842e-05,
+      "loss": 1.1794,
+      "step": 21720
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.794268012046814,
+      "learning_rate": 3.806947662761142e-05,
+      "loss": 1.3931,
+      "step": 21725
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 1.1213440895080566,
+      "learning_rate": 3.8064467836420815e-05,
+      "loss": 1.2736,
+      "step": 21730
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 4.008541584014893,
+      "learning_rate": 3.805945832369322e-05,
+      "loss": 1.247,
+      "step": 21735
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 1.5506445169448853,
+      "learning_rate": 3.805444808970533e-05,
+      "loss": 1.024,
+      "step": 21740
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 1.3730429410934448,
+      "learning_rate": 3.8049437134733834e-05,
+      "loss": 1.2982,
+      "step": 21745
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 1.3284574747085571,
+      "learning_rate": 3.80444254590555e-05,
+      "loss": 1.2488,
+      "step": 21750
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 3.1834280490875244,
+      "learning_rate": 3.80394130629471e-05,
+      "loss": 1.1852,
+      "step": 21755
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 1.5427137613296509,
+      "learning_rate": 3.8034399946685466e-05,
+      "loss": 1.2332,
+      "step": 21760
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 2.425147533416748,
+      "learning_rate": 3.802938611054747e-05,
+      "loss": 1.1036,
+      "step": 21765
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 16.511322021484375,
+      "learning_rate": 3.802437155481003e-05,
+      "loss": 1.0699,
+      "step": 21770
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 1.3702104091644287,
+      "learning_rate": 3.801935627975008e-05,
+      "loss": 1.2717,
+      "step": 21775
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 2.96932315826416,
+      "learning_rate": 3.8014340285644625e-05,
+      "loss": 1.2624,
+      "step": 21780
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 1.8278169631958008,
+      "learning_rate": 3.8009323572770684e-05,
+      "loss": 1.1947,
+      "step": 21785
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 1.3895366191864014,
+      "learning_rate": 3.8004306141405314e-05,
+      "loss": 1.1731,
+      "step": 21790
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 2.925675868988037,
+      "learning_rate": 3.799928799182564e-05,
+      "loss": 1.22,
+      "step": 21795
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 1.1851224899291992,
+      "learning_rate": 3.79942691243088e-05,
+      "loss": 1.0118,
+      "step": 21800
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 2.040398120880127,
+      "learning_rate": 3.7989249539131984e-05,
+      "loss": 1.0473,
+      "step": 21805
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 3.938803195953369,
+      "learning_rate": 3.798422923657241e-05,
+      "loss": 1.0397,
+      "step": 21810
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 2.930795907974243,
+      "learning_rate": 3.797920821690736e-05,
+      "loss": 1.1349,
+      "step": 21815
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 2.4224655628204346,
+      "learning_rate": 3.797418648041413e-05,
+      "loss": 1.2474,
+      "step": 21820
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 2.1420652866363525,
+      "learning_rate": 3.796916402737007e-05,
+      "loss": 1.1601,
+      "step": 21825
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 2.40444016456604,
+      "learning_rate": 3.7964140858052555e-05,
+      "loss": 1.252,
+      "step": 21830
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 3.1540682315826416,
+      "learning_rate": 3.795911697273902e-05,
+      "loss": 1.2526,
+      "step": 21835
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 1.9476728439331055,
+      "learning_rate": 3.7954092371706925e-05,
+      "loss": 1.0338,
+      "step": 21840
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 1.7248507738113403,
+      "learning_rate": 3.7949067055233774e-05,
+      "loss": 1.1222,
+      "step": 21845
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 3.3292136192321777,
+      "learning_rate": 3.7944041023597105e-05,
+      "loss": 1.1172,
+      "step": 21850
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 2.2748050689697266,
+      "learning_rate": 3.793901427707451e-05,
+      "loss": 1.1926,
+      "step": 21855
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 1.3561511039733887,
+      "learning_rate": 3.79339868159436e-05,
+      "loss": 1.2057,
+      "step": 21860
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 4.775590896606445,
+      "learning_rate": 3.7928958640482046e-05,
+      "loss": 1.112,
+      "step": 21865
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 1.4836394786834717,
+      "learning_rate": 3.792392975096754e-05,
+      "loss": 1.3004,
+      "step": 21870
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 1.8242735862731934,
+      "learning_rate": 3.791890014767783e-05,
+      "loss": 1.3185,
+      "step": 21875
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 3.001962661743164,
+      "learning_rate": 3.79138698308907e-05,
+      "loss": 1.082,
+      "step": 21880
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 1.9447215795516968,
+      "learning_rate": 3.790883880088396e-05,
+      "loss": 1.3185,
+      "step": 21885
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 2.3919167518615723,
+      "learning_rate": 3.790380705793547e-05,
+      "loss": 1.2078,
+      "step": 21890
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 6.295918941497803,
+      "learning_rate": 3.789877460232313e-05,
+      "loss": 0.922,
+      "step": 21895
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 2.696068525314331,
+      "learning_rate": 3.789374143432487e-05,
+      "loss": 1.3105,
+      "step": 21900
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 3.568495035171509,
+      "learning_rate": 3.788870755421867e-05,
+      "loss": 1.128,
+      "step": 21905
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 1.2198554277420044,
+      "learning_rate": 3.7883672962282555e-05,
+      "loss": 1.2149,
+      "step": 21910
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 5.560957908630371,
+      "learning_rate": 3.787863765879458e-05,
+      "loss": 1.1389,
+      "step": 21915
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 3.2426624298095703,
+      "learning_rate": 3.787360164403283e-05,
+      "loss": 1.0907,
+      "step": 21920
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 2.8648648262023926,
+      "learning_rate": 3.786856491827544e-05,
+      "loss": 0.966,
+      "step": 21925
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 1.9964818954467773,
+      "learning_rate": 3.786352748180059e-05,
+      "loss": 1.2,
+      "step": 21930
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 1.6703797578811646,
+      "learning_rate": 3.7858489334886477e-05,
+      "loss": 1.2223,
+      "step": 21935
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 2.2925643920898438,
+      "learning_rate": 3.785345047781137e-05,
+      "loss": 1.3513,
+      "step": 21940
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 1.0569705963134766,
+      "learning_rate": 3.784841091085356e-05,
+      "loss": 1.1399,
+      "step": 21945
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 3.3244879245758057,
+      "learning_rate": 3.784337063429136e-05,
+      "loss": 1.3942,
+      "step": 21950
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 3.7458391189575195,
+      "learning_rate": 3.783832964840316e-05,
+      "loss": 1.1377,
+      "step": 21955
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 1.8675755262374878,
+      "learning_rate": 3.7833287953467354e-05,
+      "loss": 1.249,
+      "step": 21960
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 2.4439284801483154,
+      "learning_rate": 3.782824554976239e-05,
+      "loss": 1.2981,
+      "step": 21965
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 5.8578877449035645,
+      "learning_rate": 3.7823202437566764e-05,
+      "loss": 1.2135,
+      "step": 21970
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 1.4064592123031616,
+      "learning_rate": 3.7818158617158995e-05,
+      "loss": 1.3478,
+      "step": 21975
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 2.7772834300994873,
+      "learning_rate": 3.781311408881765e-05,
+      "loss": 1.2668,
+      "step": 21980
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 2.9502522945404053,
+      "learning_rate": 3.7808068852821334e-05,
+      "loss": 1.1482,
+      "step": 21985
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 4.143762588500977,
+      "learning_rate": 3.780302290944868e-05,
+      "loss": 1.1882,
+      "step": 21990
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 5.523019313812256,
+      "learning_rate": 3.7797976258978386e-05,
+      "loss": 1.3205,
+      "step": 21995
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 1.1239092350006104,
+      "learning_rate": 3.779292890168916e-05,
+      "loss": 1.3228,
+      "step": 22000
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 2.243865966796875,
+      "learning_rate": 3.7787880837859767e-05,
+      "loss": 1.3075,
+      "step": 22005
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 2.5616250038146973,
+      "learning_rate": 3.778283206776902e-05,
+      "loss": 1.2133,
+      "step": 22010
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 3.7556142807006836,
+      "learning_rate": 3.777778259169574e-05,
+      "loss": 1.1166,
+      "step": 22015
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 2.411454439163208,
+      "learning_rate": 3.77727324099188e-05,
+      "loss": 1.2067,
+      "step": 22020
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 7.664334297180176,
+      "learning_rate": 3.776768152271713e-05,
+      "loss": 1.0729,
+      "step": 22025
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 1.7373452186584473,
+      "learning_rate": 3.776262993036969e-05,
+      "loss": 1.1381,
+      "step": 22030
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 1.6284407377243042,
+      "learning_rate": 3.7757577633155464e-05,
+      "loss": 1.1592,
+      "step": 22035
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 1.5111100673675537,
+      "learning_rate": 3.7752524631353484e-05,
+      "loss": 1.1819,
+      "step": 22040
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 2.4430501461029053,
+      "learning_rate": 3.7747470925242815e-05,
+      "loss": 1.3295,
+      "step": 22045
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 1.4049407243728638,
+      "learning_rate": 3.774241651510258e-05,
+      "loss": 1.1802,
+      "step": 22050
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 1.4053436517715454,
+      "learning_rate": 3.773736140121193e-05,
+      "loss": 1.2387,
+      "step": 22055
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 1.9086112976074219,
+      "learning_rate": 3.7732305583850044e-05,
+      "loss": 1.1988,
+      "step": 22060
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 2.53025221824646,
+      "learning_rate": 3.772724906329616e-05,
+      "loss": 1.1728,
+      "step": 22065
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 2.8648312091827393,
+      "learning_rate": 3.772219183982953e-05,
+      "loss": 1.0513,
+      "step": 22070
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 1.3704460859298706,
+      "learning_rate": 3.771713391372949e-05,
+      "loss": 1.2856,
+      "step": 22075
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 5.734248161315918,
+      "learning_rate": 3.771207528527534e-05,
+      "loss": 1.1455,
+      "step": 22080
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 2.4370834827423096,
+      "learning_rate": 3.77070159547465e-05,
+      "loss": 1.1426,
+      "step": 22085
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 13.942612648010254,
+      "learning_rate": 3.770195592242237e-05,
+      "loss": 1.2256,
+      "step": 22090
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 2.1054158210754395,
+      "learning_rate": 3.769689518858242e-05,
+      "loss": 1.2302,
+      "step": 22095
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 2.0124964714050293,
+      "learning_rate": 3.7691833753506145e-05,
+      "loss": 1.2412,
+      "step": 22100
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 2.5039613246917725,
+      "learning_rate": 3.7686771617473094e-05,
+      "loss": 1.0837,
+      "step": 22105
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 1.1294031143188477,
+      "learning_rate": 3.768170878076283e-05,
+      "loss": 1.2652,
+      "step": 22110
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 1.8122518062591553,
+      "learning_rate": 3.767664524365496e-05,
+      "loss": 1.3259,
+      "step": 22115
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 2.7276618480682373,
+      "learning_rate": 3.767158100642916e-05,
+      "loss": 1.0632,
+      "step": 22120
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 6.643666744232178,
+      "learning_rate": 3.766651606936511e-05,
+      "loss": 1.1819,
+      "step": 22125
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 1.3447577953338623,
+      "learning_rate": 3.7661450432742534e-05,
+      "loss": 1.3445,
+      "step": 22130
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 4.023565769195557,
+      "learning_rate": 3.7656384096841226e-05,
+      "loss": 1.1085,
+      "step": 22135
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 2.0256614685058594,
+      "learning_rate": 3.7651317061940975e-05,
+      "loss": 1.2472,
+      "step": 22140
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 1.7763872146606445,
+      "learning_rate": 3.764624932832163e-05,
+      "loss": 1.1751,
+      "step": 22145
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 2.526588201522827,
+      "learning_rate": 3.7641180896263085e-05,
+      "loss": 1.0481,
+      "step": 22150
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 2.3050858974456787,
+      "learning_rate": 3.7636111766045254e-05,
+      "loss": 1.1158,
+      "step": 22155
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 3.152416944503784,
+      "learning_rate": 3.76310419379481e-05,
+      "loss": 1.188,
+      "step": 22160
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 2.187168598175049,
+      "learning_rate": 3.7625971412251636e-05,
+      "loss": 1.2462,
+      "step": 22165
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 1.528957486152649,
+      "learning_rate": 3.76209001892359e-05,
+      "loss": 1.0216,
+      "step": 22170
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 2.985142230987549,
+      "learning_rate": 3.7615828269180955e-05,
+      "loss": 1.1851,
+      "step": 22175
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 1.8914645910263062,
+      "learning_rate": 3.7610755652366933e-05,
+      "loss": 1.1384,
+      "step": 22180
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 1.485687017440796,
+      "learning_rate": 3.7605682339073986e-05,
+      "loss": 1.1725,
+      "step": 22185
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 2.7884113788604736,
+      "learning_rate": 3.76006083295823e-05,
+      "loss": 1.0369,
+      "step": 22190
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 2.6795337200164795,
+      "learning_rate": 3.759553362417212e-05,
+      "loss": 1.2506,
+      "step": 22195
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 3.732194662094116,
+      "learning_rate": 3.759045822312371e-05,
+      "loss": 1.2024,
+      "step": 22200
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 2.955843210220337,
+      "learning_rate": 3.758538212671738e-05,
+      "loss": 1.092,
+      "step": 22205
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 2.543309450149536,
+      "learning_rate": 3.7580305335233466e-05,
+      "loss": 1.1417,
+      "step": 22210
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 1.1416964530944824,
+      "learning_rate": 3.757522784895238e-05,
+      "loss": 1.2911,
+      "step": 22215
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 2.8476202487945557,
+      "learning_rate": 3.757014966815452e-05,
+      "loss": 1.1666,
+      "step": 22220
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 3.7398064136505127,
+      "learning_rate": 3.7565070793120366e-05,
+      "loss": 1.2501,
+      "step": 22225
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 1.3023401498794556,
+      "learning_rate": 3.755999122413041e-05,
+      "loss": 1.0223,
+      "step": 22230
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 3.140843391418457,
+      "learning_rate": 3.7554910961465204e-05,
+      "loss": 1.1367,
+      "step": 22235
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 1.410656452178955,
+      "learning_rate": 3.75498300054053e-05,
+      "loss": 1.227,
+      "step": 22240
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 3.442096710205078,
+      "learning_rate": 3.754474835623134e-05,
+      "loss": 1.0639,
+      "step": 22245
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 2.1821417808532715,
+      "learning_rate": 3.753966601422396e-05,
+      "loss": 1.2744,
+      "step": 22250
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 1.2946699857711792,
+      "learning_rate": 3.753458297966387e-05,
+      "loss": 1.1973,
+      "step": 22255
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 1.561705231666565,
+      "learning_rate": 3.752949925283178e-05,
+      "loss": 1.1897,
+      "step": 22260
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 1.594902515411377,
+      "learning_rate": 3.7524414834008475e-05,
+      "loss": 1.1862,
+      "step": 22265
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 2.135448932647705,
+      "learning_rate": 3.751932972347476e-05,
+      "loss": 1.2566,
+      "step": 22270
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 4.568878650665283,
+      "learning_rate": 3.751424392151147e-05,
+      "loss": 0.9779,
+      "step": 22275
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 1.9273889064788818,
+      "learning_rate": 3.75091574283995e-05,
+      "loss": 1.2662,
+      "step": 22280
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 2.1159157752990723,
+      "learning_rate": 3.750407024441977e-05,
+      "loss": 1.1524,
+      "step": 22285
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 2.8787682056427,
+      "learning_rate": 3.7498982369853235e-05,
+      "loss": 1.1924,
+      "step": 22290
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 3.1426780223846436,
+      "learning_rate": 3.7493893804980897e-05,
+      "loss": 0.9211,
+      "step": 22295
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 2.8271982669830322,
+      "learning_rate": 3.748880455008379e-05,
+      "loss": 1.3036,
+      "step": 22300
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 3.4877443313598633,
+      "learning_rate": 3.7483714605442996e-05,
+      "loss": 1.1612,
+      "step": 22305
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 1.2759697437286377,
+      "learning_rate": 3.747862397133961e-05,
+      "loss": 1.2321,
+      "step": 22310
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 2.5716593265533447,
+      "learning_rate": 3.74735326480548e-05,
+      "loss": 1.0509,
+      "step": 22315
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 1.8768802881240845,
+      "learning_rate": 3.7468440635869746e-05,
+      "loss": 1.2344,
+      "step": 22320
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 3.4969723224639893,
+      "learning_rate": 3.746334793506567e-05,
+      "loss": 1.1697,
+      "step": 22325
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 2.126424551010132,
+      "learning_rate": 3.7458254545923855e-05,
+      "loss": 1.2722,
+      "step": 22330
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 2.6972811222076416,
+      "learning_rate": 3.745316046872558e-05,
+      "loss": 1.3685,
+      "step": 22335
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 1.37725830078125,
+      "learning_rate": 3.74480657037522e-05,
+      "loss": 1.2091,
+      "step": 22340
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 1.036205768585205,
+      "learning_rate": 3.7442970251285095e-05,
+      "loss": 1.3278,
+      "step": 22345
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 1.2511407136917114,
+      "learning_rate": 3.743787411160567e-05,
+      "loss": 1.1185,
+      "step": 22350
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 2.403078079223633,
+      "learning_rate": 3.7432777284995384e-05,
+      "loss": 1.187,
+      "step": 22355
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 2.58774471282959,
+      "learning_rate": 3.742767977173573e-05,
+      "loss": 1.2784,
+      "step": 22360
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 1.8022263050079346,
+      "learning_rate": 3.742258157210825e-05,
+      "loss": 1.1472,
+      "step": 22365
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 1.9385359287261963,
+      "learning_rate": 3.74174826863945e-05,
+      "loss": 1.0361,
+      "step": 22370
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 1.3857041597366333,
+      "learning_rate": 3.741238311487609e-05,
+      "loss": 1.2892,
+      "step": 22375
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 1.456871747970581,
+      "learning_rate": 3.740728285783466e-05,
+      "loss": 1.251,
+      "step": 22380
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 1.2048922777175903,
+      "learning_rate": 3.740218191555189e-05,
+      "loss": 1.2085,
+      "step": 22385
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 1.9001728296279907,
+      "learning_rate": 3.7397080288309504e-05,
+      "loss": 1.1827,
+      "step": 22390
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 1.7842788696289062,
+      "learning_rate": 3.739197797638927e-05,
+      "loss": 1.1712,
+      "step": 22395
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 2.8742237091064453,
+      "learning_rate": 3.7386874980072965e-05,
+      "loss": 1.346,
+      "step": 22400
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.9042024612426758,
+      "learning_rate": 3.7381771299642434e-05,
+      "loss": 1.1113,
+      "step": 22405
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 2.389392852783203,
+      "learning_rate": 3.737666693537954e-05,
+      "loss": 1.2051,
+      "step": 22410
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 1.8585515022277832,
+      "learning_rate": 3.73715618875662e-05,
+      "loss": 1.0961,
+      "step": 22415
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 1.445894479751587,
+      "learning_rate": 3.736645615648435e-05,
+      "loss": 1.205,
+      "step": 22420
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 1.4859192371368408,
+      "learning_rate": 3.736134974241599e-05,
+      "loss": 1.2583,
+      "step": 22425
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 1.2996731996536255,
+      "learning_rate": 3.735624264564313e-05,
+      "loss": 1.1439,
+      "step": 22430
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 2.0021395683288574,
+      "learning_rate": 3.735113486644783e-05,
+      "loss": 1.153,
+      "step": 22435
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 1.7902204990386963,
+      "learning_rate": 3.734602640511219e-05,
+      "loss": 1.0953,
+      "step": 22440
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 2.417541027069092,
+      "learning_rate": 3.734091726191834e-05,
+      "loss": 1.2148,
+      "step": 22445
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 3.7760798931121826,
+      "learning_rate": 3.7335807437148454e-05,
+      "loss": 1.1157,
+      "step": 22450
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 1.5918861627578735,
+      "learning_rate": 3.733069693108476e-05,
+      "loss": 1.2129,
+      "step": 22455
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.9340783953666687,
+      "learning_rate": 3.732558574400948e-05,
+      "loss": 1.2252,
+      "step": 22460
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 1.9744924306869507,
+      "learning_rate": 3.732047387620492e-05,
+      "loss": 1.1734,
+      "step": 22465
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 1.9062458276748657,
+      "learning_rate": 3.731536132795339e-05,
+      "loss": 1.1843,
+      "step": 22470
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 2.320115566253662,
+      "learning_rate": 3.731024809953726e-05,
+      "loss": 1.0732,
+      "step": 22475
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 2.1954009532928467,
+      "learning_rate": 3.7305134191238914e-05,
+      "loss": 1.195,
+      "step": 22480
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 2.1128578186035156,
+      "learning_rate": 3.73000196033408e-05,
+      "loss": 1.1004,
+      "step": 22485
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 2.868645191192627,
+      "learning_rate": 3.729490433612539e-05,
+      "loss": 1.1917,
+      "step": 22490
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 4.787483215332031,
+      "learning_rate": 3.7289788389875195e-05,
+      "loss": 1.1271,
+      "step": 22495
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 1.8350549936294556,
+      "learning_rate": 3.7284671764872764e-05,
+      "loss": 1.1971,
+      "step": 22500
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 1.9952346086502075,
+      "learning_rate": 3.7279554461400675e-05,
+      "loss": 1.2058,
+      "step": 22505
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 2.3155946731567383,
+      "learning_rate": 3.727443647974156e-05,
+      "loss": 1.0397,
+      "step": 22510
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 2.454383611679077,
+      "learning_rate": 3.726931782017807e-05,
+      "loss": 1.2128,
+      "step": 22515
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 1.4196258783340454,
+      "learning_rate": 3.726419848299292e-05,
+      "loss": 1.3271,
+      "step": 22520
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 4.044848918914795,
+      "learning_rate": 3.725907846846883e-05,
+      "loss": 1.1593,
+      "step": 22525
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 1.5444008111953735,
+      "learning_rate": 3.725395777688858e-05,
+      "loss": 1.165,
+      "step": 22530
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 2.297614097595215,
+      "learning_rate": 3.7248836408534975e-05,
+      "loss": 1.259,
+      "step": 22535
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 1.7064695358276367,
+      "learning_rate": 3.724371436369087e-05,
+      "loss": 1.2259,
+      "step": 22540
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 1.426712155342102,
+      "learning_rate": 3.7238591642639155e-05,
+      "loss": 1.1047,
+      "step": 22545
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 2.8516054153442383,
+      "learning_rate": 3.723346824566273e-05,
+      "loss": 1.1642,
+      "step": 22550
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 1.5631628036499023,
+      "learning_rate": 3.722834417304459e-05,
+      "loss": 1.1772,
+      "step": 22555
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 1.1760963201522827,
+      "learning_rate": 3.7223219425067694e-05,
+      "loss": 1.2157,
+      "step": 22560
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 1.5245758295059204,
+      "learning_rate": 3.7218094002015104e-05,
+      "loss": 1.0718,
+      "step": 22565
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 5.258598327636719,
+      "learning_rate": 3.7212967904169874e-05,
+      "loss": 1.0714,
+      "step": 22570
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 0.8288945555686951,
+      "learning_rate": 3.720784113181513e-05,
+      "loss": 1.1171,
+      "step": 22575
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 2.472649097442627,
+      "learning_rate": 3.7202713685234006e-05,
+      "loss": 1.2317,
+      "step": 22580
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 6.6396565437316895,
+      "learning_rate": 3.719758556470969e-05,
+      "loss": 1.1746,
+      "step": 22585
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 1.3930727243423462,
+      "learning_rate": 3.71924567705254e-05,
+      "loss": 1.2441,
+      "step": 22590
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 1.635603666305542,
+      "learning_rate": 3.71873273029644e-05,
+      "loss": 1.4045,
+      "step": 22595
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 2.988915205001831,
+      "learning_rate": 3.718219716230998e-05,
+      "loss": 1.2706,
+      "step": 22600
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 1.915685772895813,
+      "learning_rate": 3.717706634884547e-05,
+      "loss": 1.3597,
+      "step": 22605
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 2.3813059329986572,
+      "learning_rate": 3.7171934862854244e-05,
+      "loss": 1.2544,
+      "step": 22610
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 3.367213010787964,
+      "learning_rate": 3.716680270461971e-05,
+      "loss": 1.2843,
+      "step": 22615
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 0.9341633319854736,
+      "learning_rate": 3.7161669874425304e-05,
+      "loss": 1.1621,
+      "step": 22620
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 2.1656148433685303,
+      "learning_rate": 3.715653637255452e-05,
+      "loss": 1.0864,
+      "step": 22625
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 2.8741602897644043,
+      "learning_rate": 3.715140219929086e-05,
+      "loss": 1.331,
+      "step": 22630
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 3.68282151222229,
+      "learning_rate": 3.714626735491789e-05,
+      "loss": 1.2079,
+      "step": 22635
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 1.4136219024658203,
+      "learning_rate": 3.71411318397192e-05,
+      "loss": 1.168,
+      "step": 22640
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 2.178769588470459,
+      "learning_rate": 3.7135995653978415e-05,
+      "loss": 1.2211,
+      "step": 22645
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 2.099884033203125,
+      "learning_rate": 3.713085879797921e-05,
+      "loss": 1.2584,
+      "step": 22650
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 2.237651824951172,
+      "learning_rate": 3.7125721272005285e-05,
+      "loss": 1.1906,
+      "step": 22655
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 5.40058708190918,
+      "learning_rate": 3.7120583076340374e-05,
+      "loss": 1.177,
+      "step": 22660
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 3.936767578125,
+      "learning_rate": 3.711544421126826e-05,
+      "loss": 1.1745,
+      "step": 22665
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 9.406877517700195,
+      "learning_rate": 3.711030467707275e-05,
+      "loss": 1.1282,
+      "step": 22670
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 4.18208122253418,
+      "learning_rate": 3.7105164474037706e-05,
+      "loss": 1.3561,
+      "step": 22675
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 1.8916975259780884,
+      "learning_rate": 3.710002360244701e-05,
+      "loss": 1.2032,
+      "step": 22680
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 1.6227880716323853,
+      "learning_rate": 3.709488206258459e-05,
+      "loss": 1.1442,
+      "step": 22685
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 2.2588720321655273,
+      "learning_rate": 3.7089739854734406e-05,
+      "loss": 1.1536,
+      "step": 22690
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 2.82357120513916,
+      "learning_rate": 3.708459697918045e-05,
+      "loss": 1.1142,
+      "step": 22695
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 2.484433650970459,
+      "learning_rate": 3.707945343620677e-05,
+      "loss": 1.2712,
+      "step": 22700
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 3.071495771408081,
+      "learning_rate": 3.7074309226097424e-05,
+      "loss": 1.0004,
+      "step": 22705
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 1.225174069404602,
+      "learning_rate": 3.706916434913654e-05,
+      "loss": 0.9823,
+      "step": 22710
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 2.9054837226867676,
+      "learning_rate": 3.7064018805608244e-05,
+      "loss": 1.1946,
+      "step": 22715
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 2.1783711910247803,
+      "learning_rate": 3.705887259579674e-05,
+      "loss": 1.1034,
+      "step": 22720
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 7.1069016456604,
+      "learning_rate": 3.7053725719986234e-05,
+      "loss": 1.2409,
+      "step": 22725
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 1.7012306451797485,
+      "learning_rate": 3.704960774000956e-05,
+      "loss": 1.0688,
+      "step": 22730
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 2.2616653442382812,
+      "learning_rate": 3.704445966611721e-05,
+      "loss": 1.2547,
+      "step": 22735
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.8807179927825928,
+      "learning_rate": 3.703931092702188e-05,
+      "loss": 1.0968,
+      "step": 22740
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 3.9191973209381104,
+      "learning_rate": 3.7034161523007905e-05,
+      "loss": 1.008,
+      "step": 22745
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 2.1995856761932373,
+      "learning_rate": 3.70290114543597e-05,
+      "loss": 1.1851,
+      "step": 22750
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 2.6970908641815186,
+      "learning_rate": 3.70238607213617e-05,
+      "loss": 1.1741,
+      "step": 22755
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 1.7191367149353027,
+      "learning_rate": 3.7018709324298364e-05,
+      "loss": 1.3658,
+      "step": 22760
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 2.820333242416382,
+      "learning_rate": 3.701355726345421e-05,
+      "loss": 0.9673,
+      "step": 22765
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 3.5425102710723877,
+      "learning_rate": 3.700840453911378e-05,
+      "loss": 1.2862,
+      "step": 22770
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 2.4167778491973877,
+      "learning_rate": 3.700325115156165e-05,
+      "loss": 1.1087,
+      "step": 22775
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 1.7471559047698975,
+      "learning_rate": 3.6998097101082424e-05,
+      "loss": 1.2492,
+      "step": 22780
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 4.646815776824951,
+      "learning_rate": 3.6992942387960766e-05,
+      "loss": 1.2538,
+      "step": 22785
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 1.7695302963256836,
+      "learning_rate": 3.698778701248137e-05,
+      "loss": 1.0982,
+      "step": 22790
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 3.9096667766571045,
+      "learning_rate": 3.698263097492896e-05,
+      "loss": 1.1196,
+      "step": 22795
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 2.3695907592773438,
+      "learning_rate": 3.69774742755883e-05,
+      "loss": 0.9765,
+      "step": 22800
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 2.3105177879333496,
+      "learning_rate": 3.697231691474419e-05,
+      "loss": 1.1265,
+      "step": 22805
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 8.283111572265625,
+      "learning_rate": 3.6967158892681456e-05,
+      "loss": 1.1517,
+      "step": 22810
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 1.0593783855438232,
+      "learning_rate": 3.696200020968497e-05,
+      "loss": 1.2822,
+      "step": 22815
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 2.067911386489868,
+      "learning_rate": 3.695684086603964e-05,
+      "loss": 1.1125,
+      "step": 22820
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 7.895202159881592,
+      "learning_rate": 3.695168086203044e-05,
+      "loss": 1.1565,
+      "step": 22825
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 2.0913262367248535,
+      "learning_rate": 3.694652019794231e-05,
+      "loss": 1.159,
+      "step": 22830
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 2.38482403755188,
+      "learning_rate": 3.6941358874060295e-05,
+      "loss": 1.2542,
+      "step": 22835
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 3.0547502040863037,
+      "learning_rate": 3.693619689066944e-05,
+      "loss": 1.3171,
+      "step": 22840
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 4.048938751220703,
+      "learning_rate": 3.6931034248054834e-05,
+      "loss": 1.0786,
+      "step": 22845
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 4.187254428863525,
+      "learning_rate": 3.6925870946501615e-05,
+      "loss": 1.1805,
+      "step": 22850
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 2.219910144805908,
+      "learning_rate": 3.692070698629493e-05,
+      "loss": 1.2842,
+      "step": 22855
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 3.7236175537109375,
+      "learning_rate": 3.6915542367720005e-05,
+      "loss": 1.2186,
+      "step": 22860
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 3.3912012577056885,
+      "learning_rate": 3.691037709106205e-05,
+      "loss": 1.0233,
+      "step": 22865
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 3.296980857849121,
+      "learning_rate": 3.6905211156606347e-05,
+      "loss": 1.2314,
+      "step": 22870
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 2.4683995246887207,
+      "learning_rate": 3.690004456463821e-05,
+      "loss": 1.3122,
+      "step": 22875
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 2.9964542388916016,
+      "learning_rate": 3.689487731544298e-05,
+      "loss": 1.2645,
+      "step": 22880
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 2.661977529525757,
+      "learning_rate": 3.688970940930604e-05,
+      "loss": 0.9723,
+      "step": 22885
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 1.609717607498169,
+      "learning_rate": 3.6884540846512806e-05,
+      "loss": 1.0939,
+      "step": 22890
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 2.814666271209717,
+      "learning_rate": 3.687937162734874e-05,
+      "loss": 1.2832,
+      "step": 22895
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 6.085357666015625,
+      "learning_rate": 3.687420175209933e-05,
+      "loss": 1.3523,
+      "step": 22900
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 2.151413679122925,
+      "learning_rate": 3.686903122105009e-05,
+      "loss": 1.0956,
+      "step": 22905
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 1.2115709781646729,
+      "learning_rate": 3.686386003448659e-05,
+      "loss": 0.9451,
+      "step": 22910
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 1.7795747518539429,
+      "learning_rate": 3.685868819269444e-05,
+      "loss": 1.272,
+      "step": 22915
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 1.4842565059661865,
+      "learning_rate": 3.6853515695959264e-05,
+      "loss": 1.3025,
+      "step": 22920
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 1.450868010520935,
+      "learning_rate": 3.684834254456674e-05,
+      "loss": 1.2952,
+      "step": 22925
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 2.6749181747436523,
+      "learning_rate": 3.684316873880257e-05,
+      "loss": 1.2248,
+      "step": 22930
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 1.4082355499267578,
+      "learning_rate": 3.6837994278952506e-05,
+      "loss": 1.2363,
+      "step": 22935
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 1.7880902290344238,
+      "learning_rate": 3.6832819165302325e-05,
+      "loss": 1.1894,
+      "step": 22940
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 2.6306281089782715,
+      "learning_rate": 3.682764339813783e-05,
+      "loss": 1.1545,
+      "step": 22945
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 4.578315734863281,
+      "learning_rate": 3.682246697774489e-05,
+      "loss": 1.0853,
+      "step": 22950
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 1.6041889190673828,
+      "learning_rate": 3.6817289904409394e-05,
+      "loss": 1.2719,
+      "step": 22955
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 4.389896869659424,
+      "learning_rate": 3.6812112178417244e-05,
+      "loss": 1.12,
+      "step": 22960
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 3.254216432571411,
+      "learning_rate": 3.6806933800054425e-05,
+      "loss": 1.2594,
+      "step": 22965
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 1.4171181917190552,
+      "learning_rate": 3.680175476960693e-05,
+      "loss": 1.2488,
+      "step": 22970
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 8.962324142456055,
+      "learning_rate": 3.679657508736078e-05,
+      "loss": 1.1462,
+      "step": 22975
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 4.2052321434021,
+      "learning_rate": 3.6791394753602054e-05,
+      "loss": 1.241,
+      "step": 22980
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 1.6290329694747925,
+      "learning_rate": 3.678621376861685e-05,
+      "loss": 1.2826,
+      "step": 22985
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 2.770118236541748,
+      "learning_rate": 3.678103213269131e-05,
+      "loss": 1.3042,
+      "step": 22990
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 2.3370397090911865,
+      "learning_rate": 3.6775849846111614e-05,
+      "loss": 1.1647,
+      "step": 22995
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 2.9405107498168945,
+      "learning_rate": 3.677066690916397e-05,
+      "loss": 1.231,
+      "step": 23000
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 3.535926580429077,
+      "learning_rate": 3.6765483322134625e-05,
+      "loss": 1.1303,
+      "step": 23005
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 1.9250695705413818,
+      "learning_rate": 3.676029908530987e-05,
+      "loss": 1.2256,
+      "step": 23010
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 1.7612509727478027,
+      "learning_rate": 3.6755114198976016e-05,
+      "loss": 1.061,
+      "step": 23015
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 2.3679986000061035,
+      "learning_rate": 3.674992866341943e-05,
+      "loss": 1.1068,
+      "step": 23020
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 1.4026302099227905,
+      "learning_rate": 3.6744742478926485e-05,
+      "loss": 1.1352,
+      "step": 23025
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 2.6781442165374756,
+      "learning_rate": 3.673955564578364e-05,
+      "loss": 1.3133,
+      "step": 23030
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 3.6922409534454346,
+      "learning_rate": 3.673436816427732e-05,
+      "loss": 1.1815,
+      "step": 23035
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 1.3594402074813843,
+      "learning_rate": 3.6729180034694055e-05,
+      "loss": 1.2121,
+      "step": 23040
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 2.578176736831665,
+      "learning_rate": 3.672399125732036e-05,
+      "loss": 1.2374,
+      "step": 23045
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 2.196401357650757,
+      "learning_rate": 3.671880183244282e-05,
+      "loss": 1.2618,
+      "step": 23050
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 1.1504014730453491,
+      "learning_rate": 3.671361176034803e-05,
+      "loss": 1.2077,
+      "step": 23055
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 1.4980058670043945,
+      "learning_rate": 3.670842104132265e-05,
+      "loss": 1.1627,
+      "step": 23060
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 2.5145270824432373,
+      "learning_rate": 3.670322967565333e-05,
+      "loss": 1.2164,
+      "step": 23065
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 1.6475614309310913,
+      "learning_rate": 3.6698037663626815e-05,
+      "loss": 1.1775,
+      "step": 23070
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 1.1627910137176514,
+      "learning_rate": 3.669284500552983e-05,
+      "loss": 1.2284,
+      "step": 23075
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 2.22105073928833,
+      "learning_rate": 3.668765170164917e-05,
+      "loss": 1.1692,
+      "step": 23080
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 1.5115783214569092,
+      "learning_rate": 3.668245775227166e-05,
+      "loss": 1.0566,
+      "step": 23085
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 1.735473394393921,
+      "learning_rate": 3.667726315768415e-05,
+      "loss": 1.0803,
+      "step": 23090
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 2.8493192195892334,
+      "learning_rate": 3.667206791817354e-05,
+      "loss": 1.2563,
+      "step": 23095
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 2.09541916847229,
+      "learning_rate": 3.6666872034026746e-05,
+      "loss": 1.2383,
+      "step": 23100
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 2.5542640686035156,
+      "learning_rate": 3.666167550553073e-05,
+      "loss": 1.1892,
+      "step": 23105
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 3.2478058338165283,
+      "learning_rate": 3.66564783329725e-05,
+      "loss": 1.3445,
+      "step": 23110
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 3.8739407062530518,
+      "learning_rate": 3.665128051663909e-05,
+      "loss": 1.2336,
+      "step": 23115
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 1.4774131774902344,
+      "learning_rate": 3.664608205681757e-05,
+      "loss": 1.0818,
+      "step": 23120
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 2.790187120437622,
+      "learning_rate": 3.664088295379505e-05,
+      "loss": 1.1317,
+      "step": 23125
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 5.5581278800964355,
+      "learning_rate": 3.663568320785865e-05,
+      "loss": 1.2463,
+      "step": 23130
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 1.6298890113830566,
+      "learning_rate": 3.6630482819295564e-05,
+      "loss": 1.1815,
+      "step": 23135
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 2.767247200012207,
+      "learning_rate": 3.662528178839301e-05,
+      "loss": 1.1912,
+      "step": 23140
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 2.707919120788574,
+      "learning_rate": 3.6620080115438214e-05,
+      "loss": 1.1293,
+      "step": 23145
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 4.199367046356201,
+      "learning_rate": 3.661487780071848e-05,
+      "loss": 1.0369,
+      "step": 23150
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 0.9766875505447388,
+      "learning_rate": 3.6609674844521113e-05,
+      "loss": 1.2359,
+      "step": 23155
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 2.218275308609009,
+      "learning_rate": 3.6604471247133476e-05,
+      "loss": 1.314,
+      "step": 23160
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 3.2757742404937744,
+      "learning_rate": 3.659926700884296e-05,
+      "loss": 1.1953,
+      "step": 23165
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 1.7651225328445435,
+      "learning_rate": 3.6594062129936974e-05,
+      "loss": 1.2148,
+      "step": 23170
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 6.852691650390625,
+      "learning_rate": 3.6588856610702994e-05,
+      "loss": 1.1669,
+      "step": 23175
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 4.03468132019043,
+      "learning_rate": 3.65836504514285e-05,
+      "loss": 1.3234,
+      "step": 23180
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 1.6668856143951416,
+      "learning_rate": 3.657844365240104e-05,
+      "loss": 1.2555,
+      "step": 23185
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 2.5867419242858887,
+      "learning_rate": 3.657323621390818e-05,
+      "loss": 1.2328,
+      "step": 23190
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 4.226202011108398,
+      "learning_rate": 3.65680281362375e-05,
+      "loss": 1.2422,
+      "step": 23195
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 1.6796958446502686,
+      "learning_rate": 3.656281941967665e-05,
+      "loss": 1.3767,
+      "step": 23200
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 2.898170232772827,
+      "learning_rate": 3.6557610064513314e-05,
+      "loss": 1.2212,
+      "step": 23205
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 2.079902172088623,
+      "learning_rate": 3.6552400071035184e-05,
+      "loss": 1.1753,
+      "step": 23210
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 2.3818612098693848,
+      "learning_rate": 3.654718943953e-05,
+      "loss": 1.2634,
+      "step": 23215
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 2.0267412662506104,
+      "learning_rate": 3.654197817028555e-05,
+      "loss": 1.294,
+      "step": 23220
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 2.046330213546753,
+      "learning_rate": 3.6536766263589646e-05,
+      "loss": 1.2195,
+      "step": 23225
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 1.5229865312576294,
+      "learning_rate": 3.653155371973012e-05,
+      "loss": 1.2128,
+      "step": 23230
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 1.3038949966430664,
+      "learning_rate": 3.652634053899487e-05,
+      "loss": 1.1954,
+      "step": 23235
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 2.71051025390625,
+      "learning_rate": 3.652112672167183e-05,
+      "loss": 1.1658,
+      "step": 23240
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 2.3642067909240723,
+      "learning_rate": 3.651591226804892e-05,
+      "loss": 1.2206,
+      "step": 23245
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 1.8010873794555664,
+      "learning_rate": 3.6510697178414156e-05,
+      "loss": 1.2646,
+      "step": 23250
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 2.7175979614257812,
+      "learning_rate": 3.6505481453055554e-05,
+      "loss": 1.097,
+      "step": 23255
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 7.800134181976318,
+      "learning_rate": 3.6500265092261164e-05,
+      "loss": 0.961,
+      "step": 23260
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 3.0702438354492188,
+      "learning_rate": 3.6495048096319084e-05,
+      "loss": 1.1067,
+      "step": 23265
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 2.858368158340454,
+      "learning_rate": 3.648983046551745e-05,
+      "loss": 1.1169,
+      "step": 23270
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 4.364511966705322,
+      "learning_rate": 3.6484612200144425e-05,
+      "loss": 1.057,
+      "step": 23275
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 1.3396683931350708,
+      "learning_rate": 3.64793933004882e-05,
+      "loss": 1.1468,
+      "step": 23280
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 2.03324294090271,
+      "learning_rate": 3.647417376683702e-05,
+      "loss": 1.126,
+      "step": 23285
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 2.8686647415161133,
+      "learning_rate": 3.646895359947915e-05,
+      "loss": 1.1806,
+      "step": 23290
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 1.482798457145691,
+      "learning_rate": 3.646373279870289e-05,
+      "loss": 1.2207,
+      "step": 23295
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 9.18463134765625,
+      "learning_rate": 3.6458511364796585e-05,
+      "loss": 1.2255,
+      "step": 23300
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 2.2677526473999023,
+      "learning_rate": 3.645328929804861e-05,
+      "loss": 1.3509,
+      "step": 23305
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 1.5335679054260254,
+      "learning_rate": 3.6448066598747365e-05,
+      "loss": 1.2393,
+      "step": 23310
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 3.603954792022705,
+      "learning_rate": 3.644284326718131e-05,
+      "loss": 1.1396,
+      "step": 23315
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 2.735201835632324,
+      "learning_rate": 3.6437619303638906e-05,
+      "loss": 1.0535,
+      "step": 23320
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 2.017690420150757,
+      "learning_rate": 3.6432394708408684e-05,
+      "loss": 1.1505,
+      "step": 23325
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 2.649900436401367,
+      "learning_rate": 3.6427169481779185e-05,
+      "loss": 1.1615,
+      "step": 23330
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 3.7284982204437256,
+      "learning_rate": 3.642194362403899e-05,
+      "loss": 1.0333,
+      "step": 23335
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 1.5677194595336914,
+      "learning_rate": 3.6416717135476726e-05,
+      "loss": 1.1199,
+      "step": 23340
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 1.4740290641784668,
+      "learning_rate": 3.6411490016381036e-05,
+      "loss": 1.2606,
+      "step": 23345
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 2.8093674182891846,
+      "learning_rate": 3.640626226704063e-05,
+      "loss": 1.0608,
+      "step": 23350
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 2.619047164916992,
+      "learning_rate": 3.640103388774419e-05,
+      "loss": 1.2281,
+      "step": 23355
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 3.5901520252227783,
+      "learning_rate": 3.6395804878780514e-05,
+      "loss": 1.0943,
+      "step": 23360
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 2.7254862785339355,
+      "learning_rate": 3.639057524043838e-05,
+      "loss": 1.1382,
+      "step": 23365
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 2.1018872261047363,
+      "learning_rate": 3.638534497300661e-05,
+      "loss": 1.1479,
+      "step": 23370
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 3.5326809883117676,
+      "learning_rate": 3.6380114076774076e-05,
+      "loss": 1.2146,
+      "step": 23375
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 3.996938467025757,
+      "learning_rate": 3.637488255202967e-05,
+      "loss": 1.1037,
+      "step": 23380
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 5.027413845062256,
+      "learning_rate": 3.636965039906232e-05,
+      "loss": 1.1256,
+      "step": 23385
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 2.1053760051727295,
+      "learning_rate": 3.6364417618161e-05,
+      "loss": 1.1905,
+      "step": 23390
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 2.8059771060943604,
+      "learning_rate": 3.635918420961471e-05,
+      "loss": 1.2107,
+      "step": 23395
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 3.2289247512817383,
+      "learning_rate": 3.635395017371248e-05,
+      "loss": 1.0534,
+      "step": 23400
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 2.3690922260284424,
+      "learning_rate": 3.634871551074338e-05,
+      "loss": 1.0033,
+      "step": 23405
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 1.940168857574463,
+      "learning_rate": 3.634348022099652e-05,
+      "loss": 1.1098,
+      "step": 23410
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 2.454968214035034,
+      "learning_rate": 3.633824430476104e-05,
+      "loss": 1.0398,
+      "step": 23415
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 2.5789737701416016,
+      "learning_rate": 3.6333007762326114e-05,
+      "loss": 1.1326,
+      "step": 23420
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 2.819851875305176,
+      "learning_rate": 3.6327770593980946e-05,
+      "loss": 0.9739,
+      "step": 23425
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": Infinity,
+      "learning_rate": 3.6323580408843814e-05,
+      "loss": 1.1279,
+      "step": 23430
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 2.4087419509887695,
+      "learning_rate": 3.631834211458914e-05,
+      "loss": 1.0759,
+      "step": 23435
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 1.9813141822814941,
+      "learning_rate": 3.6313103195234194e-05,
+      "loss": 1.2067,
+      "step": 23440
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 1.0360766649246216,
+      "learning_rate": 3.6307863651068315e-05,
+      "loss": 1.1645,
+      "step": 23445
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 2.9632136821746826,
+      "learning_rate": 3.6302623482380876e-05,
+      "loss": 1.1632,
+      "step": 23450
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 2.059884786605835,
+      "learning_rate": 3.62973826894613e-05,
+      "loss": 1.0065,
+      "step": 23455
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 1.9883657693862915,
+      "learning_rate": 3.629214127259901e-05,
+      "loss": 1.1301,
+      "step": 23460
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 1.7367918491363525,
+      "learning_rate": 3.628689923208351e-05,
+      "loss": 1.1835,
+      "step": 23465
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 1.1696290969848633,
+      "learning_rate": 3.6281656568204285e-05,
+      "loss": 1.241,
+      "step": 23470
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 2.884432315826416,
+      "learning_rate": 3.62764132812509e-05,
+      "loss": 1.0894,
+      "step": 23475
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 1.4623018503189087,
+      "learning_rate": 3.627116937151293e-05,
+      "loss": 1.1655,
+      "step": 23480
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 4.936588287353516,
+      "learning_rate": 3.626592483927999e-05,
+      "loss": 1.0457,
+      "step": 23485
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 2.926048517227173,
+      "learning_rate": 3.626067968484172e-05,
+      "loss": 1.1325,
+      "step": 23490
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 2.242962598800659,
+      "learning_rate": 3.625543390848783e-05,
+      "loss": 1.0459,
+      "step": 23495
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 1.152270793914795,
+      "learning_rate": 3.625018751050803e-05,
+      "loss": 0.974,
+      "step": 23500
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 2.016425371170044,
+      "learning_rate": 3.624494049119205e-05,
+      "loss": 1.1342,
+      "step": 23505
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 2.164924144744873,
+      "learning_rate": 3.623969285082971e-05,
+      "loss": 1.1836,
+      "step": 23510
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 2.6692936420440674,
+      "learning_rate": 3.623444458971081e-05,
+      "loss": 1.2556,
+      "step": 23515
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 1.107765793800354,
+      "learning_rate": 3.622919570812521e-05,
+      "loss": 1.2507,
+      "step": 23520
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 1.9478957653045654,
+      "learning_rate": 3.622394620636281e-05,
+      "loss": 1.0116,
+      "step": 23525
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 2.816403865814209,
+      "learning_rate": 3.621869608471352e-05,
+      "loss": 0.9359,
+      "step": 23530
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 2.90093994140625,
+      "learning_rate": 3.621344534346732e-05,
+      "loss": 1.109,
+      "step": 23535
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 2.196246385574341,
+      "learning_rate": 3.620819398291418e-05,
+      "loss": 1.0691,
+      "step": 23540
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 2.3277361392974854,
+      "learning_rate": 3.6202942003344134e-05,
+      "loss": 1.0831,
+      "step": 23545
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 3.5473828315734863,
+      "learning_rate": 3.619768940504725e-05,
+      "loss": 1.0406,
+      "step": 23550
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 3.1161794662475586,
+      "learning_rate": 3.619243618831362e-05,
+      "loss": 1.145,
+      "step": 23555
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 1.5470253229141235,
+      "learning_rate": 3.618718235343337e-05,
+      "loss": 1.0475,
+      "step": 23560
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 1.8705344200134277,
+      "learning_rate": 3.618192790069668e-05,
+      "loss": 1.1717,
+      "step": 23565
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 2.805480480194092,
+      "learning_rate": 3.617667283039372e-05,
+      "loss": 1.1476,
+      "step": 23570
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 1.1502296924591064,
+      "learning_rate": 3.617141714281473e-05,
+      "loss": 0.9713,
+      "step": 23575
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 2.4745311737060547,
+      "learning_rate": 3.616616083825e-05,
+      "loss": 1.1653,
+      "step": 23580
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 1.45720636844635,
+      "learning_rate": 3.616090391698981e-05,
+      "loss": 1.2729,
+      "step": 23585
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 1.0925184488296509,
+      "learning_rate": 3.615564637932449e-05,
+      "loss": 0.9977,
+      "step": 23590
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 2.9500515460968018,
+      "learning_rate": 3.615038822554442e-05,
+      "loss": 1.0472,
+      "step": 23595
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 1.715450644493103,
+      "learning_rate": 3.614512945594001e-05,
+      "loss": 1.2471,
+      "step": 23600
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 1.3677220344543457,
+      "learning_rate": 3.6139870070801675e-05,
+      "loss": 1.0056,
+      "step": 23605
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 1.0849230289459229,
+      "learning_rate": 3.613461007041989e-05,
+      "loss": 1.0685,
+      "step": 23610
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 1.6670022010803223,
+      "learning_rate": 3.612934945508517e-05,
+      "loss": 1.1105,
+      "step": 23615
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 1.0212535858154297,
+      "learning_rate": 3.612408822508805e-05,
+      "loss": 1.1527,
+      "step": 23620
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 3.0256199836730957,
+      "learning_rate": 3.611882638071909e-05,
+      "loss": 1.1113,
+      "step": 23625
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 7.463032245635986,
+      "learning_rate": 3.6113563922268926e-05,
+      "loss": 1.104,
+      "step": 23630
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 2.4572513103485107,
+      "learning_rate": 3.6108300850028165e-05,
+      "loss": 1.2712,
+      "step": 23635
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 2.344938278198242,
+      "learning_rate": 3.61030371642875e-05,
+      "loss": 0.9881,
+      "step": 23640
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 2.6021649837493896,
+      "learning_rate": 3.609777286533763e-05,
+      "loss": 1.2082,
+      "step": 23645
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 2.0514180660247803,
+      "learning_rate": 3.6092507953469305e-05,
+      "loss": 0.9766,
+      "step": 23650
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 4.5105881690979,
+      "learning_rate": 3.608724242897328e-05,
+      "loss": 0.937,
+      "step": 23655
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 3.2438158988952637,
+      "learning_rate": 3.6081976292140395e-05,
+      "loss": 1.138,
+      "step": 23660
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 3.015331745147705,
+      "learning_rate": 3.607670954326147e-05,
+      "loss": 1.1613,
+      "step": 23665
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 1.9946476221084595,
+      "learning_rate": 3.6071442182627395e-05,
+      "loss": 1.2047,
+      "step": 23670
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 3.2494962215423584,
+      "learning_rate": 3.606617421052908e-05,
+      "loss": 1.0119,
+      "step": 23675
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 2.6945230960845947,
+      "learning_rate": 3.6060905627257455e-05,
+      "loss": 1.0004,
+      "step": 23680
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 2.1307501792907715,
+      "learning_rate": 3.6055636433103524e-05,
+      "loss": 0.9985,
+      "step": 23685
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 1.7671560049057007,
+      "learning_rate": 3.6050366628358276e-05,
+      "loss": 1.1638,
+      "step": 23690
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 4.131464004516602,
+      "learning_rate": 3.6045096213312766e-05,
+      "loss": 1.0299,
+      "step": 23695
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 1.8803554773330688,
+      "learning_rate": 3.6039825188258075e-05,
+      "loss": 1.1465,
+      "step": 23700
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 4.313520908355713,
+      "learning_rate": 3.603455355348531e-05,
+      "loss": 1.1605,
+      "step": 23705
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 1.4319663047790527,
+      "learning_rate": 3.602928130928563e-05,
+      "loss": 1.13,
+      "step": 23710
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 1.6968344449996948,
+      "learning_rate": 3.60240084559502e-05,
+      "loss": 1.0394,
+      "step": 23715
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 2.2089247703552246,
+      "learning_rate": 3.601873499377024e-05,
+      "loss": 1.0584,
+      "step": 23720
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 2.618340253829956,
+      "learning_rate": 3.601346092303701e-05,
+      "loss": 1.1993,
+      "step": 23725
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 1.327785849571228,
+      "learning_rate": 3.600818624404177e-05,
+      "loss": 1.0758,
+      "step": 23730
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 3.1127471923828125,
+      "learning_rate": 3.600291095707585e-05,
+      "loss": 1.1543,
+      "step": 23735
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 1.783038854598999,
+      "learning_rate": 3.59976350624306e-05,
+      "loss": 1.0839,
+      "step": 23740
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 1.7736161947250366,
+      "learning_rate": 3.5992358560397394e-05,
+      "loss": 1.0779,
+      "step": 23745
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 2.9156086444854736,
+      "learning_rate": 3.5987081451267646e-05,
+      "loss": 1.1455,
+      "step": 23750
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 3.304860830307007,
+      "learning_rate": 3.598180373533281e-05,
+      "loss": 1.0505,
+      "step": 23755
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 1.9659478664398193,
+      "learning_rate": 3.597652541288438e-05,
+      "loss": 1.1696,
+      "step": 23760
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 3.989715814590454,
+      "learning_rate": 3.597124648421384e-05,
+      "loss": 1.1278,
+      "step": 23765
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 2.289445400238037,
+      "learning_rate": 3.596596694961278e-05,
+      "loss": 1.2564,
+      "step": 23770
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 1.6638239622116089,
+      "learning_rate": 3.5960686809372756e-05,
+      "loss": 1.2167,
+      "step": 23775
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 2.3342809677124023,
+      "learning_rate": 3.595540606378539e-05,
+      "loss": 1.1625,
+      "step": 23780
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 2.215494155883789,
+      "learning_rate": 3.595012471314234e-05,
+      "loss": 0.9338,
+      "step": 23785
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 2.7711009979248047,
+      "learning_rate": 3.594484275773529e-05,
+      "loss": 1.0951,
+      "step": 23790
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 2.727311611175537,
+      "learning_rate": 3.593956019785594e-05,
+      "loss": 1.1065,
+      "step": 23795
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 1.7198123931884766,
+      "learning_rate": 3.5934277033796055e-05,
+      "loss": 1.102,
+      "step": 23800
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 1.2289972305297852,
+      "learning_rate": 3.592899326584741e-05,
+      "loss": 1.1591,
+      "step": 23805
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 9.438529968261719,
+      "learning_rate": 3.5923708894301836e-05,
+      "loss": 1.2098,
+      "step": 23810
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 2.6810221672058105,
+      "learning_rate": 3.591842391945117e-05,
+      "loss": 1.1817,
+      "step": 23815
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 3.3912224769592285,
+      "learning_rate": 3.591313834158729e-05,
+      "loss": 1.1386,
+      "step": 23820
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 3.810546398162842,
+      "learning_rate": 3.590785216100214e-05,
+      "loss": 1.1549,
+      "step": 23825
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 2.7697300910949707,
+      "learning_rate": 3.590256537798765e-05,
+      "loss": 1.2004,
+      "step": 23830
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 1.2456960678100586,
+      "learning_rate": 3.58972779928358e-05,
+      "loss": 1.1418,
+      "step": 23835
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 2.144900321960449,
+      "learning_rate": 3.589199000583862e-05,
+      "loss": 1.199,
+      "step": 23840
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 3.3847482204437256,
+      "learning_rate": 3.588670141728815e-05,
+      "loss": 1.1028,
+      "step": 23845
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 2.134644031524658,
+      "learning_rate": 3.5881412227476476e-05,
+      "loss": 1.0711,
+      "step": 23850
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 1.6338741779327393,
+      "learning_rate": 3.587612243669571e-05,
+      "loss": 1.1405,
+      "step": 23855
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 4.937558650970459,
+      "learning_rate": 3.5870832045238013e-05,
+      "loss": 1.1193,
+      "step": 23860
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 2.3749191761016846,
+      "learning_rate": 3.5865541053395564e-05,
+      "loss": 1.0552,
+      "step": 23865
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 1.3177196979522705,
+      "learning_rate": 3.586024946146057e-05,
+      "loss": 1.2673,
+      "step": 23870
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 5.730068206787109,
+      "learning_rate": 3.585495726972529e-05,
+      "loss": 1.181,
+      "step": 23875
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 2.8973639011383057,
+      "learning_rate": 3.584966447848201e-05,
+      "loss": 1.1447,
+      "step": 23880
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 2.603912830352783,
+      "learning_rate": 3.584437108802303e-05,
+      "loss": 1.1617,
+      "step": 23885
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 2.1435694694519043,
+      "learning_rate": 3.583907709864072e-05,
+      "loss": 1.1438,
+      "step": 23890
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 3.071255922317505,
+      "learning_rate": 3.5833782510627436e-05,
+      "loss": 1.1503,
+      "step": 23895
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 2.3412327766418457,
+      "learning_rate": 3.582848732427561e-05,
+      "loss": 1.2069,
+      "step": 23900
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 2.986461639404297,
+      "learning_rate": 3.5823191539877674e-05,
+      "loss": 1.0559,
+      "step": 23905
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 1.189259648323059,
+      "learning_rate": 3.581789515772613e-05,
+      "loss": 1.2836,
+      "step": 23910
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 2.6002511978149414,
+      "learning_rate": 3.5812598178113476e-05,
+      "loss": 1.2934,
+      "step": 23915
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 1.7060097455978394,
+      "learning_rate": 3.580730060133227e-05,
+      "loss": 1.1425,
+      "step": 23920
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 2.59647536277771,
+      "learning_rate": 3.580200242767508e-05,
+      "loss": 1.0499,
+      "step": 23925
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 1.8929245471954346,
+      "learning_rate": 3.5796703657434526e-05,
+      "loss": 1.2425,
+      "step": 23930
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 1.3510212898254395,
+      "learning_rate": 3.579140429090325e-05,
+      "loss": 1.0355,
+      "step": 23935
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 1.923653483390808,
+      "learning_rate": 3.578610432837393e-05,
+      "loss": 1.1344,
+      "step": 23940
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 2.9542462825775146,
+      "learning_rate": 3.578080377013928e-05,
+      "loss": 1.2874,
+      "step": 23945
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 3.119513988494873,
+      "learning_rate": 3.577550261649204e-05,
+      "loss": 1.1657,
+      "step": 23950
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 2.2563698291778564,
+      "learning_rate": 3.5770200867725e-05,
+      "loss": 1.3417,
+      "step": 23955
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 3.3380630016326904,
+      "learning_rate": 3.5764898524130965e-05,
+      "loss": 1.2012,
+      "step": 23960
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 3.131051540374756,
+      "learning_rate": 3.5759595586002765e-05,
+      "loss": 1.0301,
+      "step": 23965
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 4.349595069885254,
+      "learning_rate": 3.575429205363329e-05,
+      "loss": 1.1835,
+      "step": 23970
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 1.5668723583221436,
+      "learning_rate": 3.574898792731544e-05,
+      "loss": 1.2774,
+      "step": 23975
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 6.6469855308532715,
+      "learning_rate": 3.5743683207342154e-05,
+      "loss": 1.2421,
+      "step": 23980
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 1.9514319896697998,
+      "learning_rate": 3.573837789400643e-05,
+      "loss": 1.1226,
+      "step": 23985
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 1.3996671438217163,
+      "learning_rate": 3.5733071987601235e-05,
+      "loss": 1.195,
+      "step": 23990
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 0.9684573411941528,
+      "learning_rate": 3.572776548841964e-05,
+      "loss": 1.1764,
+      "step": 23995
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 1.6927798986434937,
+      "learning_rate": 3.572245839675471e-05,
+      "loss": 1.094,
+      "step": 24000
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 3.316277265548706,
+      "learning_rate": 3.571715071289954e-05,
+      "loss": 1.0117,
+      "step": 24005
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 3.6141722202301025,
+      "learning_rate": 3.571184243714729e-05,
+      "loss": 1.0562,
+      "step": 24010
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 3.6086862087249756,
+      "learning_rate": 3.570653356979111e-05,
+      "loss": 1.1708,
+      "step": 24015
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 1.5085479021072388,
+      "learning_rate": 3.5701224111124206e-05,
+      "loss": 1.1285,
+      "step": 24020
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 3.2829747200012207,
+      "learning_rate": 3.5695914061439816e-05,
+      "loss": 1.227,
+      "step": 24025
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 3.2367653846740723,
+      "learning_rate": 3.569060342103121e-05,
+      "loss": 1.0892,
+      "step": 24030
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 1.7712103128433228,
+      "learning_rate": 3.56852921901917e-05,
+      "loss": 1.0643,
+      "step": 24035
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 3.2912991046905518,
+      "learning_rate": 3.56799803692146e-05,
+      "loss": 1.1242,
+      "step": 24040
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 2.6889803409576416,
+      "learning_rate": 3.5674667958393286e-05,
+      "loss": 1.1923,
+      "step": 24045
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 3.47308349609375,
+      "learning_rate": 3.566935495802117e-05,
+      "loss": 1.2753,
+      "step": 24050
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 2.2733964920043945,
+      "learning_rate": 3.566404136839165e-05,
+      "loss": 1.2105,
+      "step": 24055
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 1.1582354307174683,
+      "learning_rate": 3.565872718979822e-05,
+      "loss": 0.9991,
+      "step": 24060
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 2.0912468433380127,
+      "learning_rate": 3.565341242253437e-05,
+      "loss": 1.1606,
+      "step": 24065
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 4.56016731262207,
+      "learning_rate": 3.5648097066893614e-05,
+      "loss": 1.2603,
+      "step": 24070
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 2.5408365726470947,
+      "learning_rate": 3.564278112316953e-05,
+      "loss": 1.1349,
+      "step": 24075
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 2.359050989151001,
+      "learning_rate": 3.563746459165571e-05,
+      "loss": 1.1326,
+      "step": 24080
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 2.704974889755249,
+      "learning_rate": 3.563214747264578e-05,
+      "loss": 1.2259,
+      "step": 24085
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 3.2640304565429688,
+      "learning_rate": 3.562682976643339e-05,
+      "loss": 1.0601,
+      "step": 24090
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 2.6645710468292236,
+      "learning_rate": 3.562151147331224e-05,
+      "loss": 1.2108,
+      "step": 24095
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 3.9869096279144287,
+      "learning_rate": 3.5616192593576055e-05,
+      "loss": 1.144,
+      "step": 24100
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 1.9926941394805908,
+      "learning_rate": 3.561087312751858e-05,
+      "loss": 1.0182,
+      "step": 24105
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 1.0935914516448975,
+      "learning_rate": 3.560555307543362e-05,
+      "loss": 1.3015,
+      "step": 24110
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 3.214768409729004,
+      "learning_rate": 3.5600232437614984e-05,
+      "loss": 1.3117,
+      "step": 24115
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 3.5183026790618896,
+      "learning_rate": 3.559491121435653e-05,
+      "loss": 1.1814,
+      "step": 24120
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 0.981540858745575,
+      "learning_rate": 3.558958940595214e-05,
+      "loss": 1.2303,
+      "step": 24125
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 2.553018569946289,
+      "learning_rate": 3.558426701269574e-05,
+      "loss": 1.1696,
+      "step": 24130
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 2.0565733909606934,
+      "learning_rate": 3.557894403488127e-05,
+      "loss": 1.2089,
+      "step": 24135
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 1.6102100610733032,
+      "learning_rate": 3.5573620472802714e-05,
+      "loss": 1.1943,
+      "step": 24140
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 1.952919363975525,
+      "learning_rate": 3.55682963267541e-05,
+      "loss": 1.0149,
+      "step": 24145
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 2.255753517150879,
+      "learning_rate": 3.556297159702946e-05,
+      "loss": 1.1736,
+      "step": 24150
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 2.6610512733459473,
+      "learning_rate": 3.5557646283922875e-05,
+      "loss": 1.2155,
+      "step": 24155
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 3.2673702239990234,
+      "learning_rate": 3.555232038772846e-05,
+      "loss": 0.9963,
+      "step": 24160
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 3.3183624744415283,
+      "learning_rate": 3.554699390874036e-05,
+      "loss": 1.101,
+      "step": 24165
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 1.4142159223556519,
+      "learning_rate": 3.554166684725275e-05,
+      "loss": 1.0708,
+      "step": 24170
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 2.9164345264434814,
+      "learning_rate": 3.553633920355983e-05,
+      "loss": 1.1302,
+      "step": 24175
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 2.0309815406799316,
+      "learning_rate": 3.5531010977955865e-05,
+      "loss": 1.0906,
+      "step": 24180
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 2.8416526317596436,
+      "learning_rate": 3.55256821707351e-05,
+      "loss": 1.0474,
+      "step": 24185
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 0.9157441854476929,
+      "learning_rate": 3.5520352782191845e-05,
+      "loss": 1.2212,
+      "step": 24190
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 6.699018478393555,
+      "learning_rate": 3.5515022812620446e-05,
+      "loss": 1.1924,
+      "step": 24195
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 4.399092674255371,
+      "learning_rate": 3.550969226231527e-05,
+      "loss": 1.047,
+      "step": 24200
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 2.4295222759246826,
+      "learning_rate": 3.550436113157071e-05,
+      "loss": 1.2108,
+      "step": 24205
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 3.0615289211273193,
+      "learning_rate": 3.54990294206812e-05,
+      "loss": 1.0198,
+      "step": 24210
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 5.101573467254639,
+      "learning_rate": 3.549369712994122e-05,
+      "loss": 1.0861,
+      "step": 24215
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 2.2024030685424805,
+      "learning_rate": 3.548836425964524e-05,
+      "loss": 1.0598,
+      "step": 24220
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 3.093062400817871,
+      "learning_rate": 3.548303081008781e-05,
+      "loss": 1.1864,
+      "step": 24225
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 1.283988118171692,
+      "learning_rate": 3.547769678156349e-05,
+      "loss": 1.2008,
+      "step": 24230
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 1.7267249822616577,
+      "learning_rate": 3.547236217436686e-05,
+      "loss": 1.1673,
+      "step": 24235
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 2.7360587120056152,
+      "learning_rate": 3.546702698879256e-05,
+      "loss": 1.0917,
+      "step": 24240
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 3.3460683822631836,
+      "learning_rate": 3.546169122513524e-05,
+      "loss": 1.1798,
+      "step": 24245
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 3.3018956184387207,
+      "learning_rate": 3.545635488368959e-05,
+      "loss": 1.1226,
+      "step": 24250
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 1.762122392654419,
+      "learning_rate": 3.5451017964750316e-05,
+      "loss": 1.1001,
+      "step": 24255
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 2.407038688659668,
+      "learning_rate": 3.544568046861219e-05,
+      "loss": 1.1731,
+      "step": 24260
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 2.4787652492523193,
+      "learning_rate": 3.544034239556999e-05,
+      "loss": 1.0903,
+      "step": 24265
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 3.5315353870391846,
+      "learning_rate": 3.543500374591853e-05,
+      "loss": 1.2661,
+      "step": 24270
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 1.81735360622406,
+      "learning_rate": 3.5429664519952664e-05,
+      "loss": 1.0972,
+      "step": 24275
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 3.0193166732788086,
+      "learning_rate": 3.542432471796726e-05,
+      "loss": 1.227,
+      "step": 24280
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 2.1791579723358154,
+      "learning_rate": 3.5418984340257245e-05,
+      "loss": 1.1118,
+      "step": 24285
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 2.2652621269226074,
+      "learning_rate": 3.541364338711755e-05,
+      "loss": 1.2193,
+      "step": 24290
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 3.2821550369262695,
+      "learning_rate": 3.5408301858843155e-05,
+      "loss": 1.1751,
+      "step": 24295
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 1.8815666437149048,
+      "learning_rate": 3.540295975572907e-05,
+      "loss": 1.0835,
+      "step": 24300
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 6.063453674316406,
+      "learning_rate": 3.539761707807032e-05,
+      "loss": 1.081,
+      "step": 24305
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 1.7175376415252686,
+      "learning_rate": 3.5392273826162004e-05,
+      "loss": 1.1278,
+      "step": 24310
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 1.4363011121749878,
+      "learning_rate": 3.538693000029919e-05,
+      "loss": 1.0402,
+      "step": 24315
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 2.4211199283599854,
+      "learning_rate": 3.538158560077704e-05,
+      "loss": 1.3065,
+      "step": 24320
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 4.585165023803711,
+      "learning_rate": 3.537624062789071e-05,
+      "loss": 1.0919,
+      "step": 24325
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 2.929621934890747,
+      "learning_rate": 3.537089508193539e-05,
+      "loss": 1.2163,
+      "step": 24330
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 3.6469857692718506,
+      "learning_rate": 3.536554896320632e-05,
+      "loss": 1.3069,
+      "step": 24335
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 2.2444517612457275,
+      "learning_rate": 3.536020227199875e-05,
+      "loss": 1.2386,
+      "step": 24340
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 1.8699992895126343,
+      "learning_rate": 3.535485500860798e-05,
+      "loss": 1.1221,
+      "step": 24345
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 2.0010950565338135,
+      "learning_rate": 3.5349507173329324e-05,
+      "loss": 1.157,
+      "step": 24350
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 1.865695595741272,
+      "learning_rate": 3.534415876645815e-05,
+      "loss": 0.9495,
+      "step": 24355
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 1.9279874563217163,
+      "learning_rate": 3.533880978828984e-05,
+      "loss": 1.2705,
+      "step": 24360
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 2.339104175567627,
+      "learning_rate": 3.5333460239119814e-05,
+      "loss": 1.081,
+      "step": 24365
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 2.058978319168091,
+      "learning_rate": 3.5328110119243515e-05,
+      "loss": 0.9748,
+      "step": 24370
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 5.336772441864014,
+      "learning_rate": 3.532275942895644e-05,
+      "loss": 1.1164,
+      "step": 24375
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 13.560040473937988,
+      "learning_rate": 3.531740816855408e-05,
+      "loss": 1.2611,
+      "step": 24380
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 3.361783742904663,
+      "learning_rate": 3.5312056338331986e-05,
+      "loss": 1.0729,
+      "step": 24385
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 1.754549503326416,
+      "learning_rate": 3.530670393858575e-05,
+      "loss": 1.1315,
+      "step": 24390
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 1.895077109336853,
+      "learning_rate": 3.530135096961097e-05,
+      "loss": 1.1235,
+      "step": 24395
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 2.7250521183013916,
+      "learning_rate": 3.529599743170328e-05,
+      "loss": 0.9917,
+      "step": 24400
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 3.637592077255249,
+      "learning_rate": 3.529064332515836e-05,
+      "loss": 1.015,
+      "step": 24405
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 2.967895984649658,
+      "learning_rate": 3.5285288650271896e-05,
+      "loss": 1.285,
+      "step": 24410
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 3.107884407043457,
+      "learning_rate": 3.527993340733964e-05,
+      "loss": 1.0319,
+      "step": 24415
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 1.9659065008163452,
+      "learning_rate": 3.527457759665734e-05,
+      "loss": 1.1129,
+      "step": 24420
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 3.5118746757507324,
+      "learning_rate": 3.52692212185208e-05,
+      "loss": 1.1486,
+      "step": 24425
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 3.2912468910217285,
+      "learning_rate": 3.526386427322585e-05,
+      "loss": 1.1864,
+      "step": 24430
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 3.768181562423706,
+      "learning_rate": 3.5258506761068344e-05,
+      "loss": 1.2152,
+      "step": 24435
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 4.644777774810791,
+      "learning_rate": 3.525314868234417e-05,
+      "loss": 1.145,
+      "step": 24440
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 1.425614595413208,
+      "learning_rate": 3.524779003734925e-05,
+      "loss": 1.1543,
+      "step": 24445
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 2.4408340454101562,
+      "learning_rate": 3.524243082637954e-05,
+      "loss": 1.1951,
+      "step": 24450
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 1.1493414640426636,
+      "learning_rate": 3.523707104973102e-05,
+      "loss": 1.0654,
+      "step": 24455
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 2.574376106262207,
+      "learning_rate": 3.523171070769972e-05,
+      "loss": 1.0558,
+      "step": 24460
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 2.255390167236328,
+      "learning_rate": 3.522634980058166e-05,
+      "loss": 1.2315,
+      "step": 24465
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 1.4033783674240112,
+      "learning_rate": 3.5220988328672935e-05,
+      "loss": 1.0303,
+      "step": 24470
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 2.0258255004882812,
+      "learning_rate": 3.521562629226965e-05,
+      "loss": 1.1492,
+      "step": 24475
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 2.6093392372131348,
+      "learning_rate": 3.521026369166793e-05,
+      "loss": 1.1797,
+      "step": 24480
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 4.753375053405762,
+      "learning_rate": 3.520490052716397e-05,
+      "loss": 1.1525,
+      "step": 24485
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 2.713651657104492,
+      "learning_rate": 3.5199536799053965e-05,
+      "loss": 1.0279,
+      "step": 24490
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 2.4132895469665527,
+      "learning_rate": 3.5194172507634136e-05,
+      "loss": 1.0203,
+      "step": 24495
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 2.83127760887146,
+      "learning_rate": 3.5188807653200764e-05,
+      "loss": 1.2186,
+      "step": 24500
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 2.688488245010376,
+      "learning_rate": 3.5183442236050126e-05,
+      "loss": 1.2564,
+      "step": 24505
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 1.0452677011489868,
+      "learning_rate": 3.517807625647857e-05,
+      "loss": 1.1982,
+      "step": 24510
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 3.4765284061431885,
+      "learning_rate": 3.5172709714782435e-05,
+      "loss": 1.1172,
+      "step": 24515
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 4.563774108886719,
+      "learning_rate": 3.5167342611258114e-05,
+      "loss": 1.2248,
+      "step": 24520
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 3.0111424922943115,
+      "learning_rate": 3.5161974946202035e-05,
+      "loss": 0.9888,
+      "step": 24525
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 1.582155466079712,
+      "learning_rate": 3.515660671991064e-05,
+      "loss": 1.0793,
+      "step": 24530
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 2.5320088863372803,
+      "learning_rate": 3.515123793268042e-05,
+      "loss": 1.2144,
+      "step": 24535
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 1.5951532125473022,
+      "learning_rate": 3.5145868584807875e-05,
+      "loss": 1.0229,
+      "step": 24540
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 1.4666327238082886,
+      "learning_rate": 3.514049867658955e-05,
+      "loss": 1.273,
+      "step": 24545
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 3.8786826133728027,
+      "learning_rate": 3.513512820832202e-05,
+      "loss": 1.1415,
+      "step": 24550
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 1.4749125242233276,
+      "learning_rate": 3.5129757180301906e-05,
+      "loss": 1.2209,
+      "step": 24555
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 6.20326566696167,
+      "learning_rate": 3.512438559282583e-05,
+      "loss": 0.953,
+      "step": 24560
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 1.2090144157409668,
+      "learning_rate": 3.511901344619045e-05,
+      "loss": 1.195,
+      "step": 24565
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 2.7025344371795654,
+      "learning_rate": 3.5113640740692485e-05,
+      "loss": 1.1066,
+      "step": 24570
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 2.9294729232788086,
+      "learning_rate": 3.510826747662865e-05,
+      "loss": 1.1074,
+      "step": 24575
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 3.470428943634033,
+      "learning_rate": 3.510289365429571e-05,
+      "loss": 1.166,
+      "step": 24580
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 10.227189064025879,
+      "learning_rate": 3.509751927399046e-05,
+      "loss": 1.1535,
+      "step": 24585
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 1.4875662326812744,
+      "learning_rate": 3.509214433600971e-05,
+      "loss": 1.2838,
+      "step": 24590
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 2.2764317989349365,
+      "learning_rate": 3.508676884065032e-05,
+      "loss": 1.1669,
+      "step": 24595
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 3.8440630435943604,
+      "learning_rate": 3.5081392788209176e-05,
+      "loss": 1.0723,
+      "step": 24600
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 3.417236089706421,
+      "learning_rate": 3.507601617898319e-05,
+      "loss": 1.193,
+      "step": 24605
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 2.8095319271087646,
+      "learning_rate": 3.5070639013269296e-05,
+      "loss": 1.1633,
+      "step": 24610
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 4.380091190338135,
+      "learning_rate": 3.5065261291364485e-05,
+      "loss": 1.1677,
+      "step": 24615
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 1.5369083881378174,
+      "learning_rate": 3.505988301356574e-05,
+      "loss": 1.2476,
+      "step": 24620
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 1.350716233253479,
+      "learning_rate": 3.505450418017012e-05,
+      "loss": 1.0492,
+      "step": 24625
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 1.859533429145813,
+      "learning_rate": 3.5049124791474696e-05,
+      "loss": 0.9475,
+      "step": 24630
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 3.090188503265381,
+      "learning_rate": 3.504374484777655e-05,
+      "loss": 1.0635,
+      "step": 24635
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 4.572514533996582,
+      "learning_rate": 3.503836434937281e-05,
+      "loss": 1.0462,
+      "step": 24640
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 1.258280634880066,
+      "learning_rate": 3.503298329656064e-05,
+      "loss": 1.0908,
+      "step": 24645
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 3.248828887939453,
+      "learning_rate": 3.5027601689637244e-05,
+      "loss": 1.1565,
+      "step": 24650
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 3.4819695949554443,
+      "learning_rate": 3.502221952889981e-05,
+      "loss": 1.2237,
+      "step": 24655
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 2.005113363265991,
+      "learning_rate": 3.5016836814645624e-05,
+      "loss": 1.2283,
+      "step": 24660
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 4.542233943939209,
+      "learning_rate": 3.5011453547171954e-05,
+      "loss": 1.0564,
+      "step": 24665
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 1.675430178642273,
+      "learning_rate": 3.50060697267761e-05,
+      "loss": 1.0971,
+      "step": 24670
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 3.372847318649292,
+      "learning_rate": 3.500068535375543e-05,
+      "loss": 1.0592,
+      "step": 24675
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 2.804337978363037,
+      "learning_rate": 3.499530042840728e-05,
+      "loss": 0.9945,
+      "step": 24680
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 2.0974247455596924,
+      "learning_rate": 3.49899149510291e-05,
+      "loss": 1.0258,
+      "step": 24685
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 1.919437289237976,
+      "learning_rate": 3.498452892191829e-05,
+      "loss": 1.028,
+      "step": 24690
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 3.193256139755249,
+      "learning_rate": 3.4979142341372337e-05,
+      "loss": 1.1782,
+      "step": 24695
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 2.895226240158081,
+      "learning_rate": 3.4973755209688716e-05,
+      "loss": 1.1318,
+      "step": 24700
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 1.8412282466888428,
+      "learning_rate": 3.496836752716496e-05,
+      "loss": 1.335,
+      "step": 24705
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 4.568256378173828,
+      "learning_rate": 3.496297929409863e-05,
+      "loss": 0.9776,
+      "step": 24710
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 2.8621883392333984,
+      "learning_rate": 3.4957590510787306e-05,
+      "loss": 0.984,
+      "step": 24715
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 3.2574169635772705,
+      "learning_rate": 3.495220117752861e-05,
+      "loss": 1.0085,
+      "step": 24720
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 4.439530849456787,
+      "learning_rate": 3.4946811294620196e-05,
+      "loss": 1.0522,
+      "step": 24725
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 3.148686170578003,
+      "learning_rate": 3.494142086235972e-05,
+      "loss": 0.8726,
+      "step": 24730
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 8.771424293518066,
+      "learning_rate": 3.4936029881044917e-05,
+      "loss": 1.1329,
+      "step": 24735
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 2.2926177978515625,
+      "learning_rate": 3.493063835097351e-05,
+      "loss": 1.1628,
+      "step": 24740
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 1.3365390300750732,
+      "learning_rate": 3.4925246272443256e-05,
+      "loss": 1.2449,
+      "step": 24745
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 2.2112948894500732,
+      "learning_rate": 3.491985364575197e-05,
+      "loss": 1.0303,
+      "step": 24750
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 1.917383074760437,
+      "learning_rate": 3.4914460471197486e-05,
+      "loss": 1.3089,
+      "step": 24755
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 3.297748565673828,
+      "learning_rate": 3.4909066749077654e-05,
+      "loss": 1.222,
+      "step": 24760
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 2.1157360076904297,
+      "learning_rate": 3.490367247969036e-05,
+      "loss": 1.1927,
+      "step": 24765
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 3.120800733566284,
+      "learning_rate": 3.489827766333353e-05,
+      "loss": 1.2292,
+      "step": 24770
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 1.5992114543914795,
+      "learning_rate": 3.4892882300305127e-05,
+      "loss": 1.2151,
+      "step": 24775
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 1.0426512956619263,
+      "learning_rate": 3.48874863909031e-05,
+      "loss": 1.1554,
+      "step": 24780
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 2.195253372192383,
+      "learning_rate": 3.488208993542549e-05,
+      "loss": 1.042,
+      "step": 24785
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 2.7746365070343018,
+      "learning_rate": 3.487669293417032e-05,
+      "loss": 1.104,
+      "step": 24790
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 3.3738327026367188,
+      "learning_rate": 3.487129538743567e-05,
+      "loss": 1.0831,
+      "step": 24795
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 2.653986930847168,
+      "learning_rate": 3.4865897295519624e-05,
+      "loss": 1.1528,
+      "step": 24800
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 3.24540638923645,
+      "learning_rate": 3.486049865872033e-05,
+      "loss": 1.2233,
+      "step": 24805
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 3.4608511924743652,
+      "learning_rate": 3.485509947733595e-05,
+      "loss": 1.2021,
+      "step": 24810
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 1.495871663093567,
+      "learning_rate": 3.4849699751664664e-05,
+      "loss": 1.2353,
+      "step": 24815
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 1.2478872537612915,
+      "learning_rate": 3.484429948200471e-05,
+      "loss": 1.2628,
+      "step": 24820
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 4.952610015869141,
+      "learning_rate": 3.483889866865432e-05,
+      "loss": 1.1575,
+      "step": 24825
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 1.0654171705245972,
+      "learning_rate": 3.483349731191178e-05,
+      "loss": 1.265,
+      "step": 24830
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 5.432766437530518,
+      "learning_rate": 3.48280954120754e-05,
+      "loss": 1.0648,
+      "step": 24835
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 6.357416152954102,
+      "learning_rate": 3.482269296944354e-05,
+      "loss": 1.0119,
+      "step": 24840
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 2.8604278564453125,
+      "learning_rate": 3.481728998431455e-05,
+      "loss": 1.2266,
+      "step": 24845
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 3.6308391094207764,
+      "learning_rate": 3.481188645698684e-05,
+      "loss": 1.2901,
+      "step": 24850
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 2.2123422622680664,
+      "learning_rate": 3.4806482387758846e-05,
+      "loss": 1.2228,
+      "step": 24855
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 1.6754069328308105,
+      "learning_rate": 3.4801077776929016e-05,
+      "loss": 1.1403,
+      "step": 24860
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 1.3092021942138672,
+      "learning_rate": 3.479567262479584e-05,
+      "loss": 1.1213,
+      "step": 24865
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 3.2713966369628906,
+      "learning_rate": 3.479026693165786e-05,
+      "loss": 1.0768,
+      "step": 24870
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 2.951756000518799,
+      "learning_rate": 3.478486069781361e-05,
+      "loss": 1.1956,
+      "step": 24875
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 2.3628857135772705,
+      "learning_rate": 3.4779453923561675e-05,
+      "loss": 1.2273,
+      "step": 24880
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 2.3509888648986816,
+      "learning_rate": 3.477404660920066e-05,
+      "loss": 1.1566,
+      "step": 24885
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 1.5141558647155762,
+      "learning_rate": 3.4768638755029226e-05,
+      "loss": 1.1827,
+      "step": 24890
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 1.5687611103057861,
+      "learning_rate": 3.476323036134601e-05,
+      "loss": 1.1956,
+      "step": 24895
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 1.7349860668182373,
+      "learning_rate": 3.475782142844974e-05,
+      "loss": 1.1523,
+      "step": 24900
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 2.322187900543213,
+      "learning_rate": 3.475241195663913e-05,
+      "loss": 1.2337,
+      "step": 24905
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 1.1443284749984741,
+      "learning_rate": 3.4747001946212944e-05,
+      "loss": 1.0199,
+      "step": 24910
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 1.456677794456482,
+      "learning_rate": 3.4741591397469975e-05,
+      "loss": 1.0837,
+      "step": 24915
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 1.2312963008880615,
+      "learning_rate": 3.473618031070903e-05,
+      "loss": 1.1786,
+      "step": 24920
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 1.1006726026535034,
+      "learning_rate": 3.4730768686228976e-05,
+      "loss": 1.1402,
+      "step": 24925
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 3.3980886936187744,
+      "learning_rate": 3.4725356524328686e-05,
+      "loss": 1.0626,
+      "step": 24930
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 4.119764804840088,
+      "learning_rate": 3.471994382530706e-05,
+      "loss": 1.1438,
+      "step": 24935
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 1.62360417842865,
+      "learning_rate": 3.471453058946303e-05,
+      "loss": 1.1758,
+      "step": 24940
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 2.4606423377990723,
+      "learning_rate": 3.4709116817095584e-05,
+      "loss": 0.9549,
+      "step": 24945
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 2.0009765625,
+      "learning_rate": 3.4703702508503704e-05,
+      "loss": 1.2969,
+      "step": 24950
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 2.6999094486236572,
+      "learning_rate": 3.469828766398643e-05,
+      "loss": 1.1032,
+      "step": 24955
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 4.238891124725342,
+      "learning_rate": 3.46928722838428e-05,
+      "loss": 1.186,
+      "step": 24960
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 3.6732418537139893,
+      "learning_rate": 3.468745636837191e-05,
+      "loss": 1.1267,
+      "step": 24965
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 4.03202486038208,
+      "learning_rate": 3.468203991787287e-05,
+      "loss": 1.08,
+      "step": 24970
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 2.4040608406066895,
+      "learning_rate": 3.467662293264484e-05,
+      "loss": 0.9691,
+      "step": 24975
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 2.929931640625,
+      "learning_rate": 3.4671205412986975e-05,
+      "loss": 1.0961,
+      "step": 24980
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 3.1068196296691895,
+      "learning_rate": 3.466578735919849e-05,
+      "loss": 1.2342,
+      "step": 24985
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 3.1283774375915527,
+      "learning_rate": 3.466036877157862e-05,
+      "loss": 1.1537,
+      "step": 24990
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 2.7555837631225586,
+      "learning_rate": 3.465494965042662e-05,
+      "loss": 1.1126,
+      "step": 24995
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 1.7763209342956543,
+      "learning_rate": 3.4649529996041784e-05,
+      "loss": 1.1552,
+      "step": 25000
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 2.4423739910125732,
+      "learning_rate": 3.464410980872344e-05,
+      "loss": 0.9622,
+      "step": 25005
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 2.395561933517456,
+      "learning_rate": 3.463868908877094e-05,
+      "loss": 1.0734,
+      "step": 25010
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 1.395330548286438,
+      "learning_rate": 3.463326783648365e-05,
+      "loss": 0.914,
+      "step": 25015
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 2.5400102138519287,
+      "learning_rate": 3.4627846052161e-05,
+      "loss": 1.1574,
+      "step": 25020
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 1.9044983386993408,
+      "learning_rate": 3.4622423736102414e-05,
+      "loss": 1.2176,
+      "step": 25025
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 2.5244264602661133,
+      "learning_rate": 3.461700088860737e-05,
+      "loss": 1.0935,
+      "step": 25030
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 1.8527662754058838,
+      "learning_rate": 3.4611577509975366e-05,
+      "loss": 1.1516,
+      "step": 25035
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 2.225349187850952,
+      "learning_rate": 3.460615360050592e-05,
+      "loss": 1.0199,
+      "step": 25040
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 1.780076026916504,
+      "learning_rate": 3.46007291604986e-05,
+      "loss": 1.2104,
+      "step": 25045
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 4.2586669921875,
+      "learning_rate": 3.4595304190252993e-05,
+      "loss": 1.1166,
+      "step": 25050
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 2.9571564197540283,
+      "learning_rate": 3.45898786900687e-05,
+      "loss": 1.1133,
+      "step": 25055
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 1.0347540378570557,
+      "learning_rate": 3.458445266024538e-05,
+      "loss": 1.0487,
+      "step": 25060
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 1.9431984424591064,
+      "learning_rate": 3.45790261010827e-05,
+      "loss": 1.0142,
+      "step": 25065
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 4.437554359436035,
+      "learning_rate": 3.4573599012880364e-05,
+      "loss": 1.1976,
+      "step": 25070
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 1.2235515117645264,
+      "learning_rate": 3.456817139593811e-05,
+      "loss": 1.1345,
+      "step": 25075
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 3.0580387115478516,
+      "learning_rate": 3.45627432505557e-05,
+      "loss": 1.2483,
+      "step": 25080
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 5.822119235992432,
+      "learning_rate": 3.4557314577032915e-05,
+      "loss": 1.0357,
+      "step": 25085
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 4.656231880187988,
+      "learning_rate": 3.455188537566957e-05,
+      "loss": 1.2377,
+      "step": 25090
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 1.6126227378845215,
+      "learning_rate": 3.4546455646765535e-05,
+      "loss": 1.3903,
+      "step": 25095
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 2.4167838096618652,
+      "learning_rate": 3.454102539062068e-05,
+      "loss": 1.1542,
+      "step": 25100
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 1.3699878454208374,
+      "learning_rate": 3.45355946075349e-05,
+      "loss": 1.1183,
+      "step": 25105
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 0.8811430335044861,
+      "learning_rate": 3.453016329780815e-05,
+      "loss": 1.1796,
+      "step": 25110
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 2.27093768119812,
+      "learning_rate": 3.452473146174038e-05,
+      "loss": 1.0819,
+      "step": 25115
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 2.351771116256714,
+      "learning_rate": 3.45192990996316e-05,
+      "loss": 1.2461,
+      "step": 25120
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 2.03128719329834,
+      "learning_rate": 3.451386621178182e-05,
+      "loss": 1.2742,
+      "step": 25125
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 6.1879448890686035,
+      "learning_rate": 3.45084327984911e-05,
+      "loss": 1.2229,
+      "step": 25130
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 2.5790390968322754,
+      "learning_rate": 3.4502998860059514e-05,
+      "loss": 1.0641,
+      "step": 25135
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 4.369030475616455,
+      "learning_rate": 3.4497564396787185e-05,
+      "loss": 1.3368,
+      "step": 25140
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 1.9594241380691528,
+      "learning_rate": 3.449212940897425e-05,
+      "loss": 1.1571,
+      "step": 25145
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 3.453566312789917,
+      "learning_rate": 3.4486693896920874e-05,
+      "loss": 1.1184,
+      "step": 25150
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 19.127910614013672,
+      "learning_rate": 3.448125786092725e-05,
+      "loss": 1.102,
+      "step": 25155
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 4.914632320404053,
+      "learning_rate": 3.447582130129361e-05,
+      "loss": 1.1391,
+      "step": 25160
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 3.1397294998168945,
+      "learning_rate": 3.4470384218320205e-05,
+      "loss": 1.0924,
+      "step": 25165
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 3.5343799591064453,
+      "learning_rate": 3.446494661230733e-05,
+      "loss": 1.0208,
+      "step": 25170
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 1.5835540294647217,
+      "learning_rate": 3.445950848355529e-05,
+      "loss": 1.0526,
+      "step": 25175
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 2.000260591506958,
+      "learning_rate": 3.445406983236443e-05,
+      "loss": 1.0166,
+      "step": 25180
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 2.638507843017578,
+      "learning_rate": 3.4448630659035126e-05,
+      "loss": 1.2353,
+      "step": 25185
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 3.0085575580596924,
+      "learning_rate": 3.444319096386777e-05,
+      "loss": 1.2625,
+      "step": 25190
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 3.4091453552246094,
+      "learning_rate": 3.4437750747162776e-05,
+      "loss": 0.9272,
+      "step": 25195
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 2.23293137550354,
+      "learning_rate": 3.443231000922063e-05,
+      "loss": 1.2352,
+      "step": 25200
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 1.904063105583191,
+      "learning_rate": 3.4426868750341805e-05,
+      "loss": 1.1851,
+      "step": 25205
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 3.55661678314209,
+      "learning_rate": 3.4421426970826826e-05,
+      "loss": 1.0767,
+      "step": 25210
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 2.7817044258117676,
+      "learning_rate": 3.441598467097622e-05,
+      "loss": 1.2299,
+      "step": 25215
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 1.7373440265655518,
+      "learning_rate": 3.441054185109057e-05,
+      "loss": 1.1791,
+      "step": 25220
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 3.398428440093994,
+      "learning_rate": 3.440509851147047e-05,
+      "loss": 1.0867,
+      "step": 25225
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 2.2971742153167725,
+      "learning_rate": 3.4399654652416566e-05,
+      "loss": 1.109,
+      "step": 25230
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 3.605184555053711,
+      "learning_rate": 3.4394210274229496e-05,
+      "loss": 1.1545,
+      "step": 25235
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 2.362635374069214,
+      "learning_rate": 3.4388765377209964e-05,
+      "loss": 1.1122,
+      "step": 25240
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 1.9791773557662964,
+      "learning_rate": 3.438331996165868e-05,
+      "loss": 1.2522,
+      "step": 25245
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 3.294543743133545,
+      "learning_rate": 3.437787402787639e-05,
+      "loss": 1.2164,
+      "step": 25250
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 3.8431739807128906,
+      "learning_rate": 3.4372427576163856e-05,
+      "loss": 1.108,
+      "step": 25255
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 4.042465686798096,
+      "learning_rate": 3.4366980606821895e-05,
+      "loss": 1.1426,
+      "step": 25260
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 1.2647730112075806,
+      "learning_rate": 3.436153312015133e-05,
+      "loss": 1.0345,
+      "step": 25265
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 1.1158740520477295,
+      "learning_rate": 3.435608511645302e-05,
+      "loss": 1.1481,
+      "step": 25270
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 1.05342698097229,
+      "learning_rate": 3.4350636596027853e-05,
+      "loss": 1.0948,
+      "step": 25275
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 2.2656328678131104,
+      "learning_rate": 3.434518755917675e-05,
+      "loss": 1.2436,
+      "step": 25280
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 4.6103901863098145,
+      "learning_rate": 3.433973800620065e-05,
+      "loss": 1.0903,
+      "step": 25285
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 2.1559901237487793,
+      "learning_rate": 3.4334287937400526e-05,
+      "loss": 1.0622,
+      "step": 25290
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 2.366748332977295,
+      "learning_rate": 3.432883735307739e-05,
+      "loss": 1.1494,
+      "step": 25295
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 2.5366220474243164,
+      "learning_rate": 3.4323386253532254e-05,
+      "loss": 1.0923,
+      "step": 25300
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 3.0016655921936035,
+      "learning_rate": 3.431793463906619e-05,
+      "loss": 1.0665,
+      "step": 25305
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 2.379429340362549,
+      "learning_rate": 3.431248250998028e-05,
+      "loss": 1.2129,
+      "step": 25310
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 3.5074679851531982,
+      "learning_rate": 3.4307029866575645e-05,
+      "loss": 1.1391,
+      "step": 25315
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 2.7369179725646973,
+      "learning_rate": 3.4301576709153424e-05,
+      "loss": 1.2312,
+      "step": 25320
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 2.9599127769470215,
+      "learning_rate": 3.4296123038014786e-05,
+      "loss": 1.218,
+      "step": 25325
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 1.0387095212936401,
+      "learning_rate": 3.429066885346094e-05,
+      "loss": 1.2556,
+      "step": 25330
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 2.1688525676727295,
+      "learning_rate": 3.42852141557931e-05,
+      "loss": 1.2108,
+      "step": 25335
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 2.080822229385376,
+      "learning_rate": 3.427975894531255e-05,
+      "loss": 1.175,
+      "step": 25340
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 1.2066692113876343,
+      "learning_rate": 3.427430322232055e-05,
+      "loss": 1.2416,
+      "step": 25345
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 1.1532701253890991,
+      "learning_rate": 3.4268846987118426e-05,
+      "loss": 1.171,
+      "step": 25350
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 1.9684667587280273,
+      "learning_rate": 3.426339024000751e-05,
+      "loss": 1.1804,
+      "step": 25355
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 1.9980390071868896,
+      "learning_rate": 3.4257932981289184e-05,
+      "loss": 1.1734,
+      "step": 25360
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 1.5516365766525269,
+      "learning_rate": 3.4252475211264846e-05,
+      "loss": 1.2121,
+      "step": 25365
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 1.7034841775894165,
+      "learning_rate": 3.424701693023591e-05,
+      "loss": 1.1121,
+      "step": 25370
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 3.420742988586426,
+      "learning_rate": 3.424155813850385e-05,
+      "loss": 1.07,
+      "step": 25375
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 2.169203519821167,
+      "learning_rate": 3.423609883637014e-05,
+      "loss": 1.1631,
+      "step": 25380
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 4.135627746582031,
+      "learning_rate": 3.423063902413629e-05,
+      "loss": 1.2078,
+      "step": 25385
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 3.5442159175872803,
+      "learning_rate": 3.422517870210384e-05,
+      "loss": 1.1549,
+      "step": 25390
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 5.373948574066162,
+      "learning_rate": 3.421971787057436e-05,
+      "loss": 0.976,
+      "step": 25395
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 2.204765558242798,
+      "learning_rate": 3.421425652984944e-05,
+      "loss": 1.2897,
+      "step": 25400
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 4.4537811279296875,
+      "learning_rate": 3.420879468023072e-05,
+      "loss": 1.0213,
+      "step": 25405
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 2.6348259449005127,
+      "learning_rate": 3.4203332322019835e-05,
+      "loss": 1.1508,
+      "step": 25410
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 1.4462038278579712,
+      "learning_rate": 3.419786945551848e-05,
+      "loss": 1.2851,
+      "step": 25415
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 4.519046783447266,
+      "learning_rate": 3.419240608102834e-05,
+      "loss": 1.2382,
+      "step": 25420
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 2.593632936477661,
+      "learning_rate": 3.418694219885118e-05,
+      "loss": 1.1794,
+      "step": 25425
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 2.167607545852661,
+      "learning_rate": 3.418147780928875e-05,
+      "loss": 1.2187,
+      "step": 25430
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 3.661607027053833,
+      "learning_rate": 3.4176012912642844e-05,
+      "loss": 1.0878,
+      "step": 25435
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 2.0092809200286865,
+      "learning_rate": 3.4170547509215286e-05,
+      "loss": 1.134,
+      "step": 25440
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 7.421758651733398,
+      "learning_rate": 3.416508159930791e-05,
+      "loss": 0.9357,
+      "step": 25445
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 4.737497329711914,
+      "learning_rate": 3.415961518322262e-05,
+      "loss": 1.0698,
+      "step": 25450
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 2.214827299118042,
+      "learning_rate": 3.4154148261261285e-05,
+      "loss": 1.051,
+      "step": 25455
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 1.7568135261535645,
+      "learning_rate": 3.414868083372587e-05,
+      "loss": 1.2068,
+      "step": 25460
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 1.6030669212341309,
+      "learning_rate": 3.414321290091831e-05,
+      "loss": 1.1984,
+      "step": 25465
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 1.4295237064361572,
+      "learning_rate": 3.413774446314062e-05,
+      "loss": 1.1206,
+      "step": 25470
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 5.009536266326904,
+      "learning_rate": 3.41322755206948e-05,
+      "loss": 1.0359,
+      "step": 25475
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 1.4227309226989746,
+      "learning_rate": 3.4126806073882886e-05,
+      "loss": 1.1036,
+      "step": 25480
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 1.6048041582107544,
+      "learning_rate": 3.4121336123006965e-05,
+      "loss": 1.3853,
+      "step": 25485
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 1.398371696472168,
+      "learning_rate": 3.411586566836913e-05,
+      "loss": 1.076,
+      "step": 25490
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 1.4785361289978027,
+      "learning_rate": 3.4110394710271504e-05,
+      "loss": 1.1772,
+      "step": 25495
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 2.355851411819458,
+      "learning_rate": 3.410492324901626e-05,
+      "loss": 1.1964,
+      "step": 25500
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 1.825348973274231,
+      "learning_rate": 3.4099451284905556e-05,
+      "loss": 1.2284,
+      "step": 25505
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 1.8260172605514526,
+      "learning_rate": 3.409397881824163e-05,
+      "loss": 1.2915,
+      "step": 25510
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 2.6344926357269287,
+      "learning_rate": 3.4088505849326697e-05,
+      "loss": 1.1483,
+      "step": 25515
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 1.9603842496871948,
+      "learning_rate": 3.408303237846303e-05,
+      "loss": 1.1863,
+      "step": 25520
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 1.3052462339401245,
+      "learning_rate": 3.407755840595294e-05,
+      "loss": 1.2192,
+      "step": 25525
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 3.303022623062134,
+      "learning_rate": 3.407208393209872e-05,
+      "loss": 1.2268,
+      "step": 25530
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 1.622067928314209,
+      "learning_rate": 3.406660895720275e-05,
+      "loss": 1.2071,
+      "step": 25535
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 3.006840467453003,
+      "learning_rate": 3.406113348156738e-05,
+      "loss": 1.1447,
+      "step": 25540
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 2.42177152633667,
+      "learning_rate": 3.405565750549502e-05,
+      "loss": 1.1148,
+      "step": 25545
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 2.3825762271881104,
+      "learning_rate": 3.405018102928812e-05,
+      "loss": 1.2127,
+      "step": 25550
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 2.4480338096618652,
+      "learning_rate": 3.404470405324912e-05,
+      "loss": 1.0913,
+      "step": 25555
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 1.3634631633758545,
+      "learning_rate": 3.403922657768052e-05,
+      "loss": 1.2583,
+      "step": 25560
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 3.5564489364624023,
+      "learning_rate": 3.403374860288484e-05,
+      "loss": 1.14,
+      "step": 25565
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 2.949016571044922,
+      "learning_rate": 3.402827012916461e-05,
+      "loss": 1.3249,
+      "step": 25570
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 1.4162577390670776,
+      "learning_rate": 3.4022791156822395e-05,
+      "loss": 1.2632,
+      "step": 25575
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 1.9798351526260376,
+      "learning_rate": 3.401731168616081e-05,
+      "loss": 1.3101,
+      "step": 25580
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 1.9318190813064575,
+      "learning_rate": 3.401183171748248e-05,
+      "loss": 1.1082,
+      "step": 25585
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 2.191795825958252,
+      "learning_rate": 3.400635125109005e-05,
+      "loss": 1.2841,
+      "step": 25590
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 3.378875732421875,
+      "learning_rate": 3.40008702872862e-05,
+      "loss": 0.9573,
+      "step": 25595
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 3.725365400314331,
+      "learning_rate": 3.399538882637364e-05,
+      "loss": 1.1023,
+      "step": 25600
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 2.4383790493011475,
+      "learning_rate": 3.3989906868655104e-05,
+      "loss": 1.1758,
+      "step": 25605
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 3.5799477100372314,
+      "learning_rate": 3.398442441443336e-05,
+      "loss": 1.1651,
+      "step": 25610
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 3.4080264568328857,
+      "learning_rate": 3.397894146401118e-05,
+      "loss": 1.0039,
+      "step": 25615
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 1.794622540473938,
+      "learning_rate": 3.397345801769141e-05,
+      "loss": 1.1809,
+      "step": 25620
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 2.9816510677337646,
+      "learning_rate": 3.3967974075776875e-05,
+      "loss": 1.263,
+      "step": 25625
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 2.4955384731292725,
+      "learning_rate": 3.3962489638570464e-05,
+      "loss": 1.151,
+      "step": 25630
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 2.3042807579040527,
+      "learning_rate": 3.395700470637506e-05,
+      "loss": 1.0265,
+      "step": 25635
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 1.0877710580825806,
+      "learning_rate": 3.3951519279493585e-05,
+      "loss": 1.2262,
+      "step": 25640
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 2.408581018447876,
+      "learning_rate": 3.394603335822902e-05,
+      "loss": 1.1331,
+      "step": 25645
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 1.5194898843765259,
+      "learning_rate": 3.3940546942884324e-05,
+      "loss": 1.1822,
+      "step": 25650
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 1.858493447303772,
+      "learning_rate": 3.393506003376251e-05,
+      "loss": 1.0194,
+      "step": 25655
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 1.8844908475875854,
+      "learning_rate": 3.392957263116663e-05,
+      "loss": 1.1915,
+      "step": 25660
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 2.1716675758361816,
+      "learning_rate": 3.392408473539973e-05,
+      "loss": 1.1354,
+      "step": 25665
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 5.164560317993164,
+      "learning_rate": 3.391859634676491e-05,
+      "loss": 1.1757,
+      "step": 25670
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 3.698439359664917,
+      "learning_rate": 3.3913107465565274e-05,
+      "loss": 1.0446,
+      "step": 25675
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 7.462309837341309,
+      "learning_rate": 3.390761809210398e-05,
+      "loss": 1.1909,
+      "step": 25680
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 3.320486545562744,
+      "learning_rate": 3.39021282266842e-05,
+      "loss": 1.1427,
+      "step": 25685
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 1.506848931312561,
+      "learning_rate": 3.389663786960913e-05,
+      "loss": 1.2685,
+      "step": 25690
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 0.9463053345680237,
+      "learning_rate": 3.3891147021182004e-05,
+      "loss": 1.2191,
+      "step": 25695
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 1.4177323579788208,
+      "learning_rate": 3.388565568170607e-05,
+      "loss": 1.1461,
+      "step": 25700
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 1.9486898183822632,
+      "learning_rate": 3.38801638514846e-05,
+      "loss": 1.064,
+      "step": 25705
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 3.4581215381622314,
+      "learning_rate": 3.3874671530820915e-05,
+      "loss": 1.1031,
+      "step": 25710
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 2.35516095161438,
+      "learning_rate": 3.386917872001835e-05,
+      "loss": 1.1787,
+      "step": 25715
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 5.054357528686523,
+      "learning_rate": 3.3863685419380254e-05,
+      "loss": 1.2588,
+      "step": 25720
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 2.7185957431793213,
+      "learning_rate": 3.385819162921003e-05,
+      "loss": 0.9498,
+      "step": 25725
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 1.8890604972839355,
+      "learning_rate": 3.385269734981109e-05,
+      "loss": 1.2021,
+      "step": 25730
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 2.1678149700164795,
+      "learning_rate": 3.384720258148688e-05,
+      "loss": 1.1627,
+      "step": 25735
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 2.4753189086914062,
+      "learning_rate": 3.384170732454087e-05,
+      "loss": 1.2426,
+      "step": 25740
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 1.5026155710220337,
+      "learning_rate": 3.3836211579276546e-05,
+      "loss": 1.1725,
+      "step": 25745
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 2.9787325859069824,
+      "learning_rate": 3.3830715345997444e-05,
+      "loss": 1.2786,
+      "step": 25750
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 1.6327831745147705,
+      "learning_rate": 3.382521862500712e-05,
+      "loss": 1.1652,
+      "step": 25755
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 3.5435101985931396,
+      "learning_rate": 3.381972141660914e-05,
+      "loss": 1.08,
+      "step": 25760
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 2.974608898162842,
+      "learning_rate": 3.381422372110711e-05,
+      "loss": 1.2794,
+      "step": 25765
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 1.1757948398590088,
+      "learning_rate": 3.3808725538804667e-05,
+      "loss": 1.1428,
+      "step": 25770
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 2.5623114109039307,
+      "learning_rate": 3.380322687000547e-05,
+      "loss": 1.1777,
+      "step": 25775
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 1.8030344247817993,
+      "learning_rate": 3.37977277150132e-05,
+      "loss": 1.3782,
+      "step": 25780
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 2.135564088821411,
+      "learning_rate": 3.379222807413158e-05,
+      "loss": 1.2467,
+      "step": 25785
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 6.063063144683838,
+      "learning_rate": 3.3786727947664344e-05,
+      "loss": 1.1731,
+      "step": 25790
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 2.186962842941284,
+      "learning_rate": 3.378122733591525e-05,
+      "loss": 1.2181,
+      "step": 25795
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 1.757891058921814,
+      "learning_rate": 3.37757262391881e-05,
+      "loss": 1.151,
+      "step": 25800
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 17.243751525878906,
+      "learning_rate": 3.377022465778671e-05,
+      "loss": 1.1469,
+      "step": 25805
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 3.9934499263763428,
+      "learning_rate": 3.376472259201493e-05,
+      "loss": 1.055,
+      "step": 25810
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 1.4664825201034546,
+      "learning_rate": 3.375922004217663e-05,
+      "loss": 1.2219,
+      "step": 25815
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 2.3855884075164795,
+      "learning_rate": 3.3753717008575716e-05,
+      "loss": 1.2283,
+      "step": 25820
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 3.64817476272583,
+      "learning_rate": 3.374821349151611e-05,
+      "loss": 1.0902,
+      "step": 25825
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 5.835208415985107,
+      "learning_rate": 3.374270949130176e-05,
+      "loss": 1.1913,
+      "step": 25830
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 8.944743156433105,
+      "learning_rate": 3.373720500823666e-05,
+      "loss": 1.1375,
+      "step": 25835
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 1.3471863269805908,
+      "learning_rate": 3.37317000426248e-05,
+      "loss": 1.1213,
+      "step": 25840
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 1.5816328525543213,
+      "learning_rate": 3.3726194594770224e-05,
+      "loss": 1.3163,
+      "step": 25845
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 3.4675426483154297,
+      "learning_rate": 3.3720688664976996e-05,
+      "loss": 1.1634,
+      "step": 25850
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 1.961869478225708,
+      "learning_rate": 3.3715182253549205e-05,
+      "loss": 1.1437,
+      "step": 25855
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 3.2622334957122803,
+      "learning_rate": 3.3709675360790945e-05,
+      "loss": 1.1406,
+      "step": 25860
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 2.293381690979004,
+      "learning_rate": 3.370416798700637e-05,
+      "loss": 1.314,
+      "step": 25865
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 3.0880744457244873,
+      "learning_rate": 3.369866013249965e-05,
+      "loss": 1.2104,
+      "step": 25870
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 0.9818011522293091,
+      "learning_rate": 3.369315179757496e-05,
+      "loss": 1.205,
+      "step": 25875
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 1.3982211351394653,
+      "learning_rate": 3.368764298253654e-05,
+      "loss": 1.1116,
+      "step": 25880
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 2.2384934425354004,
+      "learning_rate": 3.368213368768863e-05,
+      "loss": 1.1532,
+      "step": 25885
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 2.110853433609009,
+      "learning_rate": 3.3676623913335507e-05,
+      "loss": 1.3154,
+      "step": 25890
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 5.248104095458984,
+      "learning_rate": 3.367111365978146e-05,
+      "loss": 1.187,
+      "step": 25895
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 2.6548986434936523,
+      "learning_rate": 3.3665602927330814e-05,
+      "loss": 1.0928,
+      "step": 25900
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 2.44582200050354,
+      "learning_rate": 3.3660091716287925e-05,
+      "loss": 1.2958,
+      "step": 25905
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 2.185173273086548,
+      "learning_rate": 3.3654580026957176e-05,
+      "loss": 1.0973,
+      "step": 25910
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 4.19801664352417,
+      "learning_rate": 3.364906785964297e-05,
+      "loss": 1.2329,
+      "step": 25915
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 2.9229116439819336,
+      "learning_rate": 3.364355521464974e-05,
+      "loss": 1.0089,
+      "step": 25920
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 3.5953376293182373,
+      "learning_rate": 3.363804209228192e-05,
+      "loss": 1.1683,
+      "step": 25925
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 3.6482677459716797,
+      "learning_rate": 3.363252849284404e-05,
+      "loss": 1.1713,
+      "step": 25930
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 3.2520852088928223,
+      "learning_rate": 3.3627014416640565e-05,
+      "loss": 1.4132,
+      "step": 25935
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 3.2956299781799316,
+      "learning_rate": 3.362149986397606e-05,
+      "loss": 1.3033,
+      "step": 25940
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 2.4566397666931152,
+      "learning_rate": 3.361598483515507e-05,
+      "loss": 1.1382,
+      "step": 25945
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 2.1512653827667236,
+      "learning_rate": 3.3610469330482205e-05,
+      "loss": 1.0305,
+      "step": 25950
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 2.5325844287872314,
+      "learning_rate": 3.360495335026207e-05,
+      "loss": 1.0878,
+      "step": 25955
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 10.988487243652344,
+      "learning_rate": 3.35994368947993e-05,
+      "loss": 1.0437,
+      "step": 25960
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 1.2049709558486938,
+      "learning_rate": 3.359391996439857e-05,
+      "loss": 1.2115,
+      "step": 25965
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 4.1528143882751465,
+      "learning_rate": 3.358840255936457e-05,
+      "loss": 1.2731,
+      "step": 25970
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 1.840437889099121,
+      "learning_rate": 3.3582884680002024e-05,
+      "loss": 1.1867,
+      "step": 25975
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 1.5152828693389893,
+      "learning_rate": 3.3577366326615676e-05,
+      "loss": 1.297,
+      "step": 25980
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 2.739409923553467,
+      "learning_rate": 3.357184749951031e-05,
+      "loss": 1.205,
+      "step": 25985
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 2.3803489208221436,
+      "learning_rate": 3.3566328198990713e-05,
+      "loss": 0.934,
+      "step": 25990
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 1.749653697013855,
+      "learning_rate": 3.356080842536171e-05,
+      "loss": 1.246,
+      "step": 25995
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 1.9984244108200073,
+      "learning_rate": 3.355528817892816e-05,
+      "loss": 1.162,
+      "step": 26000
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 1.7096799612045288,
+      "learning_rate": 3.354976745999494e-05,
+      "loss": 1.1473,
+      "step": 26005
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 2.5412979125976562,
+      "learning_rate": 3.354424626886694e-05,
+      "loss": 1.2047,
+      "step": 26010
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 1.5412076711654663,
+      "learning_rate": 3.3538724605849115e-05,
+      "loss": 1.2091,
+      "step": 26015
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 1.5192663669586182,
+      "learning_rate": 3.353320247124639e-05,
+      "loss": 1.2168,
+      "step": 26020
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 1.3354765176773071,
+      "learning_rate": 3.352767986536377e-05,
+      "loss": 1.177,
+      "step": 26025
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 2.89026141166687,
+      "learning_rate": 3.352215678850625e-05,
+      "loss": 1.0474,
+      "step": 26030
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 1.4622420072555542,
+      "learning_rate": 3.351663324097888e-05,
+      "loss": 1.0757,
+      "step": 26035
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 3.365771532058716,
+      "learning_rate": 3.35111092230867e-05,
+      "loss": 1.1737,
+      "step": 26040
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 8.960780143737793,
+      "learning_rate": 3.35055847351348e-05,
+      "loss": 1.1499,
+      "step": 26045
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 3.380896806716919,
+      "learning_rate": 3.350005977742831e-05,
+      "loss": 1.1441,
+      "step": 26050
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 2.153420925140381,
+      "learning_rate": 3.3494534350272344e-05,
+      "loss": 1.1652,
+      "step": 26055
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 3.0268092155456543,
+      "learning_rate": 3.348900845397208e-05,
+      "loss": 1.3158,
+      "step": 26060
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 5.089029788970947,
+      "learning_rate": 3.34834820888327e-05,
+      "loss": 1.093,
+      "step": 26065
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 1.8281073570251465,
+      "learning_rate": 3.347795525515942e-05,
+      "loss": 1.2282,
+      "step": 26070
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 2.1014137268066406,
+      "learning_rate": 3.347242795325749e-05,
+      "loss": 1.1495,
+      "step": 26075
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 2.044297933578491,
+      "learning_rate": 3.3466900183432164e-05,
+      "loss": 1.0403,
+      "step": 26080
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 2.4856858253479004,
+      "learning_rate": 3.346137194598874e-05,
+      "loss": 1.1332,
+      "step": 26085
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 2.4128828048706055,
+      "learning_rate": 3.345584324123254e-05,
+      "loss": 1.0393,
+      "step": 26090
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 1.2683279514312744,
+      "learning_rate": 3.345031406946891e-05,
+      "loss": 1.2443,
+      "step": 26095
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 2.806605577468872,
+      "learning_rate": 3.344478443100322e-05,
+      "loss": 1.1122,
+      "step": 26100
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 2.0906083583831787,
+      "learning_rate": 3.343925432614086e-05,
+      "loss": 1.0038,
+      "step": 26105
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 2.3186562061309814,
+      "learning_rate": 3.3433723755187255e-05,
+      "loss": 1.1656,
+      "step": 26110
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 2.0462429523468018,
+      "learning_rate": 3.342819271844787e-05,
+      "loss": 1.1485,
+      "step": 26115
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 1.9167912006378174,
+      "learning_rate": 3.342266121622814e-05,
+      "loss": 1.2493,
+      "step": 26120
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 3.755826473236084,
+      "learning_rate": 3.3417129248833596e-05,
+      "loss": 1.1989,
+      "step": 26125
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 3.0041463375091553,
+      "learning_rate": 3.3411596816569746e-05,
+      "loss": 1.1634,
+      "step": 26130
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 3.452164649963379,
+      "learning_rate": 3.340606391974215e-05,
+      "loss": 1.2504,
+      "step": 26135
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 6.089219093322754,
+      "learning_rate": 3.340053055865639e-05,
+      "loss": 1.0353,
+      "step": 26140
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 0.934475839138031,
+      "learning_rate": 3.339499673361805e-05,
+      "loss": 1.2358,
+      "step": 26145
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 7.745322227478027,
+      "learning_rate": 3.3389462444932765e-05,
+      "loss": 1.1976,
+      "step": 26150
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 1.4742351770401,
+      "learning_rate": 3.338392769290619e-05,
+      "loss": 1.2129,
+      "step": 26155
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 1.67327880859375,
+      "learning_rate": 3.337839247784401e-05,
+      "loss": 1.1415,
+      "step": 26160
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 4.258670806884766,
+      "learning_rate": 3.337285680005192e-05,
+      "loss": 1.0752,
+      "step": 26165
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 2.024472236633301,
+      "learning_rate": 3.336732065983565e-05,
+      "loss": 1.0806,
+      "step": 26170
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 1.1518867015838623,
+      "learning_rate": 3.336178405750095e-05,
+      "loss": 1.2075,
+      "step": 26175
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 3.1105754375457764,
+      "learning_rate": 3.3356246993353617e-05,
+      "loss": 1.1183,
+      "step": 26180
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 1.1292957067489624,
+      "learning_rate": 3.335070946769945e-05,
+      "loss": 1.381,
+      "step": 26185
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 1.872825264930725,
+      "learning_rate": 3.3345171480844275e-05,
+      "loss": 1.0752,
+      "step": 26190
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 1.4173269271850586,
+      "learning_rate": 3.3339633033093955e-05,
+      "loss": 1.0798,
+      "step": 26195
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 3.544238567352295,
+      "learning_rate": 3.333409412475437e-05,
+      "loss": 1.2383,
+      "step": 26200
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 2.445939064025879,
+      "learning_rate": 3.3328554756131423e-05,
+      "loss": 1.1121,
+      "step": 26205
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 2.592607021331787,
+      "learning_rate": 3.332301492753107e-05,
+      "loss": 1.0819,
+      "step": 26210
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 2.417942523956299,
+      "learning_rate": 3.3317474639259245e-05,
+      "loss": 1.128,
+      "step": 26215
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 1.7295610904693604,
+      "learning_rate": 3.331193389162194e-05,
+      "loss": 1.18,
+      "step": 26220
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 7.857126235961914,
+      "learning_rate": 3.330639268492517e-05,
+      "loss": 0.9577,
+      "step": 26225
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 1.7152934074401855,
+      "learning_rate": 3.330085101947496e-05,
+      "loss": 1.2462,
+      "step": 26230
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 1.7840052843093872,
+      "learning_rate": 3.3295308895577376e-05,
+      "loss": 0.9696,
+      "step": 26235
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 1.5356841087341309,
+      "learning_rate": 3.32897663135385e-05,
+      "loss": 1.1286,
+      "step": 26240
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 1.8642561435699463,
+      "learning_rate": 3.3284223273664465e-05,
+      "loss": 0.9585,
+      "step": 26245
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 1.54176664352417,
+      "learning_rate": 3.327867977626138e-05,
+      "loss": 1.177,
+      "step": 26250
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 3.919142961502075,
+      "learning_rate": 3.327313582163542e-05,
+      "loss": 1.1158,
+      "step": 26255
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 2.54856276512146,
+      "learning_rate": 3.326759141009276e-05,
+      "loss": 1.1494,
+      "step": 26260
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 2.491251230239868,
+      "learning_rate": 3.326204654193962e-05,
+      "loss": 1.2343,
+      "step": 26265
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 4.712347030639648,
+      "learning_rate": 3.325650121748225e-05,
+      "loss": 1.1057,
+      "step": 26270
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 1.5245498418807983,
+      "learning_rate": 3.325095543702688e-05,
+      "loss": 1.1575,
+      "step": 26275
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 3.232236385345459,
+      "learning_rate": 3.324540920087983e-05,
+      "loss": 1.1603,
+      "step": 26280
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 1.0489842891693115,
+      "learning_rate": 3.3239862509347396e-05,
+      "loss": 1.2751,
+      "step": 26285
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 1.4679367542266846,
+      "learning_rate": 3.3234315362735926e-05,
+      "loss": 1.0614,
+      "step": 26290
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 4.200760364532471,
+      "learning_rate": 3.3228767761351776e-05,
+      "loss": 1.251,
+      "step": 26295
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 3.265259265899658,
+      "learning_rate": 3.3223219705501334e-05,
+      "loss": 1.0798,
+      "step": 26300
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 5.21251106262207,
+      "learning_rate": 3.3217671195491016e-05,
+      "loss": 1.0074,
+      "step": 26305
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 2.8614084720611572,
+      "learning_rate": 3.3212122231627265e-05,
+      "loss": 0.9399,
+      "step": 26310
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 2.9637887477874756,
+      "learning_rate": 3.320657281421653e-05,
+      "loss": 1.1851,
+      "step": 26315
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 2.9170193672180176,
+      "learning_rate": 3.320102294356531e-05,
+      "loss": 1.3154,
+      "step": 26320
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 4.612870216369629,
+      "learning_rate": 3.319547261998012e-05,
+      "loss": 1.0731,
+      "step": 26325
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 3.057469129562378,
+      "learning_rate": 3.31899218437675e-05,
+      "loss": 1.0565,
+      "step": 26330
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 3.9015371799468994,
+      "learning_rate": 3.3184370615234004e-05,
+      "loss": 1.0798,
+      "step": 26335
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 2.964965343475342,
+      "learning_rate": 3.317881893468623e-05,
+      "loss": 1.0914,
+      "step": 26340
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 7.357722282409668,
+      "learning_rate": 3.3173266802430784e-05,
+      "loss": 1.3035,
+      "step": 26345
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 1.9388129711151123,
+      "learning_rate": 3.3167714218774315e-05,
+      "loss": 1.1861,
+      "step": 26350
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 1.838700532913208,
+      "learning_rate": 3.316216118402347e-05,
+      "loss": 1.0244,
+      "step": 26355
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 2.831838846206665,
+      "learning_rate": 3.315660769848495e-05,
+      "loss": 1.0736,
+      "step": 26360
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 4.200833320617676,
+      "learning_rate": 3.315105376246547e-05,
+      "loss": 1.2928,
+      "step": 26365
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 6.066209316253662,
+      "learning_rate": 3.3145499376271754e-05,
+      "loss": 1.0764,
+      "step": 26370
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 1.2264248132705688,
+      "learning_rate": 3.3139944540210574e-05,
+      "loss": 1.2454,
+      "step": 26375
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 1.508028268814087,
+      "learning_rate": 3.3134389254588724e-05,
+      "loss": 1.0258,
+      "step": 26380
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 2.4264774322509766,
+      "learning_rate": 3.312883351971301e-05,
+      "loss": 1.0888,
+      "step": 26385
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 3.008584499359131,
+      "learning_rate": 3.312327733589027e-05,
+      "loss": 1.2136,
+      "step": 26390
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 2.624093532562256,
+      "learning_rate": 3.3117720703427365e-05,
+      "loss": 1.0683,
+      "step": 26395
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 2.020382881164551,
+      "learning_rate": 3.311216362263119e-05,
+      "loss": 1.283,
+      "step": 26400
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 1.6679294109344482,
+      "learning_rate": 3.310660609380865e-05,
+      "loss": 1.193,
+      "step": 26405
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 1.7250815629959106,
+      "learning_rate": 3.3101048117266675e-05,
+      "loss": 1.1785,
+      "step": 26410
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 3.259265661239624,
+      "learning_rate": 3.3095489693312234e-05,
+      "loss": 1.1396,
+      "step": 26415
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 6.142773151397705,
+      "learning_rate": 3.308993082225231e-05,
+      "loss": 1.0735,
+      "step": 26420
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 2.9731175899505615,
+      "learning_rate": 3.308437150439392e-05,
+      "loss": 1.1468,
+      "step": 26425
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 2.0407536029815674,
+      "learning_rate": 3.3078811740044096e-05,
+      "loss": 1.3368,
+      "step": 26430
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 2.396974563598633,
+      "learning_rate": 3.30732515295099e-05,
+      "loss": 1.0735,
+      "step": 26435
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 1.7831975221633911,
+      "learning_rate": 3.306769087309841e-05,
+      "loss": 1.212,
+      "step": 26440
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 2.968552827835083,
+      "learning_rate": 3.3062129771116734e-05,
+      "loss": 1.0222,
+      "step": 26445
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 2.5638179779052734,
+      "learning_rate": 3.305656822387201e-05,
+      "loss": 1.1809,
+      "step": 26450
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 11.847745895385742,
+      "learning_rate": 3.30510062316714e-05,
+      "loss": 1.2684,
+      "step": 26455
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 2.3828535079956055,
+      "learning_rate": 3.304544379482209e-05,
+      "loss": 1.2027,
+      "step": 26460
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 3.3371334075927734,
+      "learning_rate": 3.303988091363128e-05,
+      "loss": 1.2209,
+      "step": 26465
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 1.5403908491134644,
+      "learning_rate": 3.3034317588406205e-05,
+      "loss": 1.0056,
+      "step": 26470
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 3.629544973373413,
+      "learning_rate": 3.302875381945412e-05,
+      "loss": 1.2427,
+      "step": 26475
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 1.5579273700714111,
+      "learning_rate": 3.30231896070823e-05,
+      "loss": 1.2536,
+      "step": 26480
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 3.202047824859619,
+      "learning_rate": 3.3017624951598066e-05,
+      "loss": 1.0117,
+      "step": 26485
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 4.338261127471924,
+      "learning_rate": 3.301205985330873e-05,
+      "loss": 0.9417,
+      "step": 26490
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.7901287078857422,
+      "learning_rate": 3.300649431252166e-05,
+      "loss": 1.005,
+      "step": 26495
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 2.2430081367492676,
+      "learning_rate": 3.300092832954425e-05,
+      "loss": 1.2206,
+      "step": 26500
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 2.6997382640838623,
+      "learning_rate": 3.2995361904683866e-05,
+      "loss": 1.3153,
+      "step": 26505
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 4.912943363189697,
+      "learning_rate": 3.2989795038247956e-05,
+      "loss": 1.1331,
+      "step": 26510
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 1.721016526222229,
+      "learning_rate": 3.298422773054397e-05,
+      "loss": 1.3072,
+      "step": 26515
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.9054770469665527,
+      "learning_rate": 3.297865998187939e-05,
+      "loss": 1.2066,
+      "step": 26520
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 1.856542706489563,
+      "learning_rate": 3.297309179256171e-05,
+      "loss": 1.2847,
+      "step": 26525
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 2.9849486351013184,
+      "learning_rate": 3.2967523162898465e-05,
+      "loss": 1.2153,
+      "step": 26530
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 2.4759957790374756,
+      "learning_rate": 3.296195409319719e-05,
+      "loss": 1.1859,
+      "step": 26535
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 1.966263771057129,
+      "learning_rate": 3.295638458376546e-05,
+      "loss": 1.2078,
+      "step": 26540
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 3.476435422897339,
+      "learning_rate": 3.295081463491089e-05,
+      "loss": 1.1706,
+      "step": 26545
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 3.4568331241607666,
+      "learning_rate": 3.294524424694109e-05,
+      "loss": 1.2634,
+      "step": 26550
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 1.4559932947158813,
+      "learning_rate": 3.2939673420163706e-05,
+      "loss": 1.0625,
+      "step": 26555
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 1.8391364812850952,
+      "learning_rate": 3.293410215488642e-05,
+      "loss": 1.1624,
+      "step": 26560
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 2.4696767330169678,
+      "learning_rate": 3.292853045141691e-05,
+      "loss": 1.2677,
+      "step": 26565
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 1.480440378189087,
+      "learning_rate": 3.2922958310062904e-05,
+      "loss": 1.0199,
+      "step": 26570
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 1.3950989246368408,
+      "learning_rate": 3.291738573113216e-05,
+      "loss": 1.2905,
+      "step": 26575
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 2.3112564086914062,
+      "learning_rate": 3.291181271493242e-05,
+      "loss": 1.1195,
+      "step": 26580
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 5.1455302238464355,
+      "learning_rate": 3.290623926177148e-05,
+      "loss": 1.1371,
+      "step": 26585
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 7.398214340209961,
+      "learning_rate": 3.290066537195717e-05,
+      "loss": 1.1394,
+      "step": 26590
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 3.4421448707580566,
+      "learning_rate": 3.2895091045797335e-05,
+      "loss": 1.372,
+      "step": 26595
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 5.3825764656066895,
+      "learning_rate": 3.288951628359982e-05,
+      "loss": 1.3115,
+      "step": 26600
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 2.211780548095703,
+      "learning_rate": 3.288394108567252e-05,
+      "loss": 1.1863,
+      "step": 26605
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 4.291854381561279,
+      "learning_rate": 3.287836545232335e-05,
+      "loss": 1.1253,
+      "step": 26610
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 2.0238168239593506,
+      "learning_rate": 3.2872789383860246e-05,
+      "loss": 1.34,
+      "step": 26615
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 2.0490705966949463,
+      "learning_rate": 3.286721288059116e-05,
+      "loss": 1.0287,
+      "step": 26620
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 2.1137607097625732,
+      "learning_rate": 3.28616359428241e-05,
+      "loss": 0.9508,
+      "step": 26625
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 3.983616352081299,
+      "learning_rate": 3.285605857086704e-05,
+      "loss": 1.3123,
+      "step": 26630
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 2.677525520324707,
+      "learning_rate": 3.285048076502805e-05,
+      "loss": 0.9677,
+      "step": 26635
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 3.7401962280273438,
+      "learning_rate": 3.284490252561515e-05,
+      "loss": 1.2097,
+      "step": 26640
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 2.0874180793762207,
+      "learning_rate": 3.283932385293644e-05,
+      "loss": 0.9952,
+      "step": 26645
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 2.5544826984405518,
+      "learning_rate": 3.283374474730003e-05,
+      "loss": 1.1315,
+      "step": 26650
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 1.7675901651382446,
+      "learning_rate": 3.2828165209014036e-05,
+      "loss": 1.1583,
+      "step": 26655
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 3.3144235610961914,
+      "learning_rate": 3.282258523838663e-05,
+      "loss": 1.1255,
+      "step": 26660
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 3.3691461086273193,
+      "learning_rate": 3.281700483572595e-05,
+      "loss": 1.2332,
+      "step": 26665
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 3.544421911239624,
+      "learning_rate": 3.281142400134023e-05,
+      "loss": 1.0824,
+      "step": 26670
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 1.068796157836914,
+      "learning_rate": 3.280584273553768e-05,
+      "loss": 1.14,
+      "step": 26675
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 2.5210955142974854,
+      "learning_rate": 3.2800261038626544e-05,
+      "loss": 1.1527,
+      "step": 26680
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 2.1437249183654785,
+      "learning_rate": 3.279467891091511e-05,
+      "loss": 0.9096,
+      "step": 26685
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 1.478752613067627,
+      "learning_rate": 3.278909635271165e-05,
+      "loss": 1.0211,
+      "step": 26690
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 4.072501182556152,
+      "learning_rate": 3.27835133643245e-05,
+      "loss": 1.1708,
+      "step": 26695
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 3.866422653198242,
+      "learning_rate": 3.2777929946062005e-05,
+      "loss": 1.1402,
+      "step": 26700
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 2.2160439491271973,
+      "learning_rate": 3.277234609823251e-05,
+      "loss": 1.2121,
+      "step": 26705
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 3.4925968647003174,
+      "learning_rate": 3.276676182114443e-05,
+      "loss": 1.188,
+      "step": 26710
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 1.9253541231155396,
+      "learning_rate": 3.276117711510616e-05,
+      "loss": 1.1559,
+      "step": 26715
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 1.081268072128296,
+      "learning_rate": 3.2755591980426146e-05,
+      "loss": 1.2158,
+      "step": 26720
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.8544811010360718,
+      "learning_rate": 3.275000641741285e-05,
+      "loss": 1.1147,
+      "step": 26725
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.938436508178711,
+      "learning_rate": 3.2744420426374755e-05,
+      "loss": 1.2673,
+      "step": 26730
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 2.7037627696990967,
+      "learning_rate": 3.273883400762037e-05,
+      "loss": 1.2175,
+      "step": 26735
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 2.8835654258728027,
+      "learning_rate": 3.2733247161458224e-05,
+      "loss": 1.1878,
+      "step": 26740
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.2548410892486572,
+      "learning_rate": 3.272765988819688e-05,
+      "loss": 1.2278,
+      "step": 26745
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 2.562904119491577,
+      "learning_rate": 3.2722072188144916e-05,
+      "loss": 1.1261,
+      "step": 26750
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.355686902999878,
+      "learning_rate": 3.271648406161092e-05,
+      "loss": 0.9932,
+      "step": 26755
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 2.448178768157959,
+      "learning_rate": 3.2710895508903546e-05,
+      "loss": 1.1223,
+      "step": 26760
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 1.8387428522109985,
+      "learning_rate": 3.270530653033142e-05,
+      "loss": 1.2345,
+      "step": 26765
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 2.6244566440582275,
+      "learning_rate": 3.269971712620322e-05,
+      "loss": 1.2351,
+      "step": 26770
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 1.3992868661880493,
+      "learning_rate": 3.269412729682765e-05,
+      "loss": 1.3125,
+      "step": 26775
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 1.4152156114578247,
+      "learning_rate": 3.268853704251342e-05,
+      "loss": 1.1555,
+      "step": 26780
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 3.2571237087249756,
+      "learning_rate": 3.2682946363569286e-05,
+      "loss": 0.986,
+      "step": 26785
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 1.7282525300979614,
+      "learning_rate": 3.267735526030402e-05,
+      "loss": 1.1626,
+      "step": 26790
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 1.6741749048233032,
+      "learning_rate": 3.267176373302639e-05,
+      "loss": 1.1052,
+      "step": 26795
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 2.290086030960083,
+      "learning_rate": 3.266617178204523e-05,
+      "loss": 0.9702,
+      "step": 26800
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 1.1538450717926025,
+      "learning_rate": 3.2660579407669374e-05,
+      "loss": 1.2276,
+      "step": 26805
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 5.698193073272705,
+      "learning_rate": 3.265498661020767e-05,
+      "loss": 0.9967,
+      "step": 26810
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 3.099355459213257,
+      "learning_rate": 3.2649393389969016e-05,
+      "loss": 0.9655,
+      "step": 26815
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 2.5500590801239014,
+      "learning_rate": 3.264379974726232e-05,
+      "loss": 1.1817,
+      "step": 26820
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 2.5358612537384033,
+      "learning_rate": 3.2638205682396504e-05,
+      "loss": 1.1314,
+      "step": 26825
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 1.88272225856781,
+      "learning_rate": 3.2632611195680535e-05,
+      "loss": 1.0866,
+      "step": 26830
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 2.205331325531006,
+      "learning_rate": 3.262701628742338e-05,
+      "loss": 1.1186,
+      "step": 26835
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 3.861008644104004,
+      "learning_rate": 3.262142095793404e-05,
+      "loss": 0.9824,
+      "step": 26840
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 1.2620456218719482,
+      "learning_rate": 3.261582520752154e-05,
+      "loss": 1.1634,
+      "step": 26845
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 2.01745343208313,
+      "learning_rate": 3.261022903649494e-05,
+      "loss": 1.1332,
+      "step": 26850
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 2.3341500759124756,
+      "learning_rate": 3.26046324451633e-05,
+      "loss": 1.1015,
+      "step": 26855
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 2.3520326614379883,
+      "learning_rate": 3.2599035433835706e-05,
+      "loss": 0.9539,
+      "step": 26860
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 2.2860143184661865,
+      "learning_rate": 3.2593438002821286e-05,
+      "loss": 1.098,
+      "step": 26865
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 4.713287830352783,
+      "learning_rate": 3.2587840152429186e-05,
+      "loss": 1.0839,
+      "step": 26870
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 1.481998324394226,
+      "learning_rate": 3.258224188296855e-05,
+      "loss": 1.096,
+      "step": 26875
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 2.8222692012786865,
+      "learning_rate": 3.257664319474858e-05,
+      "loss": 1.1853,
+      "step": 26880
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 2.920170307159424,
+      "learning_rate": 3.257104408807848e-05,
+      "loss": 1.0507,
+      "step": 26885
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 3.3776135444641113,
+      "learning_rate": 3.25654445632675e-05,
+      "loss": 1.0174,
+      "step": 26890
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 2.022411346435547,
+      "learning_rate": 3.255984462062487e-05,
+      "loss": 1.1012,
+      "step": 26895
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 4.32869291305542,
+      "learning_rate": 3.255424426045987e-05,
+      "loss": 1.0786,
+      "step": 26900
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 1.3376225233078003,
+      "learning_rate": 3.254864348308182e-05,
+      "loss": 1.1797,
+      "step": 26905
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 1.1576924324035645,
+      "learning_rate": 3.2543042288800035e-05,
+      "loss": 1.0853,
+      "step": 26910
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 1.6126631498336792,
+      "learning_rate": 3.2537440677923864e-05,
+      "loss": 1.0122,
+      "step": 26915
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 2.130544900894165,
+      "learning_rate": 3.253183865076269e-05,
+      "loss": 1.1658,
+      "step": 26920
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 3.746469259262085,
+      "learning_rate": 3.252623620762589e-05,
+      "loss": 1.1788,
+      "step": 26925
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 2.503143548965454,
+      "learning_rate": 3.2520633348822884e-05,
+      "loss": 1.1909,
+      "step": 26930
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 2.250931978225708,
+      "learning_rate": 3.251503007466311e-05,
+      "loss": 1.1565,
+      "step": 26935
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 1.9409987926483154,
+      "learning_rate": 3.2509426385456046e-05,
+      "loss": 1.1417,
+      "step": 26940
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 2.2973732948303223,
+      "learning_rate": 3.250382228151116e-05,
+      "loss": 1.1895,
+      "step": 26945
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 3.5512242317199707,
+      "learning_rate": 3.249821776313798e-05,
+      "loss": 1.1509,
+      "step": 26950
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 3.031540870666504,
+      "learning_rate": 3.2492612830646025e-05,
+      "loss": 1.1068,
+      "step": 26955
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 2.2181246280670166,
+      "learning_rate": 3.248700748434485e-05,
+      "loss": 1.1758,
+      "step": 26960
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 1.102636456489563,
+      "learning_rate": 3.248140172454403e-05,
+      "loss": 1.0737,
+      "step": 26965
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 1.951006293296814,
+      "learning_rate": 3.2475795551553166e-05,
+      "loss": 1.1618,
+      "step": 26970
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 1.9825594425201416,
+      "learning_rate": 3.2470188965681894e-05,
+      "loss": 1.0879,
+      "step": 26975
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 4.088083267211914,
+      "learning_rate": 3.246458196723985e-05,
+      "loss": 1.1725,
+      "step": 26980
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 6.844773292541504,
+      "learning_rate": 3.2458974556536694e-05,
+      "loss": 1.0785,
+      "step": 26985
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 1.5896713733673096,
+      "learning_rate": 3.245336673388213e-05,
+      "loss": 1.0559,
+      "step": 26990
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 1.1929066181182861,
+      "learning_rate": 3.244775849958587e-05,
+      "loss": 1.0536,
+      "step": 26995
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 1.1825475692749023,
+      "learning_rate": 3.244214985395765e-05,
+      "loss": 1.2364,
+      "step": 27000
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 3.0042402744293213,
+      "learning_rate": 3.2436540797307224e-05,
+      "loss": 1.1558,
+      "step": 27005
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 1.305233359336853,
+      "learning_rate": 3.2430931329944384e-05,
+      "loss": 1.1471,
+      "step": 27010
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 2.5870091915130615,
+      "learning_rate": 3.242532145217894e-05,
+      "loss": 1.2085,
+      "step": 27015
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 1.0035154819488525,
+      "learning_rate": 3.24197111643207e-05,
+      "loss": 0.9627,
+      "step": 27020
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 2.0611767768859863,
+      "learning_rate": 3.241410046667952e-05,
+      "loss": 1.0554,
+      "step": 27025
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 1.6869128942489624,
+      "learning_rate": 3.2408489359565286e-05,
+      "loss": 1.0618,
+      "step": 27030
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 1.8879674673080444,
+      "learning_rate": 3.240287784328789e-05,
+      "loss": 1.1033,
+      "step": 27035
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 3.194080114364624,
+      "learning_rate": 3.239726591815724e-05,
+      "loss": 0.9735,
+      "step": 27040
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 2.4983654022216797,
+      "learning_rate": 3.239165358448327e-05,
+      "loss": 1.1598,
+      "step": 27045
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 1.8231905698776245,
+      "learning_rate": 3.2386040842575976e-05,
+      "loss": 1.1378,
+      "step": 27050
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 4.191860198974609,
+      "learning_rate": 3.238042769274531e-05,
+      "loss": 0.9522,
+      "step": 27055
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 3.744335412979126,
+      "learning_rate": 3.23748141353013e-05,
+      "loss": 1.3162,
+      "step": 27060
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 2.127502202987671,
+      "learning_rate": 3.236920017055397e-05,
+      "loss": 0.9964,
+      "step": 27065
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 3.770505666732788,
+      "learning_rate": 3.2363585798813376e-05,
+      "loss": 1.2214,
+      "step": 27070
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 3.2522053718566895,
+      "learning_rate": 3.2357971020389586e-05,
+      "loss": 1.1507,
+      "step": 27075
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 1.4464082717895508,
+      "learning_rate": 3.235235583559271e-05,
+      "loss": 1.0405,
+      "step": 27080
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 2.175246238708496,
+      "learning_rate": 3.2346740244732866e-05,
+      "loss": 1.284,
+      "step": 27085
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 3.6082305908203125,
+      "learning_rate": 3.234112424812019e-05,
+      "loss": 1.0077,
+      "step": 27090
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 2.128209114074707,
+      "learning_rate": 3.233550784606486e-05,
+      "loss": 1.1639,
+      "step": 27095
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 1.130769968032837,
+      "learning_rate": 3.232989103887704e-05,
+      "loss": 1.2106,
+      "step": 27100
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 1.5438563823699951,
+      "learning_rate": 3.232427382686697e-05,
+      "loss": 1.1424,
+      "step": 27105
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 5.550532817840576,
+      "learning_rate": 3.231865621034486e-05,
+      "loss": 1.1006,
+      "step": 27110
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 20.100221633911133,
+      "learning_rate": 3.2313038189620995e-05,
+      "loss": 0.8935,
+      "step": 27115
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 2.989924430847168,
+      "learning_rate": 3.230741976500562e-05,
+      "loss": 1.3188,
+      "step": 27120
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 2.9077560901641846,
+      "learning_rate": 3.2301800936809044e-05,
+      "loss": 1.1537,
+      "step": 27125
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 1.805272102355957,
+      "learning_rate": 3.229618170534159e-05,
+      "loss": 1.1026,
+      "step": 27130
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 3.8493740558624268,
+      "learning_rate": 3.2290562070913613e-05,
+      "loss": 1.1274,
+      "step": 27135
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 3.3296542167663574,
+      "learning_rate": 3.2284942033835464e-05,
+      "loss": 1.0671,
+      "step": 27140
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 1.8140013217926025,
+      "learning_rate": 3.2279321594417546e-05,
+      "loss": 1.0956,
+      "step": 27145
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 2.7062671184539795,
+      "learning_rate": 3.227370075297026e-05,
+      "loss": 0.9826,
+      "step": 27150
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 1.546039342880249,
+      "learning_rate": 3.226807950980404e-05,
+      "loss": 1.2878,
+      "step": 27155
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 1.861967921257019,
+      "learning_rate": 3.2262457865229337e-05,
+      "loss": 0.9757,
+      "step": 27160
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 2.4952914714813232,
+      "learning_rate": 3.2256835819556643e-05,
+      "loss": 1.0902,
+      "step": 27165
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 4.320837497711182,
+      "learning_rate": 3.225121337309645e-05,
+      "loss": 1.2187,
+      "step": 27170
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 3.9226810932159424,
+      "learning_rate": 3.224559052615928e-05,
+      "loss": 1.06,
+      "step": 27175
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 1.381098985671997,
+      "learning_rate": 3.2239967279055675e-05,
+      "loss": 1.1461,
+      "step": 27180
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 2.233602285385132,
+      "learning_rate": 3.22343436320962e-05,
+      "loss": 1.2388,
+      "step": 27185
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 1.6678953170776367,
+      "learning_rate": 3.222871958559144e-05,
+      "loss": 1.2351,
+      "step": 27190
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 2.178392171859741,
+      "learning_rate": 3.2223095139852024e-05,
+      "loss": 1.0681,
+      "step": 27195
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 1.1121859550476074,
+      "learning_rate": 3.221747029518857e-05,
+      "loss": 1.3043,
+      "step": 27200
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 1.504831075668335,
+      "learning_rate": 3.221184505191173e-05,
+      "loss": 1.164,
+      "step": 27205
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 1.5870822668075562,
+      "learning_rate": 3.2206219410332184e-05,
+      "loss": 1.1253,
+      "step": 27210
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 2.400480270385742,
+      "learning_rate": 3.220059337076063e-05,
+      "loss": 1.065,
+      "step": 27215
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 6.77662992477417,
+      "learning_rate": 3.2194966933507794e-05,
+      "loss": 1.043,
+      "step": 27220
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 2.4106528759002686,
+      "learning_rate": 3.2189340098884405e-05,
+      "loss": 1.33,
+      "step": 27225
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 6.033102989196777,
+      "learning_rate": 3.2183712867201236e-05,
+      "loss": 1.1468,
+      "step": 27230
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 6.416871547698975,
+      "learning_rate": 3.2178085238769076e-05,
+      "loss": 1.0036,
+      "step": 27235
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 3.3531298637390137,
+      "learning_rate": 3.217245721389873e-05,
+      "loss": 1.2439,
+      "step": 27240
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 1.8699911832809448,
+      "learning_rate": 3.2166828792901025e-05,
+      "loss": 1.0194,
+      "step": 27245
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 2.013090133666992,
+      "learning_rate": 3.216119997608682e-05,
+      "loss": 1.013,
+      "step": 27250
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 1.574511170387268,
+      "learning_rate": 3.215557076376698e-05,
+      "loss": 1.1492,
+      "step": 27255
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 1.5347402095794678,
+      "learning_rate": 3.2149941156252406e-05,
+      "loss": 1.0025,
+      "step": 27260
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 2.3538808822631836,
+      "learning_rate": 3.214431115385401e-05,
+      "loss": 1.0706,
+      "step": 27265
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 2.6728079319000244,
+      "learning_rate": 3.213868075688273e-05,
+      "loss": 1.0859,
+      "step": 27270
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 2.588715076446533,
+      "learning_rate": 3.213304996564955e-05,
+      "loss": 1.058,
+      "step": 27275
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 2.8418338298797607,
+      "learning_rate": 3.2127418780465423e-05,
+      "loss": 1.2622,
+      "step": 27280
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 3.8776187896728516,
+      "learning_rate": 3.212178720164136e-05,
+      "loss": 1.0961,
+      "step": 27285
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 1.6147956848144531,
+      "learning_rate": 3.2116155229488404e-05,
+      "loss": 1.0547,
+      "step": 27290
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 1.412688136100769,
+      "learning_rate": 3.211052286431759e-05,
+      "loss": 1.2792,
+      "step": 27295
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 4.9000654220581055,
+      "learning_rate": 3.210489010643998e-05,
+      "loss": 1.2015,
+      "step": 27300
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 0.983841061592102,
+      "learning_rate": 3.2099256956166684e-05,
+      "loss": 1.022,
+      "step": 27305
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 2.0872743129730225,
+      "learning_rate": 3.20936234138088e-05,
+      "loss": 0.9733,
+      "step": 27310
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 2.8705453872680664,
+      "learning_rate": 3.208798947967748e-05,
+      "loss": 1.1142,
+      "step": 27315
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 2.3172919750213623,
+      "learning_rate": 3.208235515408385e-05,
+      "loss": 1.217,
+      "step": 27320
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 2.136756658554077,
+      "learning_rate": 3.207672043733912e-05,
+      "loss": 1.1509,
+      "step": 27325
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 1.0697053670883179,
+      "learning_rate": 3.207108532975447e-05,
+      "loss": 0.9292,
+      "step": 27330
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 2.2047574520111084,
+      "learning_rate": 3.206544983164113e-05,
+      "loss": 1.1085,
+      "step": 27335
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 1.8211325407028198,
+      "learning_rate": 3.205981394331035e-05,
+      "loss": 1.1218,
+      "step": 27340
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 10.550779342651367,
+      "learning_rate": 3.205417766507336e-05,
+      "loss": 1.009,
+      "step": 27345
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 1.852550745010376,
+      "learning_rate": 3.204854099724148e-05,
+      "loss": 1.0607,
+      "step": 27350
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 2.7020022869110107,
+      "learning_rate": 3.2042903940126015e-05,
+      "loss": 1.0989,
+      "step": 27355
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 1.7700706720352173,
+      "learning_rate": 3.203726649403828e-05,
+      "loss": 1.0878,
+      "step": 27360
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 6.152558326721191,
+      "learning_rate": 3.203162865928963e-05,
+      "loss": 1.1536,
+      "step": 27365
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 1.028457522392273,
+      "learning_rate": 3.202599043619145e-05,
+      "loss": 1.3372,
+      "step": 27370
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 2.56754994392395,
+      "learning_rate": 3.2020351825055114e-05,
+      "loss": 1.1325,
+      "step": 27375
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 8.181985855102539,
+      "learning_rate": 3.201471282619204e-05,
+      "loss": 1.1545,
+      "step": 27380
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.9413318037986755,
+      "learning_rate": 3.200907343991367e-05,
+      "loss": 1.1858,
+      "step": 27385
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 2.4057810306549072,
+      "learning_rate": 3.2003433666531456e-05,
+      "loss": 1.0307,
+      "step": 27390
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 1.1477302312850952,
+      "learning_rate": 3.199779350635688e-05,
+      "loss": 1.1957,
+      "step": 27395
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 2.2809982299804688,
+      "learning_rate": 3.199215295970145e-05,
+      "loss": 1.2828,
+      "step": 27400
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 10.860541343688965,
+      "learning_rate": 3.198651202687668e-05,
+      "loss": 1.0256,
+      "step": 27405
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 2.4457149505615234,
+      "learning_rate": 3.198087070819411e-05,
+      "loss": 1.1488,
+      "step": 27410
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 2.509361743927002,
+      "learning_rate": 3.1975229003965305e-05,
+      "loss": 0.9908,
+      "step": 27415
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 2.225059747695923,
+      "learning_rate": 3.1969586914501854e-05,
+      "loss": 1.1636,
+      "step": 27420
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 4.2079386711120605,
+      "learning_rate": 3.196394444011536e-05,
+      "loss": 1.0119,
+      "step": 27425
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 3.070852041244507,
+      "learning_rate": 3.1958301581117455e-05,
+      "loss": 1.0048,
+      "step": 27430
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 1.48086416721344,
+      "learning_rate": 3.195265833781979e-05,
+      "loss": 0.8762,
+      "step": 27435
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 2.2743024826049805,
+      "learning_rate": 3.1947014710534024e-05,
+      "loss": 1.0763,
+      "step": 27440
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 1.9955936670303345,
+      "learning_rate": 3.194137069957186e-05,
+      "loss": 1.1179,
+      "step": 27445
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 1.0980212688446045,
+      "learning_rate": 3.1935726305245e-05,
+      "loss": 1.0278,
+      "step": 27450
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 2.3399531841278076,
+      "learning_rate": 3.19300815278652e-05,
+      "loss": 1.0881,
+      "step": 27455
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 2.5593791007995605,
+      "learning_rate": 3.192443636774419e-05,
+      "loss": 1.003,
+      "step": 27460
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 2.6024117469787598,
+      "learning_rate": 3.1918790825193764e-05,
+      "loss": 0.9777,
+      "step": 27465
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 3.8112480640411377,
+      "learning_rate": 3.191314490052572e-05,
+      "loss": 1.1839,
+      "step": 27470
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 3.738812208175659,
+      "learning_rate": 3.190749859405185e-05,
+      "loss": 0.9292,
+      "step": 27475
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 1.4232354164123535,
+      "learning_rate": 3.1901851906084025e-05,
+      "loss": 1.1799,
+      "step": 27480
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 2.850757122039795,
+      "learning_rate": 3.189620483693409e-05,
+      "loss": 1.1459,
+      "step": 27485
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 9.836990356445312,
+      "learning_rate": 3.189055738691393e-05,
+      "loss": 1.2023,
+      "step": 27490
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 2.8823812007904053,
+      "learning_rate": 3.188490955633545e-05,
+      "loss": 1.1401,
+      "step": 27495
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 4.259209632873535,
+      "learning_rate": 3.187926134551057e-05,
+      "loss": 1.0947,
+      "step": 27500
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 1.9353116750717163,
+      "learning_rate": 3.1873612754751234e-05,
+      "loss": 0.9909,
+      "step": 27505
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 1.2285420894622803,
+      "learning_rate": 3.1867963784369415e-05,
+      "loss": 1.1093,
+      "step": 27510
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 2.983264446258545,
+      "learning_rate": 3.186231443467709e-05,
+      "loss": 0.963,
+      "step": 27515
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 2.1744792461395264,
+      "learning_rate": 3.185666470598627e-05,
+      "loss": 1.1857,
+      "step": 27520
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 6.445066928863525,
+      "learning_rate": 3.1851014598608994e-05,
+      "loss": 1.0607,
+      "step": 27525
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 3.2869045734405518,
+      "learning_rate": 3.1845364112857294e-05,
+      "loss": 1.0861,
+      "step": 27530
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 2.0690512657165527,
+      "learning_rate": 3.183971324904325e-05,
+      "loss": 1.1103,
+      "step": 27535
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 3.5820834636688232,
+      "learning_rate": 3.183406200747896e-05,
+      "loss": 1.1151,
+      "step": 27540
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 3.357940435409546,
+      "learning_rate": 3.1828410388476526e-05,
+      "loss": 0.9717,
+      "step": 27545
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 1.1148250102996826,
+      "learning_rate": 3.182275839234808e-05,
+      "loss": 1.188,
+      "step": 27550
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 2.7538514137268066,
+      "learning_rate": 3.181710601940578e-05,
+      "loss": 1.0378,
+      "step": 27555
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 1.7496304512023926,
+      "learning_rate": 3.1811453269961804e-05,
+      "loss": 1.0702,
+      "step": 27560
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 1.491807460784912,
+      "learning_rate": 3.180580014432835e-05,
+      "loss": 1.1796,
+      "step": 27565
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 1.2797476053237915,
+      "learning_rate": 3.180014664281762e-05,
+      "loss": 0.8992,
+      "step": 27570
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 2.5790648460388184,
+      "learning_rate": 3.179449276574186e-05,
+      "loss": 1.1928,
+      "step": 27575
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 3.154162883758545,
+      "learning_rate": 3.178883851341333e-05,
+      "loss": 1.176,
+      "step": 27580
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 1.5420721769332886,
+      "learning_rate": 3.1783183886144305e-05,
+      "loss": 1.162,
+      "step": 27585
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 2.776522159576416,
+      "learning_rate": 3.177752888424708e-05,
+      "loss": 1.0938,
+      "step": 27590
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 2.355360507965088,
+      "learning_rate": 3.177187350803398e-05,
+      "loss": 0.9884,
+      "step": 27595
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 1.3630084991455078,
+      "learning_rate": 3.176621775781736e-05,
+      "loss": 1.1724,
+      "step": 27600
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 2.4475739002227783,
+      "learning_rate": 3.1760561633909546e-05,
+      "loss": 1.1558,
+      "step": 27605
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 3.5432145595550537,
+      "learning_rate": 3.175490513662295e-05,
+      "loss": 0.9509,
+      "step": 27610
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 2.6473562717437744,
+      "learning_rate": 3.1749248266269966e-05,
+      "loss": 1.0777,
+      "step": 27615
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 2.7413153648376465,
+      "learning_rate": 3.174359102316301e-05,
+      "loss": 1.2854,
+      "step": 27620
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 2.960883378982544,
+      "learning_rate": 3.173793340761453e-05,
+      "loss": 1.1436,
+      "step": 27625
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 3.2864279747009277,
+      "learning_rate": 3.1732275419937e-05,
+      "loss": 1.1531,
+      "step": 27630
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 2.264131546020508,
+      "learning_rate": 3.1726617060442884e-05,
+      "loss": 1.1501,
+      "step": 27635
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 4.641735553741455,
+      "learning_rate": 3.172095832944472e-05,
+      "loss": 1.1033,
+      "step": 27640
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 2.3437931537628174,
+      "learning_rate": 3.1715299227255e-05,
+      "loss": 1.1294,
+      "step": 27645
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 4.831884860992432,
+      "learning_rate": 3.170963975418628e-05,
+      "loss": 1.0589,
+      "step": 27650
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 5.432398796081543,
+      "learning_rate": 3.170397991055114e-05,
+      "loss": 0.97,
+      "step": 27655
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 2.6022696495056152,
+      "learning_rate": 3.1698319696662156e-05,
+      "loss": 1.1821,
+      "step": 27660
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 2.880967378616333,
+      "learning_rate": 3.1692659112831934e-05,
+      "loss": 0.9731,
+      "step": 27665
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 1.824782371520996,
+      "learning_rate": 3.16869981593731e-05,
+      "loss": 1.1546,
+      "step": 27670
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 1.6104397773742676,
+      "learning_rate": 3.168133683659832e-05,
+      "loss": 1.0998,
+      "step": 27675
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 3.1559090614318848,
+      "learning_rate": 3.167567514482025e-05,
+      "loss": 1.1066,
+      "step": 27680
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 1.1295137405395508,
+      "learning_rate": 3.1670013084351575e-05,
+      "loss": 1.2197,
+      "step": 27685
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 3.0015058517456055,
+      "learning_rate": 3.166435065550501e-05,
+      "loss": 1.1269,
+      "step": 27690
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 4.1721391677856445,
+      "learning_rate": 3.1658687858593294e-05,
+      "loss": 0.992,
+      "step": 27695
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 2.583388328552246,
+      "learning_rate": 3.165302469392917e-05,
+      "loss": 1.0925,
+      "step": 27700
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 8.460959434509277,
+      "learning_rate": 3.16473611618254e-05,
+      "loss": 1.0993,
+      "step": 27705
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 5.1278767585754395,
+      "learning_rate": 3.164169726259477e-05,
+      "loss": 1.0187,
+      "step": 27710
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 2.619326591491699,
+      "learning_rate": 3.163603299655012e-05,
+      "loss": 0.9933,
+      "step": 27715
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 2.9472780227661133,
+      "learning_rate": 3.1630368364004264e-05,
+      "loss": 1.1644,
+      "step": 27720
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 1.8340462446212769,
+      "learning_rate": 3.162470336527006e-05,
+      "loss": 1.2497,
+      "step": 27725
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 4.899162292480469,
+      "learning_rate": 3.161903800066037e-05,
+      "loss": 1.0712,
+      "step": 27730
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 3.0266971588134766,
+      "learning_rate": 3.161337227048809e-05,
+      "loss": 1.1594,
+      "step": 27735
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 1.212259292602539,
+      "learning_rate": 3.1607706175066134e-05,
+      "loss": 1.0818,
+      "step": 27740
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 2.2834479808807373,
+      "learning_rate": 3.1602039714707434e-05,
+      "loss": 1.0276,
+      "step": 27745
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 1.508610486984253,
+      "learning_rate": 3.159637288972494e-05,
+      "loss": 1.1272,
+      "step": 27750
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 1.2361581325531006,
+      "learning_rate": 3.159070570043163e-05,
+      "loss": 1.0414,
+      "step": 27755
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 2.1319968700408936,
+      "learning_rate": 3.15850381471405e-05,
+      "loss": 1.052,
+      "step": 27760
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 3.3206193447113037,
+      "learning_rate": 3.157937023016456e-05,
+      "loss": 1.0506,
+      "step": 27765
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 3.340884208679199,
+      "learning_rate": 3.157370194981683e-05,
+      "loss": 1.1607,
+      "step": 27770
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 12.616667747497559,
+      "learning_rate": 3.156803330641038e-05,
+      "loss": 1.2882,
+      "step": 27775
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 3.9304230213165283,
+      "learning_rate": 3.1562364300258275e-05,
+      "loss": 1.302,
+      "step": 27780
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 3.3284659385681152,
+      "learning_rate": 3.1556694931673615e-05,
+      "loss": 1.0514,
+      "step": 27785
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 2.4775948524475098,
+      "learning_rate": 3.155102520096951e-05,
+      "loss": 1.0072,
+      "step": 27790
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 2.661914110183716,
+      "learning_rate": 3.154535510845909e-05,
+      "loss": 1.1323,
+      "step": 27795
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 3.137718677520752,
+      "learning_rate": 3.15396846544555e-05,
+      "loss": 1.248,
+      "step": 27800
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 1.339923620223999,
+      "learning_rate": 3.153401383927194e-05,
+      "loss": 1.0493,
+      "step": 27805
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 1.203647494316101,
+      "learning_rate": 3.1528342663221576e-05,
+      "loss": 0.9536,
+      "step": 27810
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 1.4011954069137573,
+      "learning_rate": 3.152267112661764e-05,
+      "loss": 1.257,
+      "step": 27815
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 3.3855531215667725,
+      "learning_rate": 3.151699922977336e-05,
+      "loss": 1.1184,
+      "step": 27820
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 2.566340208053589,
+      "learning_rate": 3.151132697300199e-05,
+      "loss": 1.0047,
+      "step": 27825
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 2.9394350051879883,
+      "learning_rate": 3.150565435661679e-05,
+      "loss": 1.244,
+      "step": 27830
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 3.349531412124634,
+      "learning_rate": 3.149998138093107e-05,
+      "loss": 1.1824,
+      "step": 27835
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 1.7213515043258667,
+      "learning_rate": 3.149430804625812e-05,
+      "loss": 1.1984,
+      "step": 27840
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 1.7168335914611816,
+      "learning_rate": 3.148863435291129e-05,
+      "loss": 1.1489,
+      "step": 27845
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 2.286252975463867,
+      "learning_rate": 3.148296030120394e-05,
+      "loss": 1.2274,
+      "step": 27850
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 1.7521454095840454,
+      "learning_rate": 3.1477285891449434e-05,
+      "loss": 1.1893,
+      "step": 27855
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 3.142904281616211,
+      "learning_rate": 3.147161112396115e-05,
+      "loss": 1.0052,
+      "step": 27860
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 2.172449827194214,
+      "learning_rate": 3.1465935999052514e-05,
+      "loss": 1.0957,
+      "step": 27865
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 2.116454601287842,
+      "learning_rate": 3.146026051703695e-05,
+      "loss": 1.0053,
+      "step": 27870
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 0.8631601333618164,
+      "learning_rate": 3.145458467822792e-05,
+      "loss": 0.9305,
+      "step": 27875
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 2.0894715785980225,
+      "learning_rate": 3.144890848293889e-05,
+      "loss": 1.1072,
+      "step": 27880
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 4.275918006896973,
+      "learning_rate": 3.1443231931483334e-05,
+      "loss": 1.0385,
+      "step": 27885
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 2.0376479625701904,
+      "learning_rate": 3.143755502417478e-05,
+      "loss": 1.3225,
+      "step": 27890
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 2.2488062381744385,
+      "learning_rate": 3.143187776132676e-05,
+      "loss": 1.1994,
+      "step": 27895
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 2.3104982376098633,
+      "learning_rate": 3.1426200143252815e-05,
+      "loss": 1.0693,
+      "step": 27900
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 2.0975565910339355,
+      "learning_rate": 3.142052217026651e-05,
+      "loss": 1.0613,
+      "step": 27905
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 3.1652896404266357,
+      "learning_rate": 3.1414843842681455e-05,
+      "loss": 1.1656,
+      "step": 27910
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 2.0645837783813477,
+      "learning_rate": 3.1409165160811226e-05,
+      "loss": 1.1103,
+      "step": 27915
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 2.3577146530151367,
+      "learning_rate": 3.140348612496947e-05,
+      "loss": 1.0475,
+      "step": 27920
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 3.1144940853118896,
+      "learning_rate": 3.1398942641647336e-05,
+      "loss": 1.3145,
+      "step": 27925
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 2.6503536701202393,
+      "learning_rate": 3.139326296944723e-05,
+      "loss": 1.0987,
+      "step": 27930
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 1.5339775085449219,
+      "learning_rate": 3.138758294415386e-05,
+      "loss": 1.0772,
+      "step": 27935
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 2.08730149269104,
+      "learning_rate": 3.138190256608093e-05,
+      "loss": 1.1344,
+      "step": 27940
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 1.825244426727295,
+      "learning_rate": 3.137622183554215e-05,
+      "loss": 1.2571,
+      "step": 27945
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 2.94596004486084,
+      "learning_rate": 3.137054075285126e-05,
+      "loss": 1.1669,
+      "step": 27950
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 1.1492950916290283,
+      "learning_rate": 3.1364859318322025e-05,
+      "loss": 1.2209,
+      "step": 27955
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 5.7294769287109375,
+      "learning_rate": 3.135917753226823e-05,
+      "loss": 1.0503,
+      "step": 27960
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 1.8514894247055054,
+      "learning_rate": 3.1353495395003675e-05,
+      "loss": 1.0441,
+      "step": 27965
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 1.6761661767959595,
+      "learning_rate": 3.134781290684216e-05,
+      "loss": 0.9575,
+      "step": 27970
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 2.4914565086364746,
+      "learning_rate": 3.134213006809755e-05,
+      "loss": 1.2053,
+      "step": 27975
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 1.1112005710601807,
+      "learning_rate": 3.133644687908368e-05,
+      "loss": 1.1066,
+      "step": 27980
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 5.55382776260376,
+      "learning_rate": 3.133076334011443e-05,
+      "loss": 0.8715,
+      "step": 27985
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 3.3123645782470703,
+      "learning_rate": 3.1325079451503715e-05,
+      "loss": 1.0493,
+      "step": 27990
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 29.16041374206543,
+      "learning_rate": 3.131939521356543e-05,
+      "loss": 1.0889,
+      "step": 27995
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 2.6477432250976562,
+      "learning_rate": 3.131371062661351e-05,
+      "loss": 0.995,
+      "step": 28000
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 1.398253321647644,
+      "learning_rate": 3.130802569096194e-05,
+      "loss": 1.0378,
+      "step": 28005
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 1.7362818717956543,
+      "learning_rate": 3.130234040692464e-05,
+      "loss": 1.0509,
+      "step": 28010
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 2.3737027645111084,
+      "learning_rate": 3.129665477481564e-05,
+      "loss": 1.1996,
+      "step": 28015
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 2.6281936168670654,
+      "learning_rate": 3.129096879494894e-05,
+      "loss": 0.9499,
+      "step": 28020
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 5.763638019561768,
+      "learning_rate": 3.1285282467638577e-05,
+      "loss": 1.1019,
+      "step": 28025
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 3.136121988296509,
+      "learning_rate": 3.1279595793198593e-05,
+      "loss": 1.084,
+      "step": 28030
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 1.8044530153274536,
+      "learning_rate": 3.1273908771943064e-05,
+      "loss": 1.2044,
+      "step": 28035
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 4.135902404785156,
+      "learning_rate": 3.126822140418607e-05,
+      "loss": 1.086,
+      "step": 28040
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 2.257347822189331,
+      "learning_rate": 3.1262533690241726e-05,
+      "loss": 1.2456,
+      "step": 28045
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 3.2046022415161133,
+      "learning_rate": 3.1256845630424144e-05,
+      "loss": 1.214,
+      "step": 28050
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 3.7938222885131836,
+      "learning_rate": 3.125115722504749e-05,
+      "loss": 0.989,
+      "step": 28055
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 3.6777851581573486,
+      "learning_rate": 3.124546847442593e-05,
+      "loss": 1.0389,
+      "step": 28060
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 2.9035661220550537,
+      "learning_rate": 3.123977937887363e-05,
+      "loss": 0.9988,
+      "step": 28065
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 1.5526783466339111,
+      "learning_rate": 3.1234089938704805e-05,
+      "loss": 1.1668,
+      "step": 28070
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 3.05960750579834,
+      "learning_rate": 3.122840015423367e-05,
+      "loss": 1.1137,
+      "step": 28075
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 2.577622413635254,
+      "learning_rate": 3.122271002577446e-05,
+      "loss": 1.0632,
+      "step": 28080
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 2.9090895652770996,
+      "learning_rate": 3.121701955364146e-05,
+      "loss": 1.1106,
+      "step": 28085
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 1.0737097263336182,
+      "learning_rate": 3.121132873814892e-05,
+      "loss": 1.1976,
+      "step": 28090
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 5.77609920501709,
+      "learning_rate": 3.1205637579611154e-05,
+      "loss": 1.2236,
+      "step": 28095
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 6.517035961151123,
+      "learning_rate": 3.119994607834248e-05,
+      "loss": 0.9589,
+      "step": 28100
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 3.637007236480713,
+      "learning_rate": 3.1194254234657225e-05,
+      "loss": 1.1054,
+      "step": 28105
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 5.6374359130859375,
+      "learning_rate": 3.118856204886974e-05,
+      "loss": 1.1144,
+      "step": 28110
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 1.1041793823242188,
+      "learning_rate": 3.118286952129441e-05,
+      "loss": 1.0831,
+      "step": 28115
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 4.296421051025391,
+      "learning_rate": 3.117717665224562e-05,
+      "loss": 1.1175,
+      "step": 28120
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 1.8130741119384766,
+      "learning_rate": 3.117148344203779e-05,
+      "loss": 1.14,
+      "step": 28125
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 5.4478864669799805,
+      "learning_rate": 3.116578989098534e-05,
+      "loss": 1.1562,
+      "step": 28130
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 3.172430992126465,
+      "learning_rate": 3.116009599940273e-05,
+      "loss": 1.2238,
+      "step": 28135
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 1.4552370309829712,
+      "learning_rate": 3.1154401767604415e-05,
+      "loss": 1.2188,
+      "step": 28140
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 2.109438419342041,
+      "learning_rate": 3.114870719590489e-05,
+      "loss": 1.0155,
+      "step": 28145
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 1.051721215248108,
+      "learning_rate": 3.114301228461866e-05,
+      "loss": 1.0963,
+      "step": 28150
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 1.4635205268859863,
+      "learning_rate": 3.1137317034060236e-05,
+      "loss": 1.0117,
+      "step": 28155
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 2.354686975479126,
+      "learning_rate": 3.113162144454418e-05,
+      "loss": 1.2232,
+      "step": 28160
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 2.211465358734131,
+      "learning_rate": 3.112592551638505e-05,
+      "loss": 1.1567,
+      "step": 28165
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 2.628493070602417,
+      "learning_rate": 3.112022924989741e-05,
+      "loss": 1.0378,
+      "step": 28170
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 11.486804008483887,
+      "learning_rate": 3.111453264539588e-05,
+      "loss": 1.0214,
+      "step": 28175
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 2.4132378101348877,
+      "learning_rate": 3.110883570319507e-05,
+      "loss": 1.265,
+      "step": 28180
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 3.6286280155181885,
+      "learning_rate": 3.11031384236096e-05,
+      "loss": 1.031,
+      "step": 28185
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 2.399932622909546,
+      "learning_rate": 3.109744080695415e-05,
+      "loss": 1.0519,
+      "step": 28190
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 2.9167540073394775,
+      "learning_rate": 3.109174285354338e-05,
+      "loss": 1.1285,
+      "step": 28195
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 2.6910743713378906,
+      "learning_rate": 3.1086044563691984e-05,
+      "loss": 1.2128,
+      "step": 28200
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 2.2151172161102295,
+      "learning_rate": 3.108034593771467e-05,
+      "loss": 1.1706,
+      "step": 28205
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 4.152978897094727,
+      "learning_rate": 3.1074646975926176e-05,
+      "loss": 1.1396,
+      "step": 28210
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 4.019749164581299,
+      "learning_rate": 3.106894767864124e-05,
+      "loss": 1.1841,
+      "step": 28215
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 2.265693187713623,
+      "learning_rate": 3.106324804617463e-05,
+      "loss": 1.1932,
+      "step": 28220
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 2.9576539993286133,
+      "learning_rate": 3.105754807884113e-05,
+      "loss": 1.1587,
+      "step": 28225
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 3.3422529697418213,
+      "learning_rate": 3.105184777695555e-05,
+      "loss": 1.1099,
+      "step": 28230
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 3.2703285217285156,
+      "learning_rate": 3.104614714083271e-05,
+      "loss": 1.0236,
+      "step": 28235
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 2.098249912261963,
+      "learning_rate": 3.1040446170787444e-05,
+      "loss": 1.0275,
+      "step": 28240
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 3.4206290245056152,
+      "learning_rate": 3.103474486713462e-05,
+      "loss": 0.9629,
+      "step": 28245
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 3.716006278991699,
+      "learning_rate": 3.1029043230189106e-05,
+      "loss": 1.0353,
+      "step": 28250
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 1.844235897064209,
+      "learning_rate": 3.10233412602658e-05,
+      "loss": 1.1328,
+      "step": 28255
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 3.6148314476013184,
+      "learning_rate": 3.101763895767962e-05,
+      "loss": 1.0627,
+      "step": 28260
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 2.58854603767395,
+      "learning_rate": 3.101193632274549e-05,
+      "loss": 1.0033,
+      "step": 28265
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 1.9010263681411743,
+      "learning_rate": 3.100623335577837e-05,
+      "loss": 1.1728,
+      "step": 28270
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 3.03183913230896,
+      "learning_rate": 3.100053005709323e-05,
+      "loss": 1.133,
+      "step": 28275
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 2.4932944774627686,
+      "learning_rate": 3.0994826427005044e-05,
+      "loss": 1.1185,
+      "step": 28280
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 3.512427568435669,
+      "learning_rate": 3.098912246582884e-05,
+      "loss": 1.0649,
+      "step": 28285
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 1.8340367078781128,
+      "learning_rate": 3.098341817387961e-05,
+      "loss": 1.048,
+      "step": 28290
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 2.6581850051879883,
+      "learning_rate": 3.0977713551472424e-05,
+      "loss": 1.1243,
+      "step": 28295
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 2.2674968242645264,
+      "learning_rate": 3.097200859892232e-05,
+      "loss": 1.0713,
+      "step": 28300
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 2.0417089462280273,
+      "learning_rate": 3.09663033165444e-05,
+      "loss": 1.142,
+      "step": 28305
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 3.8348848819732666,
+      "learning_rate": 3.096059770465375e-05,
+      "loss": 1.0195,
+      "step": 28310
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 1.9417195320129395,
+      "learning_rate": 3.095489176356548e-05,
+      "loss": 1.1107,
+      "step": 28315
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 3.2145209312438965,
+      "learning_rate": 3.094918549359473e-05,
+      "loss": 1.1599,
+      "step": 28320
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 1.5097711086273193,
+      "learning_rate": 3.0943478895056645e-05,
+      "loss": 1.1153,
+      "step": 28325
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 1.1623693704605103,
+      "learning_rate": 3.09377719682664e-05,
+      "loss": 1.0584,
+      "step": 28330
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 3.9560139179229736,
+      "learning_rate": 3.093206471353918e-05,
+      "loss": 1.0023,
+      "step": 28335
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 2.495631217956543,
+      "learning_rate": 3.0926357131190196e-05,
+      "loss": 0.9913,
+      "step": 28340
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 1.4161936044692993,
+      "learning_rate": 3.092064922153466e-05,
+      "loss": 1.1667,
+      "step": 28345
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 2.1678714752197266,
+      "learning_rate": 3.091494098488783e-05,
+      "loss": 1.1517,
+      "step": 28350
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 1.4280831813812256,
+      "learning_rate": 3.090923242156496e-05,
+      "loss": 1.1162,
+      "step": 28355
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 5.203249931335449,
+      "learning_rate": 3.0903523531881325e-05,
+      "loss": 1.0205,
+      "step": 28360
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 2.362901449203491,
+      "learning_rate": 3.0897814316152214e-05,
+      "loss": 1.095,
+      "step": 28365
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 1.590612769126892,
+      "learning_rate": 3.089210477469295e-05,
+      "loss": 1.0452,
+      "step": 28370
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 1.0684711933135986,
+      "learning_rate": 3.0886394907818864e-05,
+      "loss": 1.0676,
+      "step": 28375
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 2.432785749435425,
+      "learning_rate": 3.088068471584531e-05,
+      "loss": 1.2619,
+      "step": 28380
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 4.381847381591797,
+      "learning_rate": 3.0874974199087654e-05,
+      "loss": 0.9273,
+      "step": 28385
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 2.7747817039489746,
+      "learning_rate": 3.086926335786128e-05,
+      "loss": 1.1676,
+      "step": 28390
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 2.8812899589538574,
+      "learning_rate": 3.086355219248158e-05,
+      "loss": 1.0638,
+      "step": 28395
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 3.0816097259521484,
+      "learning_rate": 3.0857840703263996e-05,
+      "loss": 0.9701,
+      "step": 28400
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 3.1254990100860596,
+      "learning_rate": 3.0852128890523954e-05,
+      "loss": 1.2058,
+      "step": 28405
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 3.639291286468506,
+      "learning_rate": 3.084641675457692e-05,
+      "loss": 1.1903,
+      "step": 28410
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 12.763455390930176,
+      "learning_rate": 3.0840704295738364e-05,
+      "loss": 0.9558,
+      "step": 28415
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 2.450136423110962,
+      "learning_rate": 3.083499151432378e-05,
+      "loss": 1.1158,
+      "step": 28420
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 2.2202651500701904,
+      "learning_rate": 3.082927841064869e-05,
+      "loss": 1.1919,
+      "step": 28425
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 5.485016822814941,
+      "learning_rate": 3.0823564985028596e-05,
+      "loss": 1.1658,
+      "step": 28430
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 3.856153964996338,
+      "learning_rate": 3.081785123777907e-05,
+      "loss": 0.9867,
+      "step": 28435
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 2.2953991889953613,
+      "learning_rate": 3.081213716921567e-05,
+      "loss": 1.1779,
+      "step": 28440
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 2.756497859954834,
+      "learning_rate": 3.0806422779653974e-05,
+      "loss": 1.1818,
+      "step": 28445
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 2.887840986251831,
+      "learning_rate": 3.080070806940958e-05,
+      "loss": 1.0941,
+      "step": 28450
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 1.7637460231781006,
+      "learning_rate": 3.0794993038798114e-05,
+      "loss": 1.1327,
+      "step": 28455
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 2.2904860973358154,
+      "learning_rate": 3.07892776881352e-05,
+      "loss": 1.2831,
+      "step": 28460
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 1.0153388977050781,
+      "learning_rate": 3.07835620177365e-05,
+      "loss": 1.0142,
+      "step": 28465
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 1.7611110210418701,
+      "learning_rate": 3.077784602791768e-05,
+      "loss": 1.2258,
+      "step": 28470
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 2.403919219970703,
+      "learning_rate": 3.077212971899443e-05,
+      "loss": 0.9231,
+      "step": 28475
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 4.131344318389893,
+      "learning_rate": 3.076641309128245e-05,
+      "loss": 1.0783,
+      "step": 28480
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 3.618623971939087,
+      "learning_rate": 3.0760696145097477e-05,
+      "loss": 1.2069,
+      "step": 28485
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 2.4561777114868164,
+      "learning_rate": 3.0754978880755246e-05,
+      "loss": 1.0412,
+      "step": 28490
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 2.350400924682617,
+      "learning_rate": 3.074926129857151e-05,
+      "loss": 1.1583,
+      "step": 28495
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 1.5594244003295898,
+      "learning_rate": 3.074354339886204e-05,
+      "loss": 1.2653,
+      "step": 28500
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 1.28977370262146,
+      "learning_rate": 3.073782518194265e-05,
+      "loss": 1.1303,
+      "step": 28505
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 1.747999668121338,
+      "learning_rate": 3.073210664812913e-05,
+      "loss": 1.0107,
+      "step": 28510
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 1.435805082321167,
+      "learning_rate": 3.072638779773732e-05,
+      "loss": 0.9404,
+      "step": 28515
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 2.7023794651031494,
+      "learning_rate": 3.0720668631083074e-05,
+      "loss": 1.1618,
+      "step": 28520
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 2.3820207118988037,
+      "learning_rate": 3.071494914848224e-05,
+      "loss": 1.2369,
+      "step": 28525
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 2.2789406776428223,
+      "learning_rate": 3.07092293502507e-05,
+      "loss": 1.0973,
+      "step": 28530
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 1.514122724533081,
+      "learning_rate": 3.0703509236704366e-05,
+      "loss": 1.0835,
+      "step": 28535
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 2.0251221656799316,
+      "learning_rate": 3.069778880815914e-05,
+      "loss": 1.1254,
+      "step": 28540
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 4.081529140472412,
+      "learning_rate": 3.069206806493095e-05,
+      "loss": 1.2861,
+      "step": 28545
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 1.8973231315612793,
+      "learning_rate": 3.068634700733577e-05,
+      "loss": 1.122,
+      "step": 28550
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 1.6018325090408325,
+      "learning_rate": 3.068062563568956e-05,
+      "loss": 1.0091,
+      "step": 28555
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 2.3443856239318848,
+      "learning_rate": 3.0674903950308295e-05,
+      "loss": 1.1682,
+      "step": 28560
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 2.2536654472351074,
+      "learning_rate": 3.0669181951507986e-05,
+      "loss": 1.0887,
+      "step": 28565
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 2.6237266063690186,
+      "learning_rate": 3.0663459639604645e-05,
+      "loss": 1.0353,
+      "step": 28570
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 2.24894118309021,
+      "learning_rate": 3.065773701491432e-05,
+      "loss": 1.2106,
+      "step": 28575
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 2.0882694721221924,
+      "learning_rate": 3.065201407775306e-05,
+      "loss": 1.1984,
+      "step": 28580
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 2.8938817977905273,
+      "learning_rate": 3.064629082843693e-05,
+      "loss": 1.0492,
+      "step": 28585
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 2.2271273136138916,
+      "learning_rate": 3.064056726728204e-05,
+      "loss": 1.165,
+      "step": 28590
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 5.680057525634766,
+      "learning_rate": 3.063484339460447e-05,
+      "loss": 1.1335,
+      "step": 28595
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 4.835309982299805,
+      "learning_rate": 3.0629119210720364e-05,
+      "loss": 1.0781,
+      "step": 28600
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 1.8358601331710815,
+      "learning_rate": 3.062339471594585e-05,
+      "loss": 1.0802,
+      "step": 28605
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 2.384139060974121,
+      "learning_rate": 3.061766991059709e-05,
+      "loss": 1.0434,
+      "step": 28610
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 1.7726037502288818,
+      "learning_rate": 3.0611944794990265e-05,
+      "loss": 0.9551,
+      "step": 28615
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 3.578003168106079,
+      "learning_rate": 3.060621936944157e-05,
+      "loss": 0.9481,
+      "step": 28620
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 2.8826417922973633,
+      "learning_rate": 3.0600493634267196e-05,
+      "loss": 1.1564,
+      "step": 28625
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 2.272235870361328,
+      "learning_rate": 3.059476758978338e-05,
+      "loss": 0.8952,
+      "step": 28630
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 2.82199764251709,
+      "learning_rate": 3.058904123630636e-05,
+      "loss": 1.1897,
+      "step": 28635
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 1.5297337770462036,
+      "learning_rate": 3.0583314574152414e-05,
+      "loss": 1.1364,
+      "step": 28640
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 2.4493231773376465,
+      "learning_rate": 3.05775876036378e-05,
+      "loss": 1.0275,
+      "step": 28645
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 4.320399761199951,
+      "learning_rate": 3.057186032507883e-05,
+      "loss": 1.205,
+      "step": 28650
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 4.2380595207214355,
+      "learning_rate": 3.05661327387918e-05,
+      "loss": 1.0884,
+      "step": 28655
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 3.2652807235717773,
+      "learning_rate": 3.056040484509304e-05,
+      "loss": 0.9676,
+      "step": 28660
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 2.164764404296875,
+      "learning_rate": 3.0554676644298906e-05,
+      "loss": 1.1527,
+      "step": 28665
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 3.5430350303649902,
+      "learning_rate": 3.0548948136725754e-05,
+      "loss": 1.285,
+      "step": 28670
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 2.4227631092071533,
+      "learning_rate": 3.0543219322689955e-05,
+      "loss": 1.0471,
+      "step": 28675
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 2.0381109714508057,
+      "learning_rate": 3.053749020250792e-05,
+      "loss": 1.0022,
+      "step": 28680
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 1.9876576662063599,
+      "learning_rate": 3.0531760776496064e-05,
+      "loss": 1.0601,
+      "step": 28685
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 2.803956985473633,
+      "learning_rate": 3.0526031044970806e-05,
+      "loss": 1.1354,
+      "step": 28690
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 1.6594746112823486,
+      "learning_rate": 3.05203010082486e-05,
+      "loss": 1.1666,
+      "step": 28695
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 2.4495763778686523,
+      "learning_rate": 3.0514570666645896e-05,
+      "loss": 1.2394,
+      "step": 28700
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 0.8862543106079102,
+      "learning_rate": 3.0508840020479194e-05,
+      "loss": 1.1907,
+      "step": 28705
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 3.8315329551696777,
+      "learning_rate": 3.0503109070064984e-05,
+      "loss": 1.026,
+      "step": 28710
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 6.697932243347168,
+      "learning_rate": 3.0497377815719787e-05,
+      "loss": 1.143,
+      "step": 28715
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 2.7949328422546387,
+      "learning_rate": 3.049164625776012e-05,
+      "loss": 1.0389,
+      "step": 28720
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 3.456163167953491,
+      "learning_rate": 3.048591439650254e-05,
+      "loss": 0.9912,
+      "step": 28725
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 4.755772113800049,
+      "learning_rate": 3.048018223226361e-05,
+      "loss": 1.0972,
+      "step": 28730
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 1.6794426441192627,
+      "learning_rate": 3.0474449765359908e-05,
+      "loss": 1.0897,
+      "step": 28735
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 1.9840363264083862,
+      "learning_rate": 3.0468716996108038e-05,
+      "loss": 1.0486,
+      "step": 28740
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 5.026776313781738,
+      "learning_rate": 3.046298392482462e-05,
+      "loss": 1.0958,
+      "step": 28745
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 3.4160451889038086,
+      "learning_rate": 3.0457250551826272e-05,
+      "loss": 1.1688,
+      "step": 28750
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 5.9285054206848145,
+      "learning_rate": 3.0451516877429648e-05,
+      "loss": 1.1551,
+      "step": 28755
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 2.421475410461426,
+      "learning_rate": 3.044578290195141e-05,
+      "loss": 1.3426,
+      "step": 28760
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 5.284264087677002,
+      "learning_rate": 3.0440048625708244e-05,
+      "loss": 1.22,
+      "step": 28765
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 1.7585577964782715,
+      "learning_rate": 3.0434314049016854e-05,
+      "loss": 1.3052,
+      "step": 28770
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 2.2892251014709473,
+      "learning_rate": 3.042857917219394e-05,
+      "loss": 1.195,
+      "step": 28775
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 2.7813527584075928,
+      "learning_rate": 3.0422843995556245e-05,
+      "loss": 1.1471,
+      "step": 28780
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 4.371838569641113,
+      "learning_rate": 3.041710851942051e-05,
+      "loss": 1.1609,
+      "step": 28785
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 5.511404991149902,
+      "learning_rate": 3.0411372744103504e-05,
+      "loss": 0.9399,
+      "step": 28790
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 6.196451663970947,
+      "learning_rate": 3.0405636669922004e-05,
+      "loss": 0.9446,
+      "step": 28795
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 1.4433752298355103,
+      "learning_rate": 3.0399900297192812e-05,
+      "loss": 1.1955,
+      "step": 28800
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 2.80629563331604,
+      "learning_rate": 3.0394163626232742e-05,
+      "loss": 1.2015,
+      "step": 28805
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 3.159404754638672,
+      "learning_rate": 3.0388426657358628e-05,
+      "loss": 1.0804,
+      "step": 28810
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 1.2445366382598877,
+      "learning_rate": 3.0382689390887297e-05,
+      "loss": 1.2395,
+      "step": 28815
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 3.2405431270599365,
+      "learning_rate": 3.0376951827135632e-05,
+      "loss": 1.0396,
+      "step": 28820
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 3.0723726749420166,
+      "learning_rate": 3.0371213966420503e-05,
+      "loss": 1.2086,
+      "step": 28825
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 2.3256523609161377,
+      "learning_rate": 3.0365475809058814e-05,
+      "loss": 1.2707,
+      "step": 28830
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 2.9459950923919678,
+      "learning_rate": 3.0359737355367467e-05,
+      "loss": 1.1848,
+      "step": 28835
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 2.410904884338379,
+      "learning_rate": 3.0353998605663403e-05,
+      "loss": 1.0366,
+      "step": 28840
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 4.633788585662842,
+      "learning_rate": 3.0348259560263563e-05,
+      "loss": 1.0301,
+      "step": 28845
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 2.1768550872802734,
+      "learning_rate": 3.0342520219484903e-05,
+      "loss": 1.2337,
+      "step": 28850
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 1.6011451482772827,
+      "learning_rate": 3.033678058364441e-05,
+      "loss": 1.1315,
+      "step": 28855
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 1.7358301877975464,
+      "learning_rate": 3.0331040653059063e-05,
+      "loss": 1.2123,
+      "step": 28860
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 2.091358184814453,
+      "learning_rate": 3.0325300428045883e-05,
+      "loss": 1.07,
+      "step": 28865
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 2.0123684406280518,
+      "learning_rate": 3.0319559908921895e-05,
+      "loss": 1.1838,
+      "step": 28870
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 2.712315082550049,
+      "learning_rate": 3.0313819096004154e-05,
+      "loss": 1.2163,
+      "step": 28875
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 2.8222601413726807,
+      "learning_rate": 3.03080779896097e-05,
+      "loss": 1.1945,
+      "step": 28880
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 4.944241046905518,
+      "learning_rate": 3.0302336590055617e-05,
+      "loss": 0.891,
+      "step": 28885
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 2.032339572906494,
+      "learning_rate": 3.0296594897658993e-05,
+      "loss": 1.2057,
+      "step": 28890
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 2.003133773803711,
+      "learning_rate": 3.0290852912736944e-05,
+      "loss": 1.052,
+      "step": 28895
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 2.7884342670440674,
+      "learning_rate": 3.0285110635606585e-05,
+      "loss": 1.1927,
+      "step": 28900
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 2.4504990577697754,
+      "learning_rate": 3.0279368066585056e-05,
+      "loss": 1.0378,
+      "step": 28905
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 6.284214496612549,
+      "learning_rate": 3.0273625205989525e-05,
+      "loss": 1.1471,
+      "step": 28910
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 12.456482887268066,
+      "learning_rate": 3.0267882054137148e-05,
+      "loss": 1.0668,
+      "step": 28915
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 2.481710195541382,
+      "learning_rate": 3.026213861134512e-05,
+      "loss": 1.2337,
+      "step": 28920
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 7.710996150970459,
+      "learning_rate": 3.025639487793065e-05,
+      "loss": 0.9881,
+      "step": 28925
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 2.2037603855133057,
+      "learning_rate": 3.0250650854210953e-05,
+      "loss": 1.1093,
+      "step": 28930
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 2.117426872253418,
+      "learning_rate": 3.0244906540503266e-05,
+      "loss": 1.1326,
+      "step": 28935
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 3.677184581756592,
+      "learning_rate": 3.023916193712485e-05,
+      "loss": 1.1629,
+      "step": 28940
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 3.842292070388794,
+      "learning_rate": 3.0233417044392953e-05,
+      "loss": 1.0908,
+      "step": 28945
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 2.213963508605957,
+      "learning_rate": 3.0227671862624878e-05,
+      "loss": 1.166,
+      "step": 28950
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 0.9696227312088013,
+      "learning_rate": 3.0221926392137922e-05,
+      "loss": 1.1106,
+      "step": 28955
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 1.315886378288269,
+      "learning_rate": 3.0216180633249396e-05,
+      "loss": 1.1739,
+      "step": 28960
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 4.016354560852051,
+      "learning_rate": 3.0210434586276637e-05,
+      "loss": 0.9315,
+      "step": 28965
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 4.51848030090332,
+      "learning_rate": 3.0204688251536994e-05,
+      "loss": 0.9756,
+      "step": 28970
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 3.8415751457214355,
+      "learning_rate": 3.0198941629347833e-05,
+      "loss": 1.0718,
+      "step": 28975
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 3.0675783157348633,
+      "learning_rate": 3.0193194720026524e-05,
+      "loss": 1.2526,
+      "step": 28980
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 1.8814668655395508,
+      "learning_rate": 3.0187447523890468e-05,
+      "loss": 1.1403,
+      "step": 28985
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 3.3150155544281006,
+      "learning_rate": 3.0181700041257077e-05,
+      "loss": 1.0186,
+      "step": 28990
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 3.6737823486328125,
+      "learning_rate": 3.017595227244378e-05,
+      "loss": 1.0857,
+      "step": 28995
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 1.9824943542480469,
+      "learning_rate": 3.0170204217768023e-05,
+      "loss": 1.2406,
+      "step": 29000
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 1.4903932809829712,
+      "learning_rate": 3.016445587754726e-05,
+      "loss": 1.3973,
+      "step": 29005
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 3.9678359031677246,
+      "learning_rate": 3.0158707252098966e-05,
+      "loss": 1.1738,
+      "step": 29010
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 3.8614068031311035,
+      "learning_rate": 3.015295834174063e-05,
+      "loss": 1.123,
+      "step": 29015
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 4.1375555992126465,
+      "learning_rate": 3.0147209146789762e-05,
+      "loss": 0.9881,
+      "step": 29020
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 3.8305399417877197,
+      "learning_rate": 3.014145966756388e-05,
+      "loss": 0.8714,
+      "step": 29025
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 2.49973464012146,
+      "learning_rate": 3.013570990438053e-05,
+      "loss": 1.0092,
+      "step": 29030
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 6.754157543182373,
+      "learning_rate": 3.0129959857557256e-05,
+      "loss": 1.1094,
+      "step": 29035
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 3.1229982376098633,
+      "learning_rate": 3.0124209527411634e-05,
+      "loss": 1.0827,
+      "step": 29040
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 5.443647384643555,
+      "learning_rate": 3.0118458914261242e-05,
+      "loss": 1.0179,
+      "step": 29045
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 3.6279354095458984,
+      "learning_rate": 3.011270801842369e-05,
+      "loss": 1.0837,
+      "step": 29050
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 1.7841243743896484,
+      "learning_rate": 3.0106956840216586e-05,
+      "loss": 1.2112,
+      "step": 29055
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 1.7607803344726562,
+      "learning_rate": 3.0101205379957563e-05,
+      "loss": 1.2424,
+      "step": 29060
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 1.5843974351882935,
+      "learning_rate": 3.0095453637964272e-05,
+      "loss": 1.0776,
+      "step": 29065
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 12.485400199890137,
+      "learning_rate": 3.0089701614554377e-05,
+      "loss": 1.0256,
+      "step": 29070
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 1.710533618927002,
+      "learning_rate": 3.008394931004555e-05,
+      "loss": 1.1107,
+      "step": 29075
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 1.6686811447143555,
+      "learning_rate": 3.007819672475548e-05,
+      "loss": 1.1631,
+      "step": 29080
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 4.799095630645752,
+      "learning_rate": 3.0072443859001893e-05,
+      "loss": 1.1524,
+      "step": 29085
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 3.90974497795105,
+      "learning_rate": 3.0066690713102497e-05,
+      "loss": 1.0848,
+      "step": 29090
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 2.303661346435547,
+      "learning_rate": 3.006093728737504e-05,
+      "loss": 1.0947,
+      "step": 29095
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 1.7615519762039185,
+      "learning_rate": 3.005518358213728e-05,
+      "loss": 0.9884,
+      "step": 29100
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 3.437652826309204,
+      "learning_rate": 3.0049429597706987e-05,
+      "loss": 0.9993,
+      "step": 29105
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 1.8464926481246948,
+      "learning_rate": 3.0043675334401943e-05,
+      "loss": 1.3269,
+      "step": 29110
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 2.6147725582122803,
+      "learning_rate": 3.0037920792539954e-05,
+      "loss": 1.1523,
+      "step": 29115
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 1.6204800605773926,
+      "learning_rate": 3.003216597243883e-05,
+      "loss": 1.0221,
+      "step": 29120
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 2.307574510574341,
+      "learning_rate": 3.0026410874416416e-05,
+      "loss": 1.1662,
+      "step": 29125
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 1.2338721752166748,
+      "learning_rate": 3.0020655498790552e-05,
+      "loss": 1.1687,
+      "step": 29130
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 1.9817173480987549,
+      "learning_rate": 3.001489984587911e-05,
+      "loss": 1.1231,
+      "step": 29135
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 2.5606281757354736,
+      "learning_rate": 3.000914391599995e-05,
+      "loss": 1.1578,
+      "step": 29140
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 2.77673602104187,
+      "learning_rate": 3.0003387709470988e-05,
+      "loss": 1.1036,
+      "step": 29145
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 1.4961824417114258,
+      "learning_rate": 2.9997631226610116e-05,
+      "loss": 1.197,
+      "step": 29150
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 4.214254379272461,
+      "learning_rate": 2.9991874467735272e-05,
+      "loss": 1.1423,
+      "step": 29155
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 4.497615337371826,
+      "learning_rate": 2.998611743316439e-05,
+      "loss": 1.1546,
+      "step": 29160
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 1.615196943283081,
+      "learning_rate": 2.998036012321543e-05,
+      "loss": 1.2521,
+      "step": 29165
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 1.5576528310775757,
+      "learning_rate": 2.997460253820635e-05,
+      "loss": 1.2029,
+      "step": 29170
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 1.751317024230957,
+      "learning_rate": 2.996884467845514e-05,
+      "loss": 1.2419,
+      "step": 29175
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 4.62571907043457,
+      "learning_rate": 2.9963086544279807e-05,
+      "loss": 1.3243,
+      "step": 29180
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 3.067814826965332,
+      "learning_rate": 2.9957328135998365e-05,
+      "loss": 1.1107,
+      "step": 29185
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 5.224938869476318,
+      "learning_rate": 2.9951569453928834e-05,
+      "loss": 1.2911,
+      "step": 29190
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 3.1084907054901123,
+      "learning_rate": 2.9945810498389275e-05,
+      "loss": 1.0711,
+      "step": 29195
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 1.8605875968933105,
+      "learning_rate": 2.994005126969775e-05,
+      "loss": 1.1007,
+      "step": 29200
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 1.8477222919464111,
+      "learning_rate": 2.9934291768172324e-05,
+      "loss": 1.1518,
+      "step": 29205
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 2.743089199066162,
+      "learning_rate": 2.9928531994131086e-05,
+      "loss": 1.1243,
+      "step": 29210
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 3.1125733852386475,
+      "learning_rate": 2.9922771947892154e-05,
+      "loss": 1.2135,
+      "step": 29215
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 2.2629575729370117,
+      "learning_rate": 2.9917011629773643e-05,
+      "loss": 1.1818,
+      "step": 29220
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 3.8765952587127686,
+      "learning_rate": 2.991125104009369e-05,
+      "loss": 1.2061,
+      "step": 29225
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 2.9777157306671143,
+      "learning_rate": 2.9905490179170446e-05,
+      "loss": 1.2431,
+      "step": 29230
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 1.9893121719360352,
+      "learning_rate": 2.9899729047322085e-05,
+      "loss": 1.066,
+      "step": 29235
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 2.7022969722747803,
+      "learning_rate": 2.989396764486677e-05,
+      "loss": 1.1172,
+      "step": 29240
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 3.0575449466705322,
+      "learning_rate": 2.988820597212272e-05,
+      "loss": 1.0959,
+      "step": 29245
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 1.885754942893982,
+      "learning_rate": 2.988244402940813e-05,
+      "loss": 1.1691,
+      "step": 29250
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 1.6529167890548706,
+      "learning_rate": 2.9876681817041235e-05,
+      "loss": 1.2623,
+      "step": 29255
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 1.5923422574996948,
+      "learning_rate": 2.987091933534027e-05,
+      "loss": 1.2028,
+      "step": 29260
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 1.1636615991592407,
+      "learning_rate": 2.9865156584623495e-05,
+      "loss": 0.9298,
+      "step": 29265
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 2.3403260707855225,
+      "learning_rate": 2.9859393565209177e-05,
+      "loss": 1.1567,
+      "step": 29270
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 3.294311761856079,
+      "learning_rate": 2.98536302774156e-05,
+      "loss": 0.9932,
+      "step": 29275
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 1.9269850254058838,
+      "learning_rate": 2.9847866721561075e-05,
+      "loss": 1.2663,
+      "step": 29280
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 4.520174026489258,
+      "learning_rate": 2.9842102897963902e-05,
+      "loss": 1.0276,
+      "step": 29285
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 1.383296012878418,
+      "learning_rate": 2.9836338806942425e-05,
+      "loss": 1.1108,
+      "step": 29290
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 2.5761561393737793,
+      "learning_rate": 2.9830574448814984e-05,
+      "loss": 1.1263,
+      "step": 29295
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 2.923858404159546,
+      "learning_rate": 2.9824809823899936e-05,
+      "loss": 1.1611,
+      "step": 29300
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 1.9109852313995361,
+      "learning_rate": 2.9819044932515655e-05,
+      "loss": 1.0511,
+      "step": 29305
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 2.4346346855163574,
+      "learning_rate": 2.9813279774980524e-05,
+      "loss": 1.1955,
+      "step": 29310
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 3.335416078567505,
+      "learning_rate": 2.9807514351612965e-05,
+      "loss": 1.086,
+      "step": 29315
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 3.1424591541290283,
+      "learning_rate": 2.9801748662731376e-05,
+      "loss": 1.2607,
+      "step": 29320
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 3.5147266387939453,
+      "learning_rate": 2.9795982708654206e-05,
+      "loss": 1.0876,
+      "step": 29325
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 3.8521475791931152,
+      "learning_rate": 2.97902164896999e-05,
+      "loss": 1.1177,
+      "step": 29330
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 4.234015464782715,
+      "learning_rate": 2.9784450006186914e-05,
+      "loss": 1.0803,
+      "step": 29335
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 1.5537238121032715,
+      "learning_rate": 2.9778683258433727e-05,
+      "loss": 1.1356,
+      "step": 29340
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 3.8383750915527344,
+      "learning_rate": 2.9772916246758825e-05,
+      "loss": 1.0342,
+      "step": 29345
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 2.1243529319763184,
+      "learning_rate": 2.9767148971480725e-05,
+      "loss": 1.0873,
+      "step": 29350
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 4.195858478546143,
+      "learning_rate": 2.976138143291794e-05,
+      "loss": 1.1394,
+      "step": 29355
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 2.7369132041931152,
+      "learning_rate": 2.9755613631389017e-05,
+      "loss": 1.0517,
+      "step": 29360
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 1.731953740119934,
+      "learning_rate": 2.9749845567212487e-05,
+      "loss": 1.1122,
+      "step": 29365
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 2.9949727058410645,
+      "learning_rate": 2.9744077240706925e-05,
+      "loss": 1.192,
+      "step": 29370
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 3.624964714050293,
+      "learning_rate": 2.9738308652190905e-05,
+      "loss": 1.1176,
+      "step": 29375
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 1.6444040536880493,
+      "learning_rate": 2.973253980198303e-05,
+      "loss": 1.1123,
+      "step": 29380
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 3.1505863666534424,
+      "learning_rate": 2.9726770690401894e-05,
+      "loss": 1.1554,
+      "step": 29385
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 2.775284767150879,
+      "learning_rate": 2.972100131776613e-05,
+      "loss": 1.0491,
+      "step": 29390
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 2.9408364295959473,
+      "learning_rate": 2.971523168439437e-05,
+      "loss": 1.0462,
+      "step": 29395
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 4.636882305145264,
+      "learning_rate": 2.9709461790605263e-05,
+      "loss": 1.1156,
+      "step": 29400
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 4.47889518737793,
+      "learning_rate": 2.970369163671748e-05,
+      "loss": 1.1295,
+      "step": 29405
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 5.22827672958374,
+      "learning_rate": 2.9697921223049697e-05,
+      "loss": 1.0539,
+      "step": 29410
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 2.5013325214385986,
+      "learning_rate": 2.9692150549920606e-05,
+      "loss": 1.1867,
+      "step": 29415
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 1.6474865674972534,
+      "learning_rate": 2.9686379617648917e-05,
+      "loss": 1.0142,
+      "step": 29420
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 2.7693967819213867,
+      "learning_rate": 2.9680608426553358e-05,
+      "loss": 1.0996,
+      "step": 29425
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 3.4689788818359375,
+      "learning_rate": 2.9674836976952657e-05,
+      "loss": 1.3735,
+      "step": 29430
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 1.6215572357177734,
+      "learning_rate": 2.966906526916557e-05,
+      "loss": 1.1179,
+      "step": 29435
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 2.069563627243042,
+      "learning_rate": 2.9663293303510857e-05,
+      "loss": 1.0961,
+      "step": 29440
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 1.665561318397522,
+      "learning_rate": 2.9657521080307305e-05,
+      "loss": 1.1638,
+      "step": 29445
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 3.0357062816619873,
+      "learning_rate": 2.9651748599873708e-05,
+      "loss": 1.2004,
+      "step": 29450
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 7.12582540512085,
+      "learning_rate": 2.9645975862528868e-05,
+      "loss": 1.1028,
+      "step": 29455
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 4.670501232147217,
+      "learning_rate": 2.9640202868591616e-05,
+      "loss": 0.9468,
+      "step": 29460
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 4.492547035217285,
+      "learning_rate": 2.9634429618380775e-05,
+      "loss": 1.161,
+      "step": 29465
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 4.036949634552002,
+      "learning_rate": 2.9628656112215202e-05,
+      "loss": 1.0556,
+      "step": 29470
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 6.104241847991943,
+      "learning_rate": 2.962288235041377e-05,
+      "loss": 1.2953,
+      "step": 29475
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 2.0466535091400146,
+      "learning_rate": 2.9617108333295345e-05,
+      "loss": 1.2064,
+      "step": 29480
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 3.5240540504455566,
+      "learning_rate": 2.9612488935986672e-05,
+      "loss": 1.0582,
+      "step": 29485
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 2.5299673080444336,
+      "learning_rate": 2.960671446010129e-05,
+      "loss": 1.1523,
+      "step": 29490
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 2.4263341426849365,
+      "learning_rate": 2.9600939729791864e-05,
+      "loss": 1.0886,
+      "step": 29495
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 5.209130764007568,
+      "learning_rate": 2.9595164745377314e-05,
+      "loss": 1.0808,
+      "step": 29500
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 2.5315446853637695,
+      "learning_rate": 2.958938950717659e-05,
+      "loss": 1.264,
+      "step": 29505
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 2.663262367248535,
+      "learning_rate": 2.9583614015508666e-05,
+      "loss": 1.201,
+      "step": 29510
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 1.5915833711624146,
+      "learning_rate": 2.9577838270692493e-05,
+      "loss": 1.0883,
+      "step": 29515
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 6.19256067276001,
+      "learning_rate": 2.9572062273047075e-05,
+      "loss": 1.1954,
+      "step": 29520
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 2.0903115272521973,
+      "learning_rate": 2.9566286022891404e-05,
+      "loss": 1.1074,
+      "step": 29525
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 7.435312271118164,
+      "learning_rate": 2.9560509520544505e-05,
+      "loss": 0.967,
+      "step": 29530
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 1.675291657447815,
+      "learning_rate": 2.9554732766325406e-05,
+      "loss": 1.0977,
+      "step": 29535
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 2.438756227493286,
+      "learning_rate": 2.9548955760553155e-05,
+      "loss": 1.1914,
+      "step": 29540
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 2.381863832473755,
+      "learning_rate": 2.9543178503546805e-05,
+      "loss": 1.0097,
+      "step": 29545
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 2.3754689693450928,
+      "learning_rate": 2.9537400995625426e-05,
+      "loss": 1.2332,
+      "step": 29550
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 1.7114568948745728,
+      "learning_rate": 2.9531623237108103e-05,
+      "loss": 1.1569,
+      "step": 29555
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 2.6389689445495605,
+      "learning_rate": 2.952584522831394e-05,
+      "loss": 1.2255,
+      "step": 29560
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 1.7149155139923096,
+      "learning_rate": 2.9520066969562056e-05,
+      "loss": 1.0352,
+      "step": 29565
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 4.823055744171143,
+      "learning_rate": 2.9514288461171557e-05,
+      "loss": 1.2276,
+      "step": 29570
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 1.9236773252487183,
+      "learning_rate": 2.9508509703461613e-05,
+      "loss": 1.3419,
+      "step": 29575
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 0.9075140357017517,
+      "learning_rate": 2.9502730696751362e-05,
+      "loss": 0.9838,
+      "step": 29580
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 1.5596219301223755,
+      "learning_rate": 2.949695144135997e-05,
+      "loss": 1.1306,
+      "step": 29585
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 4.231356620788574,
+      "learning_rate": 2.9491171937606628e-05,
+      "loss": 1.13,
+      "step": 29590
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 1.3872261047363281,
+      "learning_rate": 2.948539218581053e-05,
+      "loss": 1.1116,
+      "step": 29595
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 2.4153921604156494,
+      "learning_rate": 2.9479612186290878e-05,
+      "loss": 1.1196,
+      "step": 29600
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 2.1678645610809326,
+      "learning_rate": 2.9473831939366908e-05,
+      "loss": 1.0154,
+      "step": 29605
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 5.16516637802124,
+      "learning_rate": 2.9468051445357848e-05,
+      "loss": 0.9722,
+      "step": 29610
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 1.94868004322052,
+      "learning_rate": 2.946227070458295e-05,
+      "loss": 1.1472,
+      "step": 29615
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 1.3372802734375,
+      "learning_rate": 2.9456489717361478e-05,
+      "loss": 1.1834,
+      "step": 29620
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 1.819388508796692,
+      "learning_rate": 2.9450708484012713e-05,
+      "loss": 1.0856,
+      "step": 29625
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 2.233203649520874,
+      "learning_rate": 2.9444927004855942e-05,
+      "loss": 1.2613,
+      "step": 29630
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 1.7675302028656006,
+      "learning_rate": 2.9439145280210473e-05,
+      "loss": 1.1527,
+      "step": 29635
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 3.0035910606384277,
+      "learning_rate": 2.9433363310395634e-05,
+      "loss": 0.9933,
+      "step": 29640
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 1.98517906665802,
+      "learning_rate": 2.942758109573074e-05,
+      "loss": 1.2819,
+      "step": 29645
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 2.4812839031219482,
+      "learning_rate": 2.9421798636535137e-05,
+      "loss": 0.9064,
+      "step": 29650
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 3.8419299125671387,
+      "learning_rate": 2.9416015933128194e-05,
+      "loss": 1.1614,
+      "step": 29655
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 8.991368293762207,
+      "learning_rate": 2.941023298582929e-05,
+      "loss": 1.0835,
+      "step": 29660
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 1.9285551309585571,
+      "learning_rate": 2.9404449794957795e-05,
+      "loss": 1.1749,
+      "step": 29665
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 4.323269844055176,
+      "learning_rate": 2.9398666360833126e-05,
+      "loss": 1.0456,
+      "step": 29670
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 3.1791858673095703,
+      "learning_rate": 2.9392882683774674e-05,
+      "loss": 1.0455,
+      "step": 29675
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 1.171151041984558,
+      "learning_rate": 2.9387098764101882e-05,
+      "loss": 1.1266,
+      "step": 29680
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 3.656491994857788,
+      "learning_rate": 2.938131460213419e-05,
+      "loss": 1.1155,
+      "step": 29685
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 3.314228057861328,
+      "learning_rate": 2.937553019819104e-05,
+      "loss": 1.0785,
+      "step": 29690
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 2.453927516937256,
+      "learning_rate": 2.936974555259191e-05,
+      "loss": 1.2502,
+      "step": 29695
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 1.7844200134277344,
+      "learning_rate": 2.936396066565627e-05,
+      "loss": 1.2183,
+      "step": 29700
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 3.409106731414795,
+      "learning_rate": 2.9358175537703624e-05,
+      "loss": 1.1222,
+      "step": 29705
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 7.029350757598877,
+      "learning_rate": 2.9352390169053475e-05,
+      "loss": 1.1221,
+      "step": 29710
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 5.586660861968994,
+      "learning_rate": 2.9346604560025336e-05,
+      "loss": 1.1623,
+      "step": 29715
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 2.2339181900024414,
+      "learning_rate": 2.9340818710938745e-05,
+      "loss": 1.2503,
+      "step": 29720
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 2.694247245788574,
+      "learning_rate": 2.9335032622113252e-05,
+      "loss": 1.1939,
+      "step": 29725
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 1.8695008754730225,
+      "learning_rate": 2.9329246293868407e-05,
+      "loss": 1.1074,
+      "step": 29730
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 3.689002275466919,
+      "learning_rate": 2.9323459726523794e-05,
+      "loss": 0.9559,
+      "step": 29735
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 2.7362306118011475,
+      "learning_rate": 2.9317672920398993e-05,
+      "loss": 1.0817,
+      "step": 29740
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 1.8381929397583008,
+      "learning_rate": 2.9311885875813604e-05,
+      "loss": 1.1532,
+      "step": 29745
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 2.022266149520874,
+      "learning_rate": 2.930609859308724e-05,
+      "loss": 0.9327,
+      "step": 29750
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 3.2312119007110596,
+      "learning_rate": 2.9300311072539527e-05,
+      "loss": 0.9719,
+      "step": 29755
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 7.237270355224609,
+      "learning_rate": 2.92945233144901e-05,
+      "loss": 1.172,
+      "step": 29760
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 2.9999465942382812,
+      "learning_rate": 2.928873531925862e-05,
+      "loss": 1.0583,
+      "step": 29765
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 3.0304183959960938,
+      "learning_rate": 2.928294708716475e-05,
+      "loss": 1.0234,
+      "step": 29770
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 1.2405683994293213,
+      "learning_rate": 2.927715861852816e-05,
+      "loss": 0.9056,
+      "step": 29775
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 4.732754230499268,
+      "learning_rate": 2.9271369913668546e-05,
+      "loss": 1.2299,
+      "step": 29780
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 2.121570348739624,
+      "learning_rate": 2.9265580972905603e-05,
+      "loss": 1.0564,
+      "step": 29785
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 3.0892438888549805,
+      "learning_rate": 2.9259791796559066e-05,
+      "loss": 1.0464,
+      "step": 29790
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 2.665210247039795,
+      "learning_rate": 2.9254002384948655e-05,
+      "loss": 1.1263,
+      "step": 29795
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 2.6491611003875732,
+      "learning_rate": 2.9248212738394116e-05,
+      "loss": 1.0433,
+      "step": 29800
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 4.1141815185546875,
+      "learning_rate": 2.9242422857215195e-05,
+      "loss": 1.134,
+      "step": 29805
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 6.5142951011657715,
+      "learning_rate": 2.9236632741731673e-05,
+      "loss": 1.2787,
+      "step": 29810
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 2.864144802093506,
+      "learning_rate": 2.923084239226333e-05,
+      "loss": 1.0214,
+      "step": 29815
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 6.211786270141602,
+      "learning_rate": 2.922505180912996e-05,
+      "loss": 1.0869,
+      "step": 29820
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 4.338846206665039,
+      "learning_rate": 2.921926099265137e-05,
+      "loss": 1.1458,
+      "step": 29825
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 4.332965850830078,
+      "learning_rate": 2.9213469943147374e-05,
+      "loss": 1.2077,
+      "step": 29830
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 2.471970319747925,
+      "learning_rate": 2.920767866093782e-05,
+      "loss": 1.1077,
+      "step": 29835
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 1.8260835409164429,
+      "learning_rate": 2.920188714634255e-05,
+      "loss": 1.0752,
+      "step": 29840
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 8.06219482421875,
+      "learning_rate": 2.919609539968141e-05,
+      "loss": 1.1127,
+      "step": 29845
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 3.8866000175476074,
+      "learning_rate": 2.9190303421274288e-05,
+      "loss": 1.065,
+      "step": 29850
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 3.3143255710601807,
+      "learning_rate": 2.918451121144107e-05,
+      "loss": 1.0448,
+      "step": 29855
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 6.668207168579102,
+      "learning_rate": 2.9178718770501638e-05,
+      "loss": 1.0846,
+      "step": 29860
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 2.2254040241241455,
+      "learning_rate": 2.917292609877592e-05,
+      "loss": 1.1679,
+      "step": 29865
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 2.8223040103912354,
+      "learning_rate": 2.916713319658383e-05,
+      "loss": 0.9145,
+      "step": 29870
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 2.584289312362671,
+      "learning_rate": 2.9161340064245302e-05,
+      "loss": 1.0576,
+      "step": 29875
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 2.3711581230163574,
+      "learning_rate": 2.9155546702080282e-05,
+      "loss": 1.0433,
+      "step": 29880
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 3.263002634048462,
+      "learning_rate": 2.9149753110408744e-05,
+      "loss": 1.0901,
+      "step": 29885
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 3.8075244426727295,
+      "learning_rate": 2.9143959289550653e-05,
+      "loss": 0.8897,
+      "step": 29890
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 4.947206497192383,
+      "learning_rate": 2.913816523982601e-05,
+      "loss": 1.147,
+      "step": 29895
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 1.8365360498428345,
+      "learning_rate": 2.913237096155479e-05,
+      "loss": 1.0571,
+      "step": 29900
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 2.5439870357513428,
+      "learning_rate": 2.9126576455057014e-05,
+      "loss": 1.062,
+      "step": 29905
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 3.2245960235595703,
+      "learning_rate": 2.9120781720652713e-05,
+      "loss": 1.2575,
+      "step": 29910
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 6.086019039154053,
+      "learning_rate": 2.9114986758661922e-05,
+      "loss": 1.1212,
+      "step": 29915
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 2.142395496368408,
+      "learning_rate": 2.9109191569404693e-05,
+      "loss": 1.0457,
+      "step": 29920
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 2.1010286808013916,
+      "learning_rate": 2.9104555254582145e-05,
+      "loss": 1.0763,
+      "step": 29925
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 2.635202407836914,
+      "learning_rate": 2.909875965705188e-05,
+      "loss": 1.2506,
+      "step": 29930
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 2.7703280448913574,
+      "learning_rate": 2.9092963833151388e-05,
+      "loss": 1.1693,
+      "step": 29935
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 5.17828369140625,
+      "learning_rate": 2.9087167783200752e-05,
+      "loss": 1.1341,
+      "step": 29940
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 3.351463556289673,
+      "learning_rate": 2.908137150752008e-05,
+      "loss": 1.1165,
+      "step": 29945
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 2.4802517890930176,
+      "learning_rate": 2.9075575006429524e-05,
+      "loss": 1.1043,
+      "step": 29950
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 1.3321456909179688,
+      "learning_rate": 2.9069778280249183e-05,
+      "loss": 1.0579,
+      "step": 29955
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 5.725347518920898,
+      "learning_rate": 2.9063981329299216e-05,
+      "loss": 1.0917,
+      "step": 29960
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 1.3497544527053833,
+      "learning_rate": 2.905818415389978e-05,
+      "loss": 1.1703,
+      "step": 29965
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 2.001232385635376,
+      "learning_rate": 2.9052386754371065e-05,
+      "loss": 0.9471,
+      "step": 29970
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 2.2034993171691895,
+      "learning_rate": 2.904658913103323e-05,
+      "loss": 1.2491,
+      "step": 29975
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 7.609285354614258,
+      "learning_rate": 2.9040791284206493e-05,
+      "loss": 1.062,
+      "step": 29980
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 3.246325731277466,
+      "learning_rate": 2.9034993214211048e-05,
+      "loss": 1.1693,
+      "step": 29985
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 1.247209906578064,
+      "learning_rate": 2.902919492136712e-05,
+      "loss": 1.2194,
+      "step": 29990
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 3.488403081893921,
+      "learning_rate": 2.9023396405994946e-05,
+      "loss": 1.2346,
+      "step": 29995
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 1.7326892614364624,
+      "learning_rate": 2.901759766841477e-05,
+      "loss": 1.0722,
+      "step": 30000
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 4.2910356521606445,
+      "learning_rate": 2.901179870894685e-05,
+      "loss": 1.0043,
+      "step": 30005
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 2.8054089546203613,
+      "learning_rate": 2.900599952791146e-05,
+      "loss": 1.1417,
+      "step": 30010
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 4.5647687911987305,
+      "learning_rate": 2.9000200125628885e-05,
+      "loss": 1.1773,
+      "step": 30015
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 1.9864246845245361,
+      "learning_rate": 2.899440050241941e-05,
+      "loss": 1.212,
+      "step": 30020
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 3.7864115238189697,
+      "learning_rate": 2.898860065860335e-05,
+      "loss": 0.9945,
+      "step": 30025
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 2.867279529571533,
+      "learning_rate": 2.8982800594501014e-05,
+      "loss": 1.1722,
+      "step": 30030
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 1.8707783222198486,
+      "learning_rate": 2.8977000310432744e-05,
+      "loss": 1.2435,
+      "step": 30035
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 3.260164260864258,
+      "learning_rate": 2.8971199806718884e-05,
+      "loss": 1.2162,
+      "step": 30040
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 3.576340436935425,
+      "learning_rate": 2.896539908367979e-05,
+      "loss": 1.0317,
+      "step": 30045
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 2.4378347396850586,
+      "learning_rate": 2.8959598141635826e-05,
+      "loss": 1.0759,
+      "step": 30050
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 3.9936773777008057,
+      "learning_rate": 2.8953796980907365e-05,
+      "loss": 1.1189,
+      "step": 30055
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 2.5489132404327393,
+      "learning_rate": 2.894799560181481e-05,
+      "loss": 1.0809,
+      "step": 30060
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 2.193693161010742,
+      "learning_rate": 2.894219400467856e-05,
+      "loss": 1.1111,
+      "step": 30065
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 2.2259438037872314,
+      "learning_rate": 2.8936392189819034e-05,
+      "loss": 1.2037,
+      "step": 30070
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 6.183416843414307,
+      "learning_rate": 2.893059015755666e-05,
+      "loss": 1.2801,
+      "step": 30075
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 2.10697865486145,
+      "learning_rate": 2.892478790821187e-05,
+      "loss": 1.089,
+      "step": 30080
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 2.224777936935425,
+      "learning_rate": 2.8918985442105128e-05,
+      "loss": 1.1623,
+      "step": 30085
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 2.0744781494140625,
+      "learning_rate": 2.8913182759556894e-05,
+      "loss": 0.9921,
+      "step": 30090
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 3.06063175201416,
+      "learning_rate": 2.8907379860887645e-05,
+      "loss": 1.0767,
+      "step": 30095
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 1.988356590270996,
+      "learning_rate": 2.890157674641787e-05,
+      "loss": 1.0859,
+      "step": 30100
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 1.3868892192840576,
+      "learning_rate": 2.8895773416468063e-05,
+      "loss": 1.118,
+      "step": 30105
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 2.165102481842041,
+      "learning_rate": 2.8889969871358746e-05,
+      "loss": 1.0599,
+      "step": 30110
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 1.76089346408844,
+      "learning_rate": 2.888416611141043e-05,
+      "loss": 1.1494,
+      "step": 30115
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 3.6625218391418457,
+      "learning_rate": 2.887836213694366e-05,
+      "loss": 1.0324,
+      "step": 30120
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 2.0290017127990723,
+      "learning_rate": 2.8872557948278976e-05,
+      "loss": 1.1136,
+      "step": 30125
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 1.7238839864730835,
+      "learning_rate": 2.8866753545736946e-05,
+      "loss": 1.0594,
+      "step": 30130
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 3.4980666637420654,
+      "learning_rate": 2.8860948929638136e-05,
+      "loss": 0.9758,
+      "step": 30135
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 4.032216548919678,
+      "learning_rate": 2.885514410030313e-05,
+      "loss": 0.999,
+      "step": 30140
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 1.803316593170166,
+      "learning_rate": 2.8849339058052526e-05,
+      "loss": 1.188,
+      "step": 30145
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 1.5701510906219482,
+      "learning_rate": 2.8843533803206923e-05,
+      "loss": 1.0158,
+      "step": 30150
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 1.275262713432312,
+      "learning_rate": 2.8837728336086946e-05,
+      "loss": 0.9069,
+      "step": 30155
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 2.408705949783325,
+      "learning_rate": 2.8831922657013216e-05,
+      "loss": 1.0335,
+      "step": 30160
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 2.678744316101074,
+      "learning_rate": 2.8826116766306383e-05,
+      "loss": 1.0944,
+      "step": 30165
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 3.0019614696502686,
+      "learning_rate": 2.8820310664287096e-05,
+      "loss": 0.9107,
+      "step": 30170
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 1.4615614414215088,
+      "learning_rate": 2.881450435127603e-05,
+      "loss": 1.0898,
+      "step": 30175
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 1.9362547397613525,
+      "learning_rate": 2.8808697827593845e-05,
+      "loss": 1.1276,
+      "step": 30180
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 2.6547698974609375,
+      "learning_rate": 2.880289109356124e-05,
+      "loss": 1.2548,
+      "step": 30185
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 7.905858516693115,
+      "learning_rate": 2.8797084149498915e-05,
+      "loss": 0.9404,
+      "step": 30190
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 1.2215291261672974,
+      "learning_rate": 2.879127699572758e-05,
+      "loss": 1.003,
+      "step": 30195
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 1.8280051946640015,
+      "learning_rate": 2.878546963256795e-05,
+      "loss": 1.0391,
+      "step": 30200
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 1.2926268577575684,
+      "learning_rate": 2.8779662060340778e-05,
+      "loss": 1.2109,
+      "step": 30205
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 1.7577592134475708,
+      "learning_rate": 2.8773854279366797e-05,
+      "loss": 1.1695,
+      "step": 30210
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 2.365588903427124,
+      "learning_rate": 2.8768046289966766e-05,
+      "loss": 1.1072,
+      "step": 30215
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 2.1536929607391357,
+      "learning_rate": 2.8762238092461447e-05,
+      "loss": 1.1535,
+      "step": 30220
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 2.988088369369507,
+      "learning_rate": 2.8756429687171637e-05,
+      "loss": 0.9861,
+      "step": 30225
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 3.4814767837524414,
+      "learning_rate": 2.8750621074418115e-05,
+      "loss": 1.3146,
+      "step": 30230
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 1.8867098093032837,
+      "learning_rate": 2.874481225452169e-05,
+      "loss": 1.037,
+      "step": 30235
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 1.5579580068588257,
+      "learning_rate": 2.8739003227803184e-05,
+      "loss": 1.1022,
+      "step": 30240
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 3.373805284500122,
+      "learning_rate": 2.8733193994583412e-05,
+      "loss": 1.1459,
+      "step": 30245
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 2.02644419670105,
+      "learning_rate": 2.8727384555183217e-05,
+      "loss": 1.1493,
+      "step": 30250
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 2.24518084526062,
+      "learning_rate": 2.8721574909923445e-05,
+      "loss": 1.0157,
+      "step": 30255
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 1.2897074222564697,
+      "learning_rate": 2.871576505912496e-05,
+      "loss": 1.128,
+      "step": 30260
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 1.510237693786621,
+      "learning_rate": 2.8709955003108636e-05,
+      "loss": 1.1292,
+      "step": 30265
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 1.2689833641052246,
+      "learning_rate": 2.870414474219535e-05,
+      "loss": 1.0674,
+      "step": 30270
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 1.3254342079162598,
+      "learning_rate": 2.8698334276705995e-05,
+      "loss": 1.1342,
+      "step": 30275
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 2.8477869033813477,
+      "learning_rate": 2.8692523606961492e-05,
+      "loss": 1.1701,
+      "step": 30280
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 1.7730213403701782,
+      "learning_rate": 2.868671273328274e-05,
+      "loss": 1.1741,
+      "step": 30285
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 1.311159610748291,
+      "learning_rate": 2.8680901655990678e-05,
+      "loss": 1.0136,
+      "step": 30290
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 1.5597997903823853,
+      "learning_rate": 2.867509037540625e-05,
+      "loss": 1.07,
+      "step": 30295
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 3.055157423019409,
+      "learning_rate": 2.8669278891850392e-05,
+      "loss": 1.2124,
+      "step": 30300
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 2.4278862476348877,
+      "learning_rate": 2.8663467205644086e-05,
+      "loss": 0.9972,
+      "step": 30305
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 2.5763089656829834,
+      "learning_rate": 2.8657655317108284e-05,
+      "loss": 0.9838,
+      "step": 30310
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 2.0975847244262695,
+      "learning_rate": 2.8651843226563983e-05,
+      "loss": 1.1683,
+      "step": 30315
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 1.7405827045440674,
+      "learning_rate": 2.864603093433218e-05,
+      "loss": 1.0669,
+      "step": 30320
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 7.420092582702637,
+      "learning_rate": 2.8640218440733875e-05,
+      "loss": 1.015,
+      "step": 30325
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 3.352482795715332,
+      "learning_rate": 2.8634405746090088e-05,
+      "loss": 1.0018,
+      "step": 30330
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 4.274402141571045,
+      "learning_rate": 2.8628592850721857e-05,
+      "loss": 1.1142,
+      "step": 30335
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 1.6124602556228638,
+      "learning_rate": 2.862277975495021e-05,
+      "loss": 1.1566,
+      "step": 30340
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 1.762075424194336,
+      "learning_rate": 2.8616966459096202e-05,
+      "loss": 1.037,
+      "step": 30345
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 3.1598358154296875,
+      "learning_rate": 2.8611152963480892e-05,
+      "loss": 1.0827,
+      "step": 30350
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 4.015747547149658,
+      "learning_rate": 2.8605339268425363e-05,
+      "loss": 0.9845,
+      "step": 30355
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 6.202319145202637,
+      "learning_rate": 2.8599525374250684e-05,
+      "loss": 1.1009,
+      "step": 30360
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 3.212571620941162,
+      "learning_rate": 2.859371128127797e-05,
+      "loss": 1.1507,
+      "step": 30365
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 2.252204656600952,
+      "learning_rate": 2.8587896989828323e-05,
+      "loss": 1.1523,
+      "step": 30370
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 2.1097633838653564,
+      "learning_rate": 2.8582082500222845e-05,
+      "loss": 1.0826,
+      "step": 30375
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 2.7466039657592773,
+      "learning_rate": 2.8576267812782675e-05,
+      "loss": 1.1409,
+      "step": 30380
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 8.023296356201172,
+      "learning_rate": 2.857045292782895e-05,
+      "loss": 0.9465,
+      "step": 30385
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 3.5571467876434326,
+      "learning_rate": 2.8564637845682823e-05,
+      "loss": 1.1629,
+      "step": 30390
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 3.21684193611145,
+      "learning_rate": 2.8558822566665454e-05,
+      "loss": 1.0213,
+      "step": 30395
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 3.065678596496582,
+      "learning_rate": 2.8553007091098016e-05,
+      "loss": 0.9455,
+      "step": 30400
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 3.9140865802764893,
+      "learning_rate": 2.8547191419301687e-05,
+      "loss": 0.974,
+      "step": 30405
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 4.4788641929626465,
+      "learning_rate": 2.854137555159766e-05,
+      "loss": 0.9984,
+      "step": 30410
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 1.5970557928085327,
+      "learning_rate": 2.8535559488307145e-05,
+      "loss": 1.1595,
+      "step": 30415
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 2.536341905593872,
+      "learning_rate": 2.8529743229751354e-05,
+      "loss": 1.1607,
+      "step": 30420
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 2.384742259979248,
+      "learning_rate": 2.8523926776251514e-05,
+      "loss": 1.0816,
+      "step": 30425
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 4.2098388671875,
+      "learning_rate": 2.8518110128128863e-05,
+      "loss": 1.0073,
+      "step": 30430
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 1.5726318359375,
+      "learning_rate": 2.851229328570465e-05,
+      "loss": 1.0309,
+      "step": 30435
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 2.4862496852874756,
+      "learning_rate": 2.850647624930012e-05,
+      "loss": 1.1971,
+      "step": 30440
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 3.5220296382904053,
+      "learning_rate": 2.8500659019236553e-05,
+      "loss": 1.0254,
+      "step": 30445
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 5.456885814666748,
+      "learning_rate": 2.8494841595835226e-05,
+      "loss": 0.7803,
+      "step": 30450
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 3.647815704345703,
+      "learning_rate": 2.8489023979417435e-05,
+      "loss": 1.0921,
+      "step": 30455
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 3.8451616764068604,
+      "learning_rate": 2.8483206170304473e-05,
+      "loss": 1.048,
+      "step": 30460
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 1.4922055006027222,
+      "learning_rate": 2.8477388168817664e-05,
+      "loss": 1.0033,
+      "step": 30465
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 3.4989609718322754,
+      "learning_rate": 2.847156997527831e-05,
+      "loss": 1.2043,
+      "step": 30470
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 2.6432862281799316,
+      "learning_rate": 2.8465751590007762e-05,
+      "loss": 1.0925,
+      "step": 30475
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 2.137115478515625,
+      "learning_rate": 2.845993301332735e-05,
+      "loss": 1.1623,
+      "step": 30480
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 3.7192418575286865,
+      "learning_rate": 2.845411424555844e-05,
+      "loss": 1.0968,
+      "step": 30485
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 2.331989049911499,
+      "learning_rate": 2.8448295287022386e-05,
+      "loss": 1.066,
+      "step": 30490
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 2.4555537700653076,
+      "learning_rate": 2.8442476138040568e-05,
+      "loss": 1.0104,
+      "step": 30495
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 4.285811901092529,
+      "learning_rate": 2.8436656798934376e-05,
+      "loss": 1.0691,
+      "step": 30500
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 1.4364317655563354,
+      "learning_rate": 2.8430837270025196e-05,
+      "loss": 1.1077,
+      "step": 30505
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 1.1822254657745361,
+      "learning_rate": 2.842501755163444e-05,
+      "loss": 1.0285,
+      "step": 30510
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 7.019848823547363,
+      "learning_rate": 2.8419197644083527e-05,
+      "loss": 1.1324,
+      "step": 30515
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 1.9157739877700806,
+      "learning_rate": 2.841337754769388e-05,
+      "loss": 1.1563,
+      "step": 30520
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 2.709386110305786,
+      "learning_rate": 2.8407557262786945e-05,
+      "loss": 1.1656,
+      "step": 30525
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 4.894626140594482,
+      "learning_rate": 2.8401736789684153e-05,
+      "loss": 1.1094,
+      "step": 30530
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 2.685741662979126,
+      "learning_rate": 2.839591612870698e-05,
+      "loss": 0.9865,
+      "step": 30535
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 2.1641218662261963,
+      "learning_rate": 2.8390095280176894e-05,
+      "loss": 1.1466,
+      "step": 30540
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 3.6387031078338623,
+      "learning_rate": 2.838427424441536e-05,
+      "loss": 1.0634,
+      "step": 30545
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 2.7952628135681152,
+      "learning_rate": 2.8378453021743882e-05,
+      "loss": 1.104,
+      "step": 30550
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 2.7595205307006836,
+      "learning_rate": 2.837263161248396e-05,
+      "loss": 1.1497,
+      "step": 30555
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 2.8632593154907227,
+      "learning_rate": 2.8366810016957096e-05,
+      "loss": 0.8454,
+      "step": 30560
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 2.603904962539673,
+      "learning_rate": 2.8360988235484814e-05,
+      "loss": 1.0677,
+      "step": 30565
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 3.4272186756134033,
+      "learning_rate": 2.8355166268388643e-05,
+      "loss": 1.0451,
+      "step": 30570
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 2.422112226486206,
+      "learning_rate": 2.8349344115990127e-05,
+      "loss": 1.1719,
+      "step": 30575
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 4.006277561187744,
+      "learning_rate": 2.834352177861082e-05,
+      "loss": 1.0388,
+      "step": 30580
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 3.0967395305633545,
+      "learning_rate": 2.8337699256572282e-05,
+      "loss": 0.8718,
+      "step": 30585
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 3.621616840362549,
+      "learning_rate": 2.833187655019608e-05,
+      "loss": 1.2271,
+      "step": 30590
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 1.8164085149765015,
+      "learning_rate": 2.8326053659803803e-05,
+      "loss": 1.0748,
+      "step": 30595
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 5.264411926269531,
+      "learning_rate": 2.8320230585717034e-05,
+      "loss": 1.0605,
+      "step": 30600
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 2.898695707321167,
+      "learning_rate": 2.831440732825738e-05,
+      "loss": 1.0845,
+      "step": 30605
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 1.863917350769043,
+      "learning_rate": 2.830858388774646e-05,
+      "loss": 1.1017,
+      "step": 30610
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 2.258936643600464,
+      "learning_rate": 2.830276026450589e-05,
+      "loss": 1.2552,
+      "step": 30615
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 1.4957672357559204,
+      "learning_rate": 2.82969364588573e-05,
+      "loss": 1.1517,
+      "step": 30620
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 4.283958911895752,
+      "learning_rate": 2.8291112471122338e-05,
+      "loss": 1.2211,
+      "step": 30625
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 2.875387191772461,
+      "learning_rate": 2.8285288301622658e-05,
+      "loss": 1.1582,
+      "step": 30630
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 3.575859546661377,
+      "learning_rate": 2.827946395067991e-05,
+      "loss": 1.2838,
+      "step": 30635
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 3.4343674182891846,
+      "learning_rate": 2.8273639418615788e-05,
+      "loss": 1.0276,
+      "step": 30640
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 2.2512946128845215,
+      "learning_rate": 2.826781470575196e-05,
+      "loss": 0.8439,
+      "step": 30645
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 3.449051856994629,
+      "learning_rate": 2.826198981241012e-05,
+      "loss": 1.1702,
+      "step": 30650
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 3.092623233795166,
+      "learning_rate": 2.8256164738911977e-05,
+      "loss": 1.0548,
+      "step": 30655
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 2.316596031188965,
+      "learning_rate": 2.8250339485579248e-05,
+      "loss": 1.1654,
+      "step": 30660
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 5.095229148864746,
+      "learning_rate": 2.824451405273364e-05,
+      "loss": 1.1143,
+      "step": 30665
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 2.3152172565460205,
+      "learning_rate": 2.8238688440696885e-05,
+      "loss": 1.0509,
+      "step": 30670
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 2.434166669845581,
+      "learning_rate": 2.8232862649790742e-05,
+      "loss": 0.9805,
+      "step": 30675
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 2.4533190727233887,
+      "learning_rate": 2.8227036680336956e-05,
+      "loss": 0.964,
+      "step": 30680
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 3.624160051345825,
+      "learning_rate": 2.8221210532657283e-05,
+      "loss": 1.0408,
+      "step": 30685
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 4.148279666900635,
+      "learning_rate": 2.8215384207073515e-05,
+      "loss": 1.1267,
+      "step": 30690
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 3.3725688457489014,
+      "learning_rate": 2.820955770390741e-05,
+      "loss": 1.0548,
+      "step": 30695
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 1.4353331327438354,
+      "learning_rate": 2.820373102348076e-05,
+      "loss": 0.8967,
+      "step": 30700
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 1.5979543924331665,
+      "learning_rate": 2.8197904166115386e-05,
+      "loss": 1.2111,
+      "step": 30705
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 4.696241855621338,
+      "learning_rate": 2.819207713213309e-05,
+      "loss": 1.0566,
+      "step": 30710
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 2.242954730987549,
+      "learning_rate": 2.8186249921855683e-05,
+      "loss": 1.1283,
+      "step": 30715
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 1.5376769304275513,
+      "learning_rate": 2.818042253560501e-05,
+      "loss": 1.1125,
+      "step": 30720
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 3.4272782802581787,
+      "learning_rate": 2.8174594973702906e-05,
+      "loss": 1.1207,
+      "step": 30725
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 1.447213053703308,
+      "learning_rate": 2.8168767236471223e-05,
+      "loss": 1.2412,
+      "step": 30730
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 1.9952369928359985,
+      "learning_rate": 2.816293932423182e-05,
+      "loss": 1.0659,
+      "step": 30735
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 3.7000386714935303,
+      "learning_rate": 2.8157111237306562e-05,
+      "loss": 1.0848,
+      "step": 30740
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 0.9547794461250305,
+      "learning_rate": 2.8151282976017336e-05,
+      "loss": 1.0252,
+      "step": 30745
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 2.513737440109253,
+      "learning_rate": 2.8145454540686024e-05,
+      "loss": 0.9642,
+      "step": 30750
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 5.190265655517578,
+      "learning_rate": 2.813962593163453e-05,
+      "loss": 1.0277,
+      "step": 30755
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 2.4562807083129883,
+      "learning_rate": 2.8133797149184755e-05,
+      "loss": 1.1477,
+      "step": 30760
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 1.605533242225647,
+      "learning_rate": 2.812796819365862e-05,
+      "loss": 1.0803,
+      "step": 30765
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 1.3287237882614136,
+      "learning_rate": 2.812213906537806e-05,
+      "loss": 1.1856,
+      "step": 30770
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 2.1841320991516113,
+      "learning_rate": 2.8116309764665e-05,
+      "loss": 1.272,
+      "step": 30775
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 2.0276877880096436,
+      "learning_rate": 2.811048029184139e-05,
+      "loss": 1.0535,
+      "step": 30780
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 2.9414737224578857,
+      "learning_rate": 2.810465064722919e-05,
+      "loss": 1.0341,
+      "step": 30785
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 1.2586784362792969,
+      "learning_rate": 2.8098820831150362e-05,
+      "loss": 1.1424,
+      "step": 30790
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 4.782675743103027,
+      "learning_rate": 2.809299084392688e-05,
+      "loss": 1.0249,
+      "step": 30795
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 1.6296465396881104,
+      "learning_rate": 2.8087160685880726e-05,
+      "loss": 1.0696,
+      "step": 30800
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 1.8938437700271606,
+      "learning_rate": 2.808133035733389e-05,
+      "loss": 1.0324,
+      "step": 30805
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 3.3237104415893555,
+      "learning_rate": 2.807549985860839e-05,
+      "loss": 1.0725,
+      "step": 30810
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 1.7440217733383179,
+      "learning_rate": 2.8069669190026233e-05,
+      "loss": 1.0946,
+      "step": 30815
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 1.855573058128357,
+      "learning_rate": 2.8063838351909434e-05,
+      "loss": 1.1026,
+      "step": 30820
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 1.5629770755767822,
+      "learning_rate": 2.8058007344580027e-05,
+      "loss": 1.3132,
+      "step": 30825
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 4.092102527618408,
+      "learning_rate": 2.805217616836005e-05,
+      "loss": 1.1355,
+      "step": 30830
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 2.19754695892334,
+      "learning_rate": 2.8046344823571557e-05,
+      "loss": 1.0562,
+      "step": 30835
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 2.126832962036133,
+      "learning_rate": 2.804051331053661e-05,
+      "loss": 1.169,
+      "step": 30840
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 3.9059386253356934,
+      "learning_rate": 2.803468162957727e-05,
+      "loss": 1.0877,
+      "step": 30845
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 3.600858449935913,
+      "learning_rate": 2.8028849781015615e-05,
+      "loss": 1.0115,
+      "step": 30850
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 3.020725965499878,
+      "learning_rate": 2.802301776517374e-05,
+      "loss": 0.9752,
+      "step": 30855
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 2.0996792316436768,
+      "learning_rate": 2.801718558237374e-05,
+      "loss": 1.253,
+      "step": 30860
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 1.7010042667388916,
+      "learning_rate": 2.801135323293771e-05,
+      "loss": 1.0858,
+      "step": 30865
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 2.4270615577697754,
+      "learning_rate": 2.8005520717187773e-05,
+      "loss": 0.7829,
+      "step": 30870
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 3.7250173091888428,
+      "learning_rate": 2.7999688035446048e-05,
+      "loss": 1.1342,
+      "step": 30875
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 2.1018354892730713,
+      "learning_rate": 2.7993855188034674e-05,
+      "loss": 1.086,
+      "step": 30880
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 1.520485758781433,
+      "learning_rate": 2.7988022175275786e-05,
+      "loss": 1.0099,
+      "step": 30885
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 2.133230447769165,
+      "learning_rate": 2.7982188997491544e-05,
+      "loss": 1.1996,
+      "step": 30890
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 3.0092363357543945,
+      "learning_rate": 2.7976355655004096e-05,
+      "loss": 1.1812,
+      "step": 30895
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 4.08493185043335,
+      "learning_rate": 2.7970522148135626e-05,
+      "loss": 1.1021,
+      "step": 30900
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 2.9970130920410156,
+      "learning_rate": 2.79646884772083e-05,
+      "loss": 1.0735,
+      "step": 30905
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 3.3271689414978027,
+      "learning_rate": 2.795885464254431e-05,
+      "loss": 1.053,
+      "step": 30910
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 1.4670406579971313,
+      "learning_rate": 2.7953020644465865e-05,
+      "loss": 1.0696,
+      "step": 30915
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 5.897749900817871,
+      "learning_rate": 2.7947186483295157e-05,
+      "loss": 1.0656,
+      "step": 30920
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 2.2055165767669678,
+      "learning_rate": 2.7941352159354394e-05,
+      "loss": 0.9848,
+      "step": 30925
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 2.664703845977783,
+      "learning_rate": 2.7935517672965817e-05,
+      "loss": 1.071,
+      "step": 30930
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 5.1785454750061035,
+      "learning_rate": 2.792968302445164e-05,
+      "loss": 1.1283,
+      "step": 30935
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 4.427746772766113,
+      "learning_rate": 2.7923848214134123e-05,
+      "loss": 1.0444,
+      "step": 30940
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 2.530972957611084,
+      "learning_rate": 2.7918013242335504e-05,
+      "loss": 1.0145,
+      "step": 30945
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 16.054180145263672,
+      "learning_rate": 2.7912178109378056e-05,
+      "loss": 1.2892,
+      "step": 30950
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 2.498807430267334,
+      "learning_rate": 2.790634281558403e-05,
+      "loss": 0.9976,
+      "step": 30955
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 1.9145476818084717,
+      "learning_rate": 2.7900507361275714e-05,
+      "loss": 0.938,
+      "step": 30960
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 1.1175459623336792,
+      "learning_rate": 2.7894671746775386e-05,
+      "loss": 1.1884,
+      "step": 30965
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 4.308173656463623,
+      "learning_rate": 2.7888835972405352e-05,
+      "loss": 1.1149,
+      "step": 30970
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 2.5797321796417236,
+      "learning_rate": 2.7883000038487904e-05,
+      "loss": 1.0525,
+      "step": 30975
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 2.3308804035186768,
+      "learning_rate": 2.7877163945345368e-05,
+      "loss": 1.168,
+      "step": 30980
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 2.186554193496704,
+      "learning_rate": 2.7871327693300054e-05,
+      "loss": 1.0804,
+      "step": 30985
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 7.420773029327393,
+      "learning_rate": 2.7865491282674293e-05,
+      "loss": 1.021,
+      "step": 30990
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 3.671011447906494,
+      "learning_rate": 2.7859654713790434e-05,
+      "loss": 1.0881,
+      "step": 30995
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 1.4102942943572998,
+      "learning_rate": 2.7853817986970814e-05,
+      "loss": 1.1006,
+      "step": 31000
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 4.502505779266357,
+      "learning_rate": 2.784798110253779e-05,
+      "loss": 1.0186,
+      "step": 31005
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 2.1751863956451416,
+      "learning_rate": 2.7842144060813736e-05,
+      "loss": 0.9848,
+      "step": 31010
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 2.5720980167388916,
+      "learning_rate": 2.783630686212102e-05,
+      "loss": 1.0289,
+      "step": 31015
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 1.7621455192565918,
+      "learning_rate": 2.783046950678202e-05,
+      "loss": 0.89,
+      "step": 31020
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 1.859261155128479,
+      "learning_rate": 2.7824631995119134e-05,
+      "loss": 1.1063,
+      "step": 31025
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 2.331822156906128,
+      "learning_rate": 2.7818794327454757e-05,
+      "loss": 1.0846,
+      "step": 31030
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 5.336060523986816,
+      "learning_rate": 2.78129565041113e-05,
+      "loss": 0.9446,
+      "step": 31035
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 1.7737152576446533,
+      "learning_rate": 2.7807118525411176e-05,
+      "loss": 1.0767,
+      "step": 31040
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 10.145687103271484,
+      "learning_rate": 2.780128039167682e-05,
+      "loss": 1.118,
+      "step": 31045
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 2.768529176712036,
+      "learning_rate": 2.7795442103230656e-05,
+      "loss": 1.0003,
+      "step": 31050
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 2.1712610721588135,
+      "learning_rate": 2.778960366039513e-05,
+      "loss": 1.0129,
+      "step": 31055
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 2.940528154373169,
+      "learning_rate": 2.7783765063492696e-05,
+      "loss": 1.077,
+      "step": 31060
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 2.5205068588256836,
+      "learning_rate": 2.777792631284581e-05,
+      "loss": 1.2037,
+      "step": 31065
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 1.8787987232208252,
+      "learning_rate": 2.7772087408776937e-05,
+      "loss": 1.1164,
+      "step": 31070
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 1.8113431930541992,
+      "learning_rate": 2.776624835160856e-05,
+      "loss": 1.1057,
+      "step": 31075
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 1.184775710105896,
+      "learning_rate": 2.7760409141663164e-05,
+      "loss": 1.1547,
+      "step": 31080
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 2.488152265548706,
+      "learning_rate": 2.775456977926324e-05,
+      "loss": 1.1285,
+      "step": 31085
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 1.7316298484802246,
+      "learning_rate": 2.7748730264731287e-05,
+      "loss": 1.0315,
+      "step": 31090
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 1.8663023710250854,
+      "learning_rate": 2.774289059838982e-05,
+      "loss": 1.0133,
+      "step": 31095
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 1.8005400896072388,
+      "learning_rate": 2.7737050780561358e-05,
+      "loss": 1.0123,
+      "step": 31100
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 2.0711119174957275,
+      "learning_rate": 2.7731210811568427e-05,
+      "loss": 1.0181,
+      "step": 31105
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 4.488975524902344,
+      "learning_rate": 2.7725370691733565e-05,
+      "loss": 0.9204,
+      "step": 31110
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 1.9340286254882812,
+      "learning_rate": 2.7719530421379308e-05,
+      "loss": 0.9195,
+      "step": 31115
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 1.3298678398132324,
+      "learning_rate": 2.7713690000828213e-05,
+      "loss": 1.09,
+      "step": 31120
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 1.4798712730407715,
+      "learning_rate": 2.7707849430402838e-05,
+      "loss": 1.0828,
+      "step": 31125
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 2.92453670501709,
+      "learning_rate": 2.770200871042576e-05,
+      "loss": 1.1759,
+      "step": 31130
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 2.3763322830200195,
+      "learning_rate": 2.769616784121954e-05,
+      "loss": 1.04,
+      "step": 31135
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 1.7251625061035156,
+      "learning_rate": 2.7690326823106776e-05,
+      "loss": 1.0022,
+      "step": 31140
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 2.365755796432495,
+      "learning_rate": 2.768448565641007e-05,
+      "loss": 1.0852,
+      "step": 31145
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 2.628757953643799,
+      "learning_rate": 2.7678644341451998e-05,
+      "loss": 1.0231,
+      "step": 31150
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 2.307605743408203,
+      "learning_rate": 2.7672802878555186e-05,
+      "loss": 1.1832,
+      "step": 31155
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 4.4019951820373535,
+      "learning_rate": 2.7666961268042253e-05,
+      "loss": 1.0704,
+      "step": 31160
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 1.9276536703109741,
+      "learning_rate": 2.7661119510235816e-05,
+      "loss": 1.1429,
+      "step": 31165
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 1.0725014209747314,
+      "learning_rate": 2.7655277605458507e-05,
+      "loss": 0.9747,
+      "step": 31170
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 2.2577459812164307,
+      "learning_rate": 2.7649435554032994e-05,
+      "loss": 1.1215,
+      "step": 31175
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 1.3800833225250244,
+      "learning_rate": 2.76435933562819e-05,
+      "loss": 1.2958,
+      "step": 31180
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 1.0564745664596558,
+      "learning_rate": 2.763775101252789e-05,
+      "loss": 1.0229,
+      "step": 31185
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 1.7594860792160034,
+      "learning_rate": 2.763190852309364e-05,
+      "loss": 1.041,
+      "step": 31190
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 1.3283685445785522,
+      "learning_rate": 2.7626065888301816e-05,
+      "loss": 1.1612,
+      "step": 31195
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 2.163010597229004,
+      "learning_rate": 2.76202231084751e-05,
+      "loss": 1.0172,
+      "step": 31200
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 0.8803406953811646,
+      "learning_rate": 2.761438018393619e-05,
+      "loss": 1.0381,
+      "step": 31205
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 2.269392728805542,
+      "learning_rate": 2.7608537115007775e-05,
+      "loss": 1.0224,
+      "step": 31210
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 3.6767170429229736,
+      "learning_rate": 2.7602693902012572e-05,
+      "loss": 1.1205,
+      "step": 31215
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 2.087529182434082,
+      "learning_rate": 2.7596850545273286e-05,
+      "loss": 1.0549,
+      "step": 31220
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 2.73809552192688,
+      "learning_rate": 2.7591007045112642e-05,
+      "loss": 1.0792,
+      "step": 31225
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 4.561300277709961,
+      "learning_rate": 2.7585163401853376e-05,
+      "loss": 1.1238,
+      "step": 31230
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 3.266984462738037,
+      "learning_rate": 2.7579319615818215e-05,
+      "loss": 1.0098,
+      "step": 31235
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 2.625706434249878,
+      "learning_rate": 2.757347568732993e-05,
+      "loss": 0.9742,
+      "step": 31240
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 1.4844180345535278,
+      "learning_rate": 2.7567631616711243e-05,
+      "loss": 1.1356,
+      "step": 31245
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 1.707905888557434,
+      "learning_rate": 2.7561787404284928e-05,
+      "loss": 0.9465,
+      "step": 31250
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 1.6472175121307373,
+      "learning_rate": 2.755594305037376e-05,
+      "loss": 1.1298,
+      "step": 31255
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 4.289102554321289,
+      "learning_rate": 2.755009855530052e-05,
+      "loss": 1.0665,
+      "step": 31260
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 1.6025971174240112,
+      "learning_rate": 2.7544253919387987e-05,
+      "loss": 0.968,
+      "step": 31265
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 3.026954412460327,
+      "learning_rate": 2.7538409142958953e-05,
+      "loss": 0.775,
+      "step": 31270
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 3.0979301929473877,
+      "learning_rate": 2.7532564226336222e-05,
+      "loss": 1.1642,
+      "step": 31275
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 3.6659834384918213,
+      "learning_rate": 2.7526719169842602e-05,
+      "loss": 1.1043,
+      "step": 31280
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 2.309248208999634,
+      "learning_rate": 2.7520873973800903e-05,
+      "loss": 1.0931,
+      "step": 31285
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 2.78249192237854,
+      "learning_rate": 2.7515028638533958e-05,
+      "loss": 1.0622,
+      "step": 31290
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 1.3780916929244995,
+      "learning_rate": 2.7509183164364593e-05,
+      "loss": 1.0036,
+      "step": 31295
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 2.205402135848999,
+      "learning_rate": 2.7503337551615654e-05,
+      "loss": 1.3497,
+      "step": 31300
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 3.5160036087036133,
+      "learning_rate": 2.7497491800609986e-05,
+      "loss": 1.2502,
+      "step": 31305
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 1.4284930229187012,
+      "learning_rate": 2.7491645911670437e-05,
+      "loss": 0.9721,
+      "step": 31310
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 5.789677143096924,
+      "learning_rate": 2.748579988511988e-05,
+      "loss": 1.1206,
+      "step": 31315
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 3.2626028060913086,
+      "learning_rate": 2.747995372128117e-05,
+      "loss": 1.0979,
+      "step": 31320
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 3.9523472785949707,
+      "learning_rate": 2.7474107420477195e-05,
+      "loss": 1.0172,
+      "step": 31325
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 3.8621017932891846,
+      "learning_rate": 2.7468260983030846e-05,
+      "loss": 1.1448,
+      "step": 31330
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 2.3181259632110596,
+      "learning_rate": 2.7462414409265003e-05,
+      "loss": 1.0566,
+      "step": 31335
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 1.018276572227478,
+      "learning_rate": 2.7456567699502573e-05,
+      "loss": 1.1421,
+      "step": 31340
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 2.900003671646118,
+      "learning_rate": 2.7450720854066464e-05,
+      "loss": 1.0946,
+      "step": 31345
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 1.6421140432357788,
+      "learning_rate": 2.7444873873279586e-05,
+      "loss": 1.1454,
+      "step": 31350
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 3.2035090923309326,
+      "learning_rate": 2.743902675746487e-05,
+      "loss": 1.1674,
+      "step": 31355
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 2.1225266456604004,
+      "learning_rate": 2.743317950694524e-05,
+      "loss": 1.0314,
+      "step": 31360
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 1.5846216678619385,
+      "learning_rate": 2.742733212204363e-05,
+      "loss": 0.9834,
+      "step": 31365
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 2.3945157527923584,
+      "learning_rate": 2.7421484603083004e-05,
+      "loss": 1.1503,
+      "step": 31370
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 1.944319725036621,
+      "learning_rate": 2.7415636950386285e-05,
+      "loss": 1.1397,
+      "step": 31375
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 3.456681728363037,
+      "learning_rate": 2.7409789164276456e-05,
+      "loss": 1.0885,
+      "step": 31380
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 2.284849166870117,
+      "learning_rate": 2.740394124507647e-05,
+      "loss": 1.0706,
+      "step": 31385
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 3.1083555221557617,
+      "learning_rate": 2.7398093193109314e-05,
+      "loss": 1.2338,
+      "step": 31390
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 2.024843454360962,
+      "learning_rate": 2.739224500869796e-05,
+      "loss": 1.2683,
+      "step": 31395
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 3.8661715984344482,
+      "learning_rate": 2.7386396692165406e-05,
+      "loss": 1.213,
+      "step": 31400
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 2.7120206356048584,
+      "learning_rate": 2.738054824383464e-05,
+      "loss": 1.0801,
+      "step": 31405
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 1.732399582862854,
+      "learning_rate": 2.7374699664028668e-05,
+      "loss": 1.2241,
+      "step": 31410
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 2.247511386871338,
+      "learning_rate": 2.7368850953070503e-05,
+      "loss": 1.0742,
+      "step": 31415
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 1.6583189964294434,
+      "learning_rate": 2.736300211128316e-05,
+      "loss": 1.0932,
+      "step": 31420
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 3.329725742340088,
+      "learning_rate": 2.7357153138989668e-05,
+      "loss": 1.0847,
+      "step": 31425
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 4.0160956382751465,
+      "learning_rate": 2.735130403651306e-05,
+      "loss": 1.026,
+      "step": 31430
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 4.307696342468262,
+      "learning_rate": 2.734545480417637e-05,
+      "loss": 0.9154,
+      "step": 31435
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 1.9788671731948853,
+      "learning_rate": 2.733960544230265e-05,
+      "loss": 1.0141,
+      "step": 31440
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 10.733818054199219,
+      "learning_rate": 2.7333755951214957e-05,
+      "loss": 0.9602,
+      "step": 31445
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 4.034037113189697,
+      "learning_rate": 2.732790633123634e-05,
+      "loss": 0.9853,
+      "step": 31450
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 4.451757431030273,
+      "learning_rate": 2.7322056582689885e-05,
+      "loss": 1.1026,
+      "step": 31455
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 5.899003028869629,
+      "learning_rate": 2.7316206705898655e-05,
+      "loss": 1.2884,
+      "step": 31460
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 2.728759765625,
+      "learning_rate": 2.7310356701185747e-05,
+      "loss": 1.0355,
+      "step": 31465
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 1.9716694355010986,
+      "learning_rate": 2.7304506568874228e-05,
+      "loss": 1.1145,
+      "step": 31470
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 2.966343641281128,
+      "learning_rate": 2.7298656309287206e-05,
+      "loss": 1.2393,
+      "step": 31475
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 3.9974021911621094,
+      "learning_rate": 2.7292805922747787e-05,
+      "loss": 1.08,
+      "step": 31480
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 2.1800546646118164,
+      "learning_rate": 2.7286955409579084e-05,
+      "loss": 1.0767,
+      "step": 31485
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 7.327594757080078,
+      "learning_rate": 2.7281104770104205e-05,
+      "loss": 1.0415,
+      "step": 31490
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 1.3344082832336426,
+      "learning_rate": 2.7275254004646284e-05,
+      "loss": 1.059,
+      "step": 31495
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 3.478764772415161,
+      "learning_rate": 2.726940311352845e-05,
+      "loss": 1.1981,
+      "step": 31500
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 0.9754401445388794,
+      "learning_rate": 2.726355209707384e-05,
+      "loss": 1.0865,
+      "step": 31505
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 2.373671531677246,
+      "learning_rate": 2.72577009556056e-05,
+      "loss": 0.9644,
+      "step": 31510
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 2.6133816242218018,
+      "learning_rate": 2.7251849689446885e-05,
+      "loss": 1.0806,
+      "step": 31515
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 3.878574848175049,
+      "learning_rate": 2.724599829892085e-05,
+      "loss": 1.1225,
+      "step": 31520
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 3.734332799911499,
+      "learning_rate": 2.724014678435067e-05,
+      "loss": 1.121,
+      "step": 31525
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 8.514619827270508,
+      "learning_rate": 2.7234295146059503e-05,
+      "loss": 1.0401,
+      "step": 31530
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 1.05905282497406,
+      "learning_rate": 2.722844338437054e-05,
+      "loss": 0.9848,
+      "step": 31535
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 1.8391199111938477,
+      "learning_rate": 2.7222591499606966e-05,
+      "loss": 1.0086,
+      "step": 31540
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 2.663299560546875,
+      "learning_rate": 2.721673949209197e-05,
+      "loss": 1.134,
+      "step": 31545
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 2.3702192306518555,
+      "learning_rate": 2.7210887362148755e-05,
+      "loss": 1.1833,
+      "step": 31550
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 3.5295441150665283,
+      "learning_rate": 2.7205035110100534e-05,
+      "loss": 1.0104,
+      "step": 31555
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 6.249225616455078,
+      "learning_rate": 2.7199182736270524e-05,
+      "loss": 1.0269,
+      "step": 31560
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 3.3529887199401855,
+      "learning_rate": 2.719333024098193e-05,
+      "loss": 1.0667,
+      "step": 31565
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 2.5261456966400146,
+      "learning_rate": 2.7187477624557982e-05,
+      "loss": 0.9635,
+      "step": 31570
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 3.518795967102051,
+      "learning_rate": 2.718162488732192e-05,
+      "loss": 1.1624,
+      "step": 31575
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 8.467171669006348,
+      "learning_rate": 2.7175772029596986e-05,
+      "loss": 1.1869,
+      "step": 31580
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 3.2780685424804688,
+      "learning_rate": 2.7169919051706422e-05,
+      "loss": 1.0429,
+      "step": 31585
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 4.071305751800537,
+      "learning_rate": 2.716406595397349e-05,
+      "loss": 1.252,
+      "step": 31590
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 2.072431802749634,
+      "learning_rate": 2.7158212736721444e-05,
+      "loss": 1.0693,
+      "step": 31595
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 6.514179229736328,
+      "learning_rate": 2.7152359400273546e-05,
+      "loss": 1.147,
+      "step": 31600
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 3.2777328491210938,
+      "learning_rate": 2.7146505944953075e-05,
+      "loss": 0.9809,
+      "step": 31605
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 2.1688232421875,
+      "learning_rate": 2.714065237108332e-05,
+      "loss": 1.2513,
+      "step": 31610
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 3.6384012699127197,
+      "learning_rate": 2.7134798678987544e-05,
+      "loss": 1.1978,
+      "step": 31615
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 1.3136318922042847,
+      "learning_rate": 2.712894486898907e-05,
+      "loss": 1.0106,
+      "step": 31620
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 1.3264063596725464,
+      "learning_rate": 2.7123090941411185e-05,
+      "loss": 1.1233,
+      "step": 31625
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 1.84585440158844,
+      "learning_rate": 2.7117236896577185e-05,
+      "loss": 1.0608,
+      "step": 31630
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 2.1590003967285156,
+      "learning_rate": 2.71113827348104e-05,
+      "loss": 1.1022,
+      "step": 31635
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 5.31462287902832,
+      "learning_rate": 2.7105528456434136e-05,
+      "loss": 1.0598,
+      "step": 31640
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 2.746979236602783,
+      "learning_rate": 2.7099674061771724e-05,
+      "loss": 1.116,
+      "step": 31645
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 2.339494466781616,
+      "learning_rate": 2.70938195511465e-05,
+      "loss": 1.0503,
+      "step": 31650
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 10.379657745361328,
+      "learning_rate": 2.708796492488179e-05,
+      "loss": 1.2657,
+      "step": 31655
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 1.8636327981948853,
+      "learning_rate": 2.7082110183300962e-05,
+      "loss": 1.0569,
+      "step": 31660
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 3.62514328956604,
+      "learning_rate": 2.7076255326727344e-05,
+      "loss": 1.1566,
+      "step": 31665
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 1.6103851795196533,
+      "learning_rate": 2.7070400355484306e-05,
+      "loss": 1.2291,
+      "step": 31670
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 38.40741729736328,
+      "learning_rate": 2.70645452698952e-05,
+      "loss": 1.0219,
+      "step": 31675
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 1.3236831426620483,
+      "learning_rate": 2.705869007028341e-05,
+      "loss": 1.1879,
+      "step": 31680
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 2.328878402709961,
+      "learning_rate": 2.7052834756972307e-05,
+      "loss": 0.949,
+      "step": 31685
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 2.4937305450439453,
+      "learning_rate": 2.704697933028528e-05,
+      "loss": 1.2623,
+      "step": 31690
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 1.3484976291656494,
+      "learning_rate": 2.7041123790545703e-05,
+      "loss": 0.9774,
+      "step": 31695
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 2.5267484188079834,
+      "learning_rate": 2.703526813807698e-05,
+      "loss": 1.0935,
+      "step": 31700
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 1.9129496812820435,
+      "learning_rate": 2.702941237320252e-05,
+      "loss": 1.0674,
+      "step": 31705
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 2.022209882736206,
+      "learning_rate": 2.702355649624572e-05,
+      "loss": 1.1119,
+      "step": 31710
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 5.069640159606934,
+      "learning_rate": 2.7017700507529996e-05,
+      "loss": 1.1296,
+      "step": 31715
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 3.870335102081299,
+      "learning_rate": 2.7011844407378776e-05,
+      "loss": 1.2017,
+      "step": 31720
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 1.1030346155166626,
+      "learning_rate": 2.7005988196115482e-05,
+      "loss": 1.1276,
+      "step": 31725
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 1.269315481185913,
+      "learning_rate": 2.7000131874063545e-05,
+      "loss": 1.1485,
+      "step": 31730
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 6.841867923736572,
+      "learning_rate": 2.69942754415464e-05,
+      "loss": 0.9942,
+      "step": 31735
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 2.453371524810791,
+      "learning_rate": 2.6988418898887498e-05,
+      "loss": 1.0962,
+      "step": 31740
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 4.395910263061523,
+      "learning_rate": 2.698256224641028e-05,
+      "loss": 1.0983,
+      "step": 31745
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 3.848184108734131,
+      "learning_rate": 2.6976705484438213e-05,
+      "loss": 1.0488,
+      "step": 31750
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 3.335904121398926,
+      "learning_rate": 2.6970848613294765e-05,
+      "loss": 1.2326,
+      "step": 31755
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 6.4267706871032715,
+      "learning_rate": 2.696499163330339e-05,
+      "loss": 1.0299,
+      "step": 31760
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 2.3225696086883545,
+      "learning_rate": 2.6959134544787567e-05,
+      "loss": 1.1295,
+      "step": 31765
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 4.162220478057861,
+      "learning_rate": 2.6953277348070783e-05,
+      "loss": 1.0728,
+      "step": 31770
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 5.359671592712402,
+      "learning_rate": 2.6947420043476524e-05,
+      "loss": 1.186,
+      "step": 31775
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 2.6894781589508057,
+      "learning_rate": 2.6941562631328278e-05,
+      "loss": 0.9445,
+      "step": 31780
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 30.432933807373047,
+      "learning_rate": 2.693570511194954e-05,
+      "loss": 1.1887,
+      "step": 31785
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 4.67871618270874,
+      "learning_rate": 2.6929847485663823e-05,
+      "loss": 0.9568,
+      "step": 31790
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 5.078667163848877,
+      "learning_rate": 2.6923989752794638e-05,
+      "loss": 1.0335,
+      "step": 31795
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 3.381072521209717,
+      "learning_rate": 2.69181319136655e-05,
+      "loss": 1.1649,
+      "step": 31800
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 3.0394537448883057,
+      "learning_rate": 2.6912273968599928e-05,
+      "loss": 1.1632,
+      "step": 31805
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 3.1796021461486816,
+      "learning_rate": 2.690641591792145e-05,
+      "loss": 1.2188,
+      "step": 31810
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 6.784097671508789,
+      "learning_rate": 2.690055776195361e-05,
+      "loss": 1.0647,
+      "step": 31815
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 5.862364768981934,
+      "learning_rate": 2.6894699501019937e-05,
+      "loss": 1.034,
+      "step": 31820
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 1.8071324825286865,
+      "learning_rate": 2.688884113544398e-05,
+      "loss": 1.0983,
+      "step": 31825
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 3.1565282344818115,
+      "learning_rate": 2.6882982665549288e-05,
+      "loss": 1.1485,
+      "step": 31830
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 2.5325186252593994,
+      "learning_rate": 2.6877124091659424e-05,
+      "loss": 1.1328,
+      "step": 31835
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 1.6840946674346924,
+      "learning_rate": 2.6871265414097947e-05,
+      "loss": 0.9854,
+      "step": 31840
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 1.4422705173492432,
+      "learning_rate": 2.6865406633188423e-05,
+      "loss": 1.0145,
+      "step": 31845
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 2.4371087551116943,
+      "learning_rate": 2.6859547749254433e-05,
+      "loss": 1.1726,
+      "step": 31850
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 3.1019036769866943,
+      "learning_rate": 2.6853688762619555e-05,
+      "loss": 1.1436,
+      "step": 31855
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 3.6679959297180176,
+      "learning_rate": 2.6847829673607373e-05,
+      "loss": 1.0577,
+      "step": 31860
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 3.406115770339966,
+      "learning_rate": 2.6841970482541473e-05,
+      "loss": 1.1801,
+      "step": 31865
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 2.184192180633545,
+      "learning_rate": 2.6836111189745462e-05,
+      "loss": 1.1725,
+      "step": 31870
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 2.7702200412750244,
+      "learning_rate": 2.6830251795542938e-05,
+      "loss": 1.061,
+      "step": 31875
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 1.0884158611297607,
+      "learning_rate": 2.6824392300257505e-05,
+      "loss": 0.953,
+      "step": 31880
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 1.4687933921813965,
+      "learning_rate": 2.681853270421279e-05,
+      "loss": 1.1745,
+      "step": 31885
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 1.536842703819275,
+      "learning_rate": 2.68126730077324e-05,
+      "loss": 0.9724,
+      "step": 31890
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 3.3644862174987793,
+      "learning_rate": 2.6806813211139965e-05,
+      "loss": 1.1599,
+      "step": 31895
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 2.4303081035614014,
+      "learning_rate": 2.6800953314759108e-05,
+      "loss": 1.2175,
+      "step": 31900
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 4.207813262939453,
+      "learning_rate": 2.679509331891348e-05,
+      "loss": 1.1489,
+      "step": 31905
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 4.013348579406738,
+      "learning_rate": 2.6789233223926713e-05,
+      "loss": 1.1652,
+      "step": 31910
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 4.057195663452148,
+      "learning_rate": 2.6783373030122455e-05,
+      "loss": 1.1245,
+      "step": 31915
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 1.9660584926605225,
+      "learning_rate": 2.6777512737824358e-05,
+      "loss": 1.2141,
+      "step": 31920
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 1.8502095937728882,
+      "learning_rate": 2.6771652347356074e-05,
+      "loss": 0.9919,
+      "step": 31925
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 1.4505739212036133,
+      "learning_rate": 2.6765791859041278e-05,
+      "loss": 1.0553,
+      "step": 31930
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 3.0813705921173096,
+      "learning_rate": 2.6759931273203632e-05,
+      "loss": 0.9613,
+      "step": 31935
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 1.6187719106674194,
+      "learning_rate": 2.6754070590166808e-05,
+      "loss": 1.1784,
+      "step": 31940
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 5.102146148681641,
+      "learning_rate": 2.6748209810254493e-05,
+      "loss": 1.1304,
+      "step": 31945
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 2.7340807914733887,
+      "learning_rate": 2.674234893379037e-05,
+      "loss": 1.0532,
+      "step": 31950
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 3.3203606605529785,
+      "learning_rate": 2.6736487961098122e-05,
+      "loss": 1.0749,
+      "step": 31955
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 2.275956153869629,
+      "learning_rate": 2.6730626892501448e-05,
+      "loss": 1.0954,
+      "step": 31960
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 1.4483486413955688,
+      "learning_rate": 2.6724765728324054e-05,
+      "loss": 1.1644,
+      "step": 31965
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 2.504141330718994,
+      "learning_rate": 2.6718904468889633e-05,
+      "loss": 1.299,
+      "step": 31970
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 2.5018177032470703,
+      "learning_rate": 2.671304311452191e-05,
+      "loss": 1.1193,
+      "step": 31975
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 3.87642502784729,
+      "learning_rate": 2.6707181665544594e-05,
+      "loss": 1.0883,
+      "step": 31980
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 4.660168647766113,
+      "learning_rate": 2.6701320122281416e-05,
+      "loss": 1.0395,
+      "step": 31985
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 2.28615665435791,
+      "learning_rate": 2.669545848505609e-05,
+      "loss": 1.1664,
+      "step": 31990
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 4.775680065155029,
+      "learning_rate": 2.6689596754192355e-05,
+      "loss": 1.2104,
+      "step": 31995
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 1.285249948501587,
+      "learning_rate": 2.668373493001395e-05,
+      "loss": 1.1832,
+      "step": 32000
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 4.368350982666016,
+      "learning_rate": 2.667787301284461e-05,
+      "loss": 1.1155,
+      "step": 32005
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 1.7169698476791382,
+      "learning_rate": 2.667201100300809e-05,
+      "loss": 0.9742,
+      "step": 32010
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 2.3918094635009766,
+      "learning_rate": 2.666614890082815e-05,
+      "loss": 1.067,
+      "step": 32015
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 3.565277576446533,
+      "learning_rate": 2.666028670662853e-05,
+      "loss": 1.0911,
+      "step": 32020
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 4.275758743286133,
+      "learning_rate": 2.6654424420732997e-05,
+      "loss": 0.998,
+      "step": 32025
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 2.591794490814209,
+      "learning_rate": 2.6648562043465323e-05,
+      "loss": 1.111,
+      "step": 32030
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 5.0382561683654785,
+      "learning_rate": 2.664269957514929e-05,
+      "loss": 1.1469,
+      "step": 32035
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 2.2757861614227295,
+      "learning_rate": 2.6636837016108656e-05,
+      "loss": 1.0449,
+      "step": 32040
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 2.4247963428497314,
+      "learning_rate": 2.6630974366667226e-05,
+      "loss": 1.166,
+      "step": 32045
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 1.7559020519256592,
+      "learning_rate": 2.6625111627148768e-05,
+      "loss": 1.1379,
+      "step": 32050
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 2.4389190673828125,
+      "learning_rate": 2.661924879787709e-05,
+      "loss": 1.0711,
+      "step": 32055
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 2.227290630340576,
+      "learning_rate": 2.661338587917597e-05,
+      "loss": 1.0021,
+      "step": 32060
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 1.5972774028778076,
+      "learning_rate": 2.660752287136924e-05,
+      "loss": 1.0343,
+      "step": 32065
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 7.264094829559326,
+      "learning_rate": 2.6601659774780692e-05,
+      "loss": 1.0336,
+      "step": 32070
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 1.8394643068313599,
+      "learning_rate": 2.6595796589734136e-05,
+      "loss": 1.0561,
+      "step": 32075
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 2.3896703720092773,
+      "learning_rate": 2.6589933316553395e-05,
+      "loss": 1.2167,
+      "step": 32080
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 1.8738449811935425,
+      "learning_rate": 2.6584069955562286e-05,
+      "loss": 1.1215,
+      "step": 32085
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 1.6844857931137085,
+      "learning_rate": 2.657820650708464e-05,
+      "loss": 1.1444,
+      "step": 32090
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 6.457211494445801,
+      "learning_rate": 2.657234297144429e-05,
+      "loss": 0.9628,
+      "step": 32095
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 3.831761121749878,
+      "learning_rate": 2.656647934896508e-05,
+      "loss": 0.8621,
+      "step": 32100
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 2.6082820892333984,
+      "learning_rate": 2.6560615639970833e-05,
+      "loss": 1.1312,
+      "step": 32105
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 1.152781367301941,
+      "learning_rate": 2.6554751844785414e-05,
+      "loss": 1.1762,
+      "step": 32110
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 1.7581003904342651,
+      "learning_rate": 2.654888796373266e-05,
+      "loss": 0.9553,
+      "step": 32115
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 3.2171480655670166,
+      "learning_rate": 2.654302399713644e-05,
+      "loss": 1.0941,
+      "step": 32120
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 1.9642105102539062,
+      "learning_rate": 2.6537159945320606e-05,
+      "loss": 1.1688,
+      "step": 32125
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 9.561652183532715,
+      "learning_rate": 2.6531295808609023e-05,
+      "loss": 0.9819,
+      "step": 32130
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 7.757037162780762,
+      "learning_rate": 2.6525431587325568e-05,
+      "loss": 1.0823,
+      "step": 32135
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 4.364965438842773,
+      "learning_rate": 2.6519567281794105e-05,
+      "loss": 1.008,
+      "step": 32140
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 5.253482818603516,
+      "learning_rate": 2.6513702892338526e-05,
+      "loss": 0.9406,
+      "step": 32145
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 3.2114205360412598,
+      "learning_rate": 2.650783841928271e-05,
+      "loss": 0.9781,
+      "step": 32150
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 5.03977108001709,
+      "learning_rate": 2.6501973862950547e-05,
+      "loss": 1.1507,
+      "step": 32155
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 2.5563600063323975,
+      "learning_rate": 2.6496109223665928e-05,
+      "loss": 1.0864,
+      "step": 32160
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 3.900869369506836,
+      "learning_rate": 2.649024450175275e-05,
+      "loss": 1.0983,
+      "step": 32165
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 1.9279170036315918,
+      "learning_rate": 2.648437969753491e-05,
+      "loss": 1.0866,
+      "step": 32170
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 5.001946926116943,
+      "learning_rate": 2.6478514811336342e-05,
+      "loss": 1.0995,
+      "step": 32175
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 3.2745273113250732,
+      "learning_rate": 2.6472649843480923e-05,
+      "loss": 1.1328,
+      "step": 32180
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 5.206939697265625,
+      "learning_rate": 2.6466784794292588e-05,
+      "loss": 1.1132,
+      "step": 32185
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 2.813488483428955,
+      "learning_rate": 2.6460919664095245e-05,
+      "loss": 1.0479,
+      "step": 32190
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 3.6982460021972656,
+      "learning_rate": 2.6455054453212837e-05,
+      "loss": 1.0327,
+      "step": 32195
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 1.1332085132598877,
+      "learning_rate": 2.644918916196928e-05,
+      "loss": 1.0878,
+      "step": 32200
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 2.1422863006591797,
+      "learning_rate": 2.6443323790688517e-05,
+      "loss": 1.0685,
+      "step": 32205
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 4.674763202667236,
+      "learning_rate": 2.6437458339694483e-05,
+      "loss": 0.8133,
+      "step": 32210
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 1.626625657081604,
+      "learning_rate": 2.6431592809311112e-05,
+      "loss": 1.2528,
+      "step": 32215
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 1.6559752225875854,
+      "learning_rate": 2.642572719986236e-05,
+      "loss": 1.0107,
+      "step": 32220
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 3.3808319568634033,
+      "learning_rate": 2.6419861511672174e-05,
+      "loss": 1.2791,
+      "step": 32225
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 2.20304274559021,
+      "learning_rate": 2.6413995745064513e-05,
+      "loss": 1.0505,
+      "step": 32230
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 3.014941692352295,
+      "learning_rate": 2.6408129900363342e-05,
+      "loss": 1.0684,
+      "step": 32235
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 2.0386364459991455,
+      "learning_rate": 2.6402263977892617e-05,
+      "loss": 0.9374,
+      "step": 32240
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 2.927957057952881,
+      "learning_rate": 2.6396397977976305e-05,
+      "loss": 0.95,
+      "step": 32245
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 2.662858724594116,
+      "learning_rate": 2.639053190093839e-05,
+      "loss": 1.0665,
+      "step": 32250
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 2.204014539718628,
+      "learning_rate": 2.6384665747102842e-05,
+      "loss": 1.0675,
+      "step": 32255
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 3.048424005508423,
+      "learning_rate": 2.6378799516793645e-05,
+      "loss": 1.1372,
+      "step": 32260
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 2.034694194793701,
+      "learning_rate": 2.6372933210334788e-05,
+      "loss": 1.1936,
+      "step": 32265
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 1.9877517223358154,
+      "learning_rate": 2.636706682805026e-05,
+      "loss": 1.129,
+      "step": 32270
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 1.1743559837341309,
+      "learning_rate": 2.636120037026404e-05,
+      "loss": 1.1884,
+      "step": 32275
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 3.967900514602661,
+      "learning_rate": 2.6355333837300144e-05,
+      "loss": 1.1177,
+      "step": 32280
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 2.5938291549682617,
+      "learning_rate": 2.6349467229482566e-05,
+      "loss": 1.1809,
+      "step": 32285
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 3.2662594318389893,
+      "learning_rate": 2.6343600547135318e-05,
+      "loss": 1.0683,
+      "step": 32290
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 2.3414182662963867,
+      "learning_rate": 2.633773379058241e-05,
+      "loss": 1.2112,
+      "step": 32295
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 2.2585291862487793,
+      "learning_rate": 2.633186696014785e-05,
+      "loss": 1.1298,
+      "step": 32300
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 1.8855080604553223,
+      "learning_rate": 2.632600005615567e-05,
+      "loss": 1.0465,
+      "step": 32305
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 1.7714307308197021,
+      "learning_rate": 2.632013307892988e-05,
+      "loss": 1.0813,
+      "step": 32310
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 1.7894978523254395,
+      "learning_rate": 2.6314266028794516e-05,
+      "loss": 1.0873,
+      "step": 32315
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 2.007699489593506,
+      "learning_rate": 2.63083989060736e-05,
+      "loss": 1.0634,
+      "step": 32320
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 4.4481892585754395,
+      "learning_rate": 2.630253171109118e-05,
+      "loss": 1.1221,
+      "step": 32325
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 2.6404707431793213,
+      "learning_rate": 2.6296664444171277e-05,
+      "loss": 1.0629,
+      "step": 32330
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 4.706933975219727,
+      "learning_rate": 2.629079710563795e-05,
+      "loss": 0.9293,
+      "step": 32335
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 3.423985242843628,
+      "learning_rate": 2.628492969581524e-05,
+      "loss": 1.1948,
+      "step": 32340
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 3.1933882236480713,
+      "learning_rate": 2.62790622150272e-05,
+      "loss": 0.9721,
+      "step": 32345
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 2.481731414794922,
+      "learning_rate": 2.6273194663597877e-05,
+      "loss": 1.0354,
+      "step": 32350
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 2.2168052196502686,
+      "learning_rate": 2.626732704185134e-05,
+      "loss": 1.0921,
+      "step": 32355
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 1.8711326122283936,
+      "learning_rate": 2.626145935011165e-05,
+      "loss": 1.2218,
+      "step": 32360
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 5.039557456970215,
+      "learning_rate": 2.6255591588702872e-05,
+      "loss": 1.0336,
+      "step": 32365
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 2.2609169483184814,
+      "learning_rate": 2.624972375794908e-05,
+      "loss": 1.2382,
+      "step": 32370
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 3.6978182792663574,
+      "learning_rate": 2.624385585817434e-05,
+      "loss": 1.16,
+      "step": 32375
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 1.6256968975067139,
+      "learning_rate": 2.623798788970273e-05,
+      "loss": 1.1525,
+      "step": 32380
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 2.7560393810272217,
+      "learning_rate": 2.623211985285834e-05,
+      "loss": 1.0104,
+      "step": 32385
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 2.977550506591797,
+      "learning_rate": 2.6226251747965247e-05,
+      "loss": 1.093,
+      "step": 32390
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 2.2501907348632812,
+      "learning_rate": 2.6220383575347547e-05,
+      "loss": 1.1158,
+      "step": 32395
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 1.943388819694519,
+      "learning_rate": 2.621451533532933e-05,
+      "loss": 1.1113,
+      "step": 32400
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 3.2862977981567383,
+      "learning_rate": 2.6208647028234695e-05,
+      "loss": 1.1408,
+      "step": 32405
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 2.2214648723602295,
+      "learning_rate": 2.6202778654387737e-05,
+      "loss": 0.9565,
+      "step": 32410
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 2.63506817817688,
+      "learning_rate": 2.619691021411257e-05,
+      "loss": 1.1946,
+      "step": 32415
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 2.0027542114257812,
+      "learning_rate": 2.6191041707733293e-05,
+      "loss": 0.9843,
+      "step": 32420
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 3.6025943756103516,
+      "learning_rate": 2.618517313557402e-05,
+      "loss": 0.9494,
+      "step": 32425
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 3.030153751373291,
+      "learning_rate": 2.6179304497958855e-05,
+      "loss": 1.059,
+      "step": 32430
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 1.995133399963379,
+      "learning_rate": 2.6173435795211947e-05,
+      "loss": 0.9847,
+      "step": 32435
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 2.5811808109283447,
+      "learning_rate": 2.6167567027657397e-05,
+      "loss": 1.0383,
+      "step": 32440
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 3.842437982559204,
+      "learning_rate": 2.6161698195619327e-05,
+      "loss": 1.0719,
+      "step": 32445
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 5.087784290313721,
+      "learning_rate": 2.6155829299421875e-05,
+      "loss": 1.1697,
+      "step": 32450
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 3.2515745162963867,
+      "learning_rate": 2.6149960339389174e-05,
+      "loss": 1.0113,
+      "step": 32455
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 2.410245656967163,
+      "learning_rate": 2.6144091315845353e-05,
+      "loss": 1.1748,
+      "step": 32460
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 2.392496347427368,
+      "learning_rate": 2.6138222229114572e-05,
+      "loss": 1.1732,
+      "step": 32465
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 2.2038094997406006,
+      "learning_rate": 2.613235307952095e-05,
+      "loss": 0.9646,
+      "step": 32470
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 1.9154731035232544,
+      "learning_rate": 2.6126483867388645e-05,
+      "loss": 1.2406,
+      "step": 32475
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 1.628433108329773,
+      "learning_rate": 2.612061459304181e-05,
+      "loss": 1.1421,
+      "step": 32480
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 1.9438600540161133,
+      "learning_rate": 2.611474525680459e-05,
+      "loss": 1.0932,
+      "step": 32485
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 5.289341926574707,
+      "learning_rate": 2.6108875859001152e-05,
+      "loss": 0.9074,
+      "step": 32490
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 1.4697092771530151,
+      "learning_rate": 2.610300639995565e-05,
+      "loss": 1.178,
+      "step": 32495
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 2.8310253620147705,
+      "learning_rate": 2.6097136879992256e-05,
+      "loss": 1.0798,
+      "step": 32500
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 4.534608840942383,
+      "learning_rate": 2.609126729943513e-05,
+      "loss": 1.1111,
+      "step": 32505
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 4.378732681274414,
+      "learning_rate": 2.608539765860844e-05,
+      "loss": 1.1557,
+      "step": 32510
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 2.2998828887939453,
+      "learning_rate": 2.607952795783637e-05,
+      "loss": 1.0735,
+      "step": 32515
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 2.0673623085021973,
+      "learning_rate": 2.6073658197443095e-05,
+      "loss": 1.0948,
+      "step": 32520
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 2.536200761795044,
+      "learning_rate": 2.6067788377752793e-05,
+      "loss": 1.3438,
+      "step": 32525
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 1.4531900882720947,
+      "learning_rate": 2.6061918499089656e-05,
+      "loss": 0.9371,
+      "step": 32530
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 4.1081342697143555,
+      "learning_rate": 2.6056048561777852e-05,
+      "loss": 1.1155,
+      "step": 32535
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 8.0404691696167,
+      "learning_rate": 2.6050178566141585e-05,
+      "loss": 1.1525,
+      "step": 32540
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 2.6150429248809814,
+      "learning_rate": 2.6044308512505056e-05,
+      "loss": 1.1611,
+      "step": 32545
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 6.387349605560303,
+      "learning_rate": 2.6038438401192444e-05,
+      "loss": 1.0697,
+      "step": 32550
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 3.0312044620513916,
+      "learning_rate": 2.6032568232527964e-05,
+      "loss": 1.0854,
+      "step": 32555
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 1.371320366859436,
+      "learning_rate": 2.6026698006835814e-05,
+      "loss": 1.0813,
+      "step": 32560
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 2.4062769412994385,
+      "learning_rate": 2.60208277244402e-05,
+      "loss": 1.0397,
+      "step": 32565
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 2.3554258346557617,
+      "learning_rate": 2.601495738566533e-05,
+      "loss": 1.1501,
+      "step": 32570
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 2.7629899978637695,
+      "learning_rate": 2.6009086990835418e-05,
+      "loss": 1.039,
+      "step": 32575
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 1.9334179162979126,
+      "learning_rate": 2.6003216540274682e-05,
+      "loss": 1.1206,
+      "step": 32580
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 2.248354434967041,
+      "learning_rate": 2.5997346034307337e-05,
+      "loss": 1.0189,
+      "step": 32585
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 2.1622071266174316,
+      "learning_rate": 2.5991475473257608e-05,
+      "loss": 1.0698,
+      "step": 32590
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 1.745422601699829,
+      "learning_rate": 2.598560485744972e-05,
+      "loss": 0.8944,
+      "step": 32595
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 2.3082125186920166,
+      "learning_rate": 2.59797341872079e-05,
+      "loss": 1.0338,
+      "step": 32600
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 3.817018985748291,
+      "learning_rate": 2.5973863462856378e-05,
+      "loss": 0.965,
+      "step": 32605
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 2.986459493637085,
+      "learning_rate": 2.596799268471939e-05,
+      "loss": 1.1374,
+      "step": 32610
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 4.016047477722168,
+      "learning_rate": 2.5962121853121174e-05,
+      "loss": 1.047,
+      "step": 32615
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 3.042452812194824,
+      "learning_rate": 2.5956250968385966e-05,
+      "loss": 0.9784,
+      "step": 32620
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 2.1322600841522217,
+      "learning_rate": 2.5950380030838017e-05,
+      "loss": 1.0133,
+      "step": 32625
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 2.257575511932373,
+      "learning_rate": 2.5944509040801564e-05,
+      "loss": 1.0372,
+      "step": 32630
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 2.1339566707611084,
+      "learning_rate": 2.593863799860085e-05,
+      "loss": 1.0806,
+      "step": 32635
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 2.67110538482666,
+      "learning_rate": 2.593276690456014e-05,
+      "loss": 1.0397,
+      "step": 32640
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 4.476059913635254,
+      "learning_rate": 2.592689575900369e-05,
+      "loss": 1.0341,
+      "step": 32645
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 3.7582499980926514,
+      "learning_rate": 2.592102456225574e-05,
+      "loss": 0.9378,
+      "step": 32650
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 2.036689043045044,
+      "learning_rate": 2.5915153314640566e-05,
+      "loss": 1.0847,
+      "step": 32655
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 8.208381652832031,
+      "learning_rate": 2.5909282016482435e-05,
+      "loss": 0.9298,
+      "step": 32660
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 2.9294674396514893,
+      "learning_rate": 2.5903410668105586e-05,
+      "loss": 1.1432,
+      "step": 32665
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 1.4900659322738647,
+      "learning_rate": 2.5897539269834313e-05,
+      "loss": 1.0694,
+      "step": 32670
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 7.3784990310668945,
+      "learning_rate": 2.5891667821992883e-05,
+      "loss": 1.0586,
+      "step": 32675
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 1.4126672744750977,
+      "learning_rate": 2.588579632490556e-05,
+      "loss": 1.0634,
+      "step": 32680
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 2.6837716102600098,
+      "learning_rate": 2.587992477889663e-05,
+      "loss": 1.0635,
+      "step": 32685
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 1.9375698566436768,
+      "learning_rate": 2.5874053184290366e-05,
+      "loss": 1.1508,
+      "step": 32690
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 2.615001678466797,
+      "learning_rate": 2.586818154141105e-05,
+      "loss": 1.2275,
+      "step": 32695
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 2.7543222904205322,
+      "learning_rate": 2.5862309850582977e-05,
+      "loss": 1.072,
+      "step": 32700
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 1.6145623922348022,
+      "learning_rate": 2.5856438112130427e-05,
+      "loss": 0.9575,
+      "step": 32705
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 2.7645764350891113,
+      "learning_rate": 2.585056632637769e-05,
+      "loss": 1.1105,
+      "step": 32710
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 1.5867153406143188,
+      "learning_rate": 2.5844694493649054e-05,
+      "loss": 0.9333,
+      "step": 32715
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 1.9193617105484009,
+      "learning_rate": 2.583882261426882e-05,
+      "loss": 1.2603,
+      "step": 32720
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 2.898785352706909,
+      "learning_rate": 2.5832950688561297e-05,
+      "loss": 1.1845,
+      "step": 32725
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 1.4677989482879639,
+      "learning_rate": 2.582707871685076e-05,
+      "loss": 1.1262,
+      "step": 32730
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 2.376750946044922,
+      "learning_rate": 2.582120669946153e-05,
+      "loss": 1.183,
+      "step": 32735
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 2.5935792922973633,
+      "learning_rate": 2.5815334636717902e-05,
+      "loss": 1.0188,
+      "step": 32740
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 4.154141902923584,
+      "learning_rate": 2.5809462528944195e-05,
+      "loss": 1.1518,
+      "step": 32745
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 4.1396260261535645,
+      "learning_rate": 2.5803590376464716e-05,
+      "loss": 1.1648,
+      "step": 32750
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 2.666245698928833,
+      "learning_rate": 2.5797718179603776e-05,
+      "loss": 1.0906,
+      "step": 32755
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 2.568533420562744,
+      "learning_rate": 2.579184593868569e-05,
+      "loss": 0.9812,
+      "step": 32760
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 3.8902790546417236,
+      "learning_rate": 2.578597365403477e-05,
+      "loss": 0.9988,
+      "step": 32765
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 1.01784086227417,
+      "learning_rate": 2.578010132597534e-05,
+      "loss": 1.2201,
+      "step": 32770
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 23.18547248840332,
+      "learning_rate": 2.577422895483173e-05,
+      "loss": 0.8956,
+      "step": 32775
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 2.158190965652466,
+      "learning_rate": 2.5768356540928256e-05,
+      "loss": 1.0248,
+      "step": 32780
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 6.015243053436279,
+      "learning_rate": 2.5762484084589256e-05,
+      "loss": 1.1377,
+      "step": 32785
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 2.5766751766204834,
+      "learning_rate": 2.5756611586139044e-05,
+      "loss": 0.8034,
+      "step": 32790
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 3.290252208709717,
+      "learning_rate": 2.5750739045901966e-05,
+      "loss": 1.3023,
+      "step": 32795
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 2.5434985160827637,
+      "learning_rate": 2.574486646420235e-05,
+      "loss": 1.1181,
+      "step": 32800
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 8.717535018920898,
+      "learning_rate": 2.5738993841364535e-05,
+      "loss": 1.056,
+      "step": 32805
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 2.3016560077667236,
+      "learning_rate": 2.5733121177712856e-05,
+      "loss": 1.2113,
+      "step": 32810
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 1.7972419261932373,
+      "learning_rate": 2.5727248473571653e-05,
+      "loss": 1.1958,
+      "step": 32815
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 3.321375846862793,
+      "learning_rate": 2.5721375729265283e-05,
+      "loss": 0.9634,
+      "step": 32820
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 3.124709129333496,
+      "learning_rate": 2.5715502945118075e-05,
+      "loss": 1.0188,
+      "step": 32825
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 1.5069762468338013,
+      "learning_rate": 2.570963012145438e-05,
+      "loss": 0.8482,
+      "step": 32830
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 2.640984535217285,
+      "learning_rate": 2.5703757258598554e-05,
+      "loss": 1.1124,
+      "step": 32835
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 4.078088283538818,
+      "learning_rate": 2.5697884356874947e-05,
+      "loss": 1.0427,
+      "step": 32840
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 3.2041678428649902,
+      "learning_rate": 2.5692011416607908e-05,
+      "loss": 0.9205,
+      "step": 32845
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 2.750579595565796,
+      "learning_rate": 2.5686138438121804e-05,
+      "loss": 1.2344,
+      "step": 32850
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 2.770933151245117,
+      "learning_rate": 2.568026542174099e-05,
+      "loss": 1.056,
+      "step": 32855
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 2.695629596710205,
+      "learning_rate": 2.5674392367789818e-05,
+      "loss": 0.8705,
+      "step": 32860
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 1.1691042184829712,
+      "learning_rate": 2.5668519276592658e-05,
+      "loss": 1.0706,
+      "step": 32865
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 4.2656097412109375,
+      "learning_rate": 2.5662646148473867e-05,
+      "loss": 0.9537,
+      "step": 32870
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 1.754385232925415,
+      "learning_rate": 2.5656772983757822e-05,
+      "loss": 1.1298,
+      "step": 32875
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 1.9311494827270508,
+      "learning_rate": 2.5650899782768882e-05,
+      "loss": 1.2659,
+      "step": 32880
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 3.109220504760742,
+      "learning_rate": 2.564502654583144e-05,
+      "loss": 1.4295,
+      "step": 32885
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 2.71708607673645,
+      "learning_rate": 2.563915327326984e-05,
+      "loss": 1.1224,
+      "step": 32890
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 2.3169498443603516,
+      "learning_rate": 2.5633279965408475e-05,
+      "loss": 0.8078,
+      "step": 32895
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 1.5977541208267212,
+      "learning_rate": 2.5627406622571708e-05,
+      "loss": 1.19,
+      "step": 32900
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 3.2006161212921143,
+      "learning_rate": 2.5621533245083934e-05,
+      "loss": 1.1088,
+      "step": 32905
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 1.4427859783172607,
+      "learning_rate": 2.5615659833269516e-05,
+      "loss": 1.0645,
+      "step": 32910
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 3.558783769607544,
+      "learning_rate": 2.5609786387452855e-05,
+      "loss": 1.1666,
+      "step": 32915
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 1.977691888809204,
+      "learning_rate": 2.5603912907958323e-05,
+      "loss": 1.0334,
+      "step": 32920
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 2.796060562133789,
+      "learning_rate": 2.5598039395110307e-05,
+      "loss": 0.9684,
+      "step": 32925
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 4.863877773284912,
+      "learning_rate": 2.5592165849233196e-05,
+      "loss": 1.1839,
+      "step": 32930
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 2.7297861576080322,
+      "learning_rate": 2.558629227065138e-05,
+      "loss": 1.1317,
+      "step": 32935
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 3.373911142349243,
+      "learning_rate": 2.5580418659689255e-05,
+      "loss": 1.1497,
+      "step": 32940
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 3.8792035579681396,
+      "learning_rate": 2.5574545016671204e-05,
+      "loss": 1.1907,
+      "step": 32945
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 3.2369890213012695,
+      "learning_rate": 2.556867134192164e-05,
+      "loss": 1.1728,
+      "step": 32950
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 4.953618049621582,
+      "learning_rate": 2.5562797635764936e-05,
+      "loss": 0.9538,
+      "step": 32955
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 3.96720290184021,
+      "learning_rate": 2.555692389852551e-05,
+      "loss": 1.0353,
+      "step": 32960
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 1.2469520568847656,
+      "learning_rate": 2.5551050130527753e-05,
+      "loss": 1.0343,
+      "step": 32965
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 3.0623819828033447,
+      "learning_rate": 2.554517633209607e-05,
+      "loss": 0.9463,
+      "step": 32970
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 2.923931837081909,
+      "learning_rate": 2.553930250355487e-05,
+      "loss": 1.115,
+      "step": 32975
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 3.7751688957214355,
+      "learning_rate": 2.553342864522856e-05,
+      "loss": 1.0373,
+      "step": 32980
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 3.002441167831421,
+      "learning_rate": 2.552755475744153e-05,
+      "loss": 1.2383,
+      "step": 32985
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 2.0195484161376953,
+      "learning_rate": 2.55216808405182e-05,
+      "loss": 1.033,
+      "step": 32990
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 3.1848971843719482,
+      "learning_rate": 2.551580689478298e-05,
+      "loss": 1.0186,
+      "step": 32995
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 1.623724341392517,
+      "learning_rate": 2.5509932920560274e-05,
+      "loss": 1.0359,
+      "step": 33000
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 3.2824020385742188,
+      "learning_rate": 2.5504058918174513e-05,
+      "loss": 1.2735,
+      "step": 33005
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 2.9799790382385254,
+      "learning_rate": 2.54981848879501e-05,
+      "loss": 1.0281,
+      "step": 33010
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 3.8928287029266357,
+      "learning_rate": 2.5492310830211456e-05,
+      "loss": 0.9883,
+      "step": 33015
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 4.279067516326904,
+      "learning_rate": 2.5486436745282995e-05,
+      "loss": 1.197,
+      "step": 33020
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 3.4032154083251953,
+      "learning_rate": 2.5480562633489135e-05,
+      "loss": 1.2078,
+      "step": 33025
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 3.3113880157470703,
+      "learning_rate": 2.5474688495154298e-05,
+      "loss": 1.2098,
+      "step": 33030
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 1.2559151649475098,
+      "learning_rate": 2.5468814330602913e-05,
+      "loss": 0.9934,
+      "step": 33035
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 1.3851603269577026,
+      "learning_rate": 2.5462940140159398e-05,
+      "loss": 1.1007,
+      "step": 33040
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 4.625222206115723,
+      "learning_rate": 2.5457065924148184e-05,
+      "loss": 1.074,
+      "step": 33045
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 2.140420913696289,
+      "learning_rate": 2.545119168289369e-05,
+      "loss": 1.1087,
+      "step": 33050
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 2.4748363494873047,
+      "learning_rate": 2.544531741672035e-05,
+      "loss": 1.0391,
+      "step": 33055
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 1.9943400621414185,
+      "learning_rate": 2.543944312595259e-05,
+      "loss": 0.932,
+      "step": 33060
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 4.314105987548828,
+      "learning_rate": 2.543356881091484e-05,
+      "loss": 1.0815,
+      "step": 33065
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 3.688622236251831,
+      "learning_rate": 2.5427694471931546e-05,
+      "loss": 1.1081,
+      "step": 33070
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 5.567243576049805,
+      "learning_rate": 2.542182010932712e-05,
+      "loss": 1.0279,
+      "step": 33075
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 1.8669523000717163,
+      "learning_rate": 2.541594572342602e-05,
+      "loss": 1.0404,
+      "step": 33080
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 4.838998317718506,
+      "learning_rate": 2.5410071314552664e-05,
+      "loss": 1.1254,
+      "step": 33085
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 4.472231388092041,
+      "learning_rate": 2.540419688303149e-05,
+      "loss": 1.0976,
+      "step": 33090
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 2.0864551067352295,
+      "learning_rate": 2.539832242918695e-05,
+      "loss": 1.0639,
+      "step": 33095
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 3.6057910919189453,
+      "learning_rate": 2.539244795334347e-05,
+      "loss": 1.0898,
+      "step": 33100
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 2.7485694885253906,
+      "learning_rate": 2.5386573455825503e-05,
+      "loss": 1.0214,
+      "step": 33105
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 2.115241050720215,
+      "learning_rate": 2.5380698936957486e-05,
+      "loss": 1.1543,
+      "step": 33110
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 3.4498496055603027,
+      "learning_rate": 2.5374824397063857e-05,
+      "loss": 1.2445,
+      "step": 33115
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 1.5659513473510742,
+      "learning_rate": 2.536894983646907e-05,
+      "loss": 1.1021,
+      "step": 33120
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 8.8037748336792,
+      "learning_rate": 2.5363075255497564e-05,
+      "loss": 0.9424,
+      "step": 33125
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 4.802347183227539,
+      "learning_rate": 2.5357200654473788e-05,
+      "loss": 1.0744,
+      "step": 33130
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 2.3840949535369873,
+      "learning_rate": 2.5351326033722194e-05,
+      "loss": 1.1233,
+      "step": 33135
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 2.0759215354919434,
+      "learning_rate": 2.534545139356723e-05,
+      "loss": 0.9863,
+      "step": 33140
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 1.2032520771026611,
+      "learning_rate": 2.533957673433334e-05,
+      "loss": 1.169,
+      "step": 33145
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 4.248573303222656,
+      "learning_rate": 2.5333702056344984e-05,
+      "loss": 0.9997,
+      "step": 33150
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 2.4295151233673096,
+      "learning_rate": 2.532782735992661e-05,
+      "loss": 1.0271,
+      "step": 33155
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 2.852827310562134,
+      "learning_rate": 2.5321952645402668e-05,
+      "loss": 1.0106,
+      "step": 33160
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 4.686607837677002,
+      "learning_rate": 2.5316077913097618e-05,
+      "loss": 1.045,
+      "step": 33165
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 1.6220539808273315,
+      "learning_rate": 2.5310203163335916e-05,
+      "loss": 1.2063,
+      "step": 33170
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 1.336083173751831,
+      "learning_rate": 2.530432839644202e-05,
+      "loss": 1.3131,
+      "step": 33175
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 1.3632646799087524,
+      "learning_rate": 2.5298453612740375e-05,
+      "loss": 1.073,
+      "step": 33180
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 1.0934840440750122,
+      "learning_rate": 2.529257881255545e-05,
+      "loss": 0.9939,
+      "step": 33185
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 1.7837378978729248,
+      "learning_rate": 2.5286703996211697e-05,
+      "loss": 1.3673,
+      "step": 33190
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 1.5650957822799683,
+      "learning_rate": 2.5280829164033588e-05,
+      "loss": 1.2912,
+      "step": 33195
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 1.1075351238250732,
+      "learning_rate": 2.5274954316345568e-05,
+      "loss": 1.0113,
+      "step": 33200
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 3.3911895751953125,
+      "learning_rate": 2.5269079453472117e-05,
+      "loss": 0.9913,
+      "step": 33205
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 4.012703895568848,
+      "learning_rate": 2.5263204575737683e-05,
+      "loss": 1.0538,
+      "step": 33210
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 3.5184414386749268,
+      "learning_rate": 2.525732968346673e-05,
+      "loss": 1.291,
+      "step": 33215
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 3.051833152770996,
+      "learning_rate": 2.5251454776983724e-05,
+      "loss": 1.0442,
+      "step": 33220
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 3.1589720249176025,
+      "learning_rate": 2.5245579856613137e-05,
+      "loss": 0.8625,
+      "step": 33225
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 2.489769458770752,
+      "learning_rate": 2.523970492267943e-05,
+      "loss": 1.0129,
+      "step": 33230
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 3.1582303047180176,
+      "learning_rate": 2.5233829975507068e-05,
+      "loss": 1.0015,
+      "step": 33235
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 2.759131908416748,
+      "learning_rate": 2.522795501542052e-05,
+      "loss": 1.1334,
+      "step": 33240
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 4.12851619720459,
+      "learning_rate": 2.522208004274425e-05,
+      "loss": 1.1993,
+      "step": 33245
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 1.7142823934555054,
+      "learning_rate": 2.5216205057802732e-05,
+      "loss": 1.2478,
+      "step": 33250
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 2.8924169540405273,
+      "learning_rate": 2.521033006092044e-05,
+      "loss": 1.0277,
+      "step": 33255
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 5.144822597503662,
+      "learning_rate": 2.5204455052421828e-05,
+      "loss": 1.3498,
+      "step": 33260
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 2.7403883934020996,
+      "learning_rate": 2.519858003263138e-05,
+      "loss": 0.9622,
+      "step": 33265
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 3.224583864212036,
+      "learning_rate": 2.5192705001873566e-05,
+      "loss": 1.2568,
+      "step": 33270
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 2.203730821609497,
+      "learning_rate": 2.518682996047285e-05,
+      "loss": 1.2091,
+      "step": 33275
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 2.4827523231506348,
+      "learning_rate": 2.5180954908753716e-05,
+      "loss": 1.0987,
+      "step": 33280
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 3.099918842315674,
+      "learning_rate": 2.5175079847040626e-05,
+      "loss": 1.0014,
+      "step": 33285
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 1.661093831062317,
+      "learning_rate": 2.5169204775658055e-05,
+      "loss": 1.0222,
+      "step": 33290
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 5.493398189544678,
+      "learning_rate": 2.5163329694930488e-05,
+      "loss": 1.0784,
+      "step": 33295
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 1.5556894540786743,
+      "learning_rate": 2.5157454605182386e-05,
+      "loss": 1.1672,
+      "step": 33300
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 3.597127676010132,
+      "learning_rate": 2.515157950673823e-05,
+      "loss": 1.0432,
+      "step": 33305
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 1.7815759181976318,
+      "learning_rate": 2.51457043999225e-05,
+      "loss": 1.0406,
+      "step": 33310
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 4.965347766876221,
+      "learning_rate": 2.5139829285059664e-05,
+      "loss": 1.0041,
+      "step": 33315
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 1.7741583585739136,
+      "learning_rate": 2.5133954162474195e-05,
+      "loss": 1.0557,
+      "step": 33320
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 2.6704812049865723,
+      "learning_rate": 2.512807903249058e-05,
+      "loss": 1.1279,
+      "step": 33325
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 1.5926697254180908,
+      "learning_rate": 2.51222038954333e-05,
+      "loss": 1.0901,
+      "step": 33330
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 1.6942665576934814,
+      "learning_rate": 2.5116328751626827e-05,
+      "loss": 1.034,
+      "step": 33335
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 3.1640069484710693,
+      "learning_rate": 2.5110453601395633e-05,
+      "loss": 1.0341,
+      "step": 33340
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 2.211376905441284,
+      "learning_rate": 2.5104578445064202e-05,
+      "loss": 1.0243,
+      "step": 33345
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 2.296888589859009,
+      "learning_rate": 2.5098703282957013e-05,
+      "loss": 1.0835,
+      "step": 33350
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 2.575103998184204,
+      "learning_rate": 2.5092828115398544e-05,
+      "loss": 0.963,
+      "step": 33355
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 11.325212478637695,
+      "learning_rate": 2.508695294271327e-05,
+      "loss": 0.9614,
+      "step": 33360
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 2.817223310470581,
+      "learning_rate": 2.5081077765225682e-05,
+      "loss": 1.0673,
+      "step": 33365
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 2.326043128967285,
+      "learning_rate": 2.5075202583260256e-05,
+      "loss": 1.145,
+      "step": 33370
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 2.059147357940674,
+      "learning_rate": 2.506932739714146e-05,
+      "loss": 1.3188,
+      "step": 33375
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 1.470616102218628,
+      "learning_rate": 2.506345220719379e-05,
+      "loss": 1.1262,
+      "step": 33380
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 2.974729537963867,
+      "learning_rate": 2.5057577013741716e-05,
+      "loss": 1.207,
+      "step": 33385
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 1.9753620624542236,
+      "learning_rate": 2.505170181710973e-05,
+      "loss": 1.1227,
+      "step": 33390
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 2.3012120723724365,
+      "learning_rate": 2.5045826617622298e-05,
+      "loss": 1.1889,
+      "step": 33395
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 1.2801812887191772,
+      "learning_rate": 2.503995141560392e-05,
+      "loss": 1.1202,
+      "step": 33400
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 4.384020805358887,
+      "learning_rate": 2.5034076211379053e-05,
+      "loss": 0.9629,
+      "step": 33405
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 2.4980461597442627,
+      "learning_rate": 2.5028201005272206e-05,
+      "loss": 1.1078,
+      "step": 33410
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 1.83892023563385,
+      "learning_rate": 2.5022325797607837e-05,
+      "loss": 1.1261,
+      "step": 33415
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 2.998216390609741,
+      "learning_rate": 2.5016450588710443e-05,
+      "loss": 1.0531,
+      "step": 33420
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 3.9695932865142822,
+      "learning_rate": 2.50105753789045e-05,
+      "loss": 0.8376,
+      "step": 33425
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 2.823977470397949,
+      "learning_rate": 2.5004700168514482e-05,
+      "loss": 0.9155,
+      "step": 33430
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 1.843510627746582,
+      "learning_rate": 2.4998824957864885e-05,
+      "loss": 1.1598,
+      "step": 33435
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 1.946144461631775,
+      "learning_rate": 2.499294974728019e-05,
+      "loss": 0.9176,
+      "step": 33440
+    },
+    {
+      "epoch": 10.01,
+      "grad_norm": 2.370626211166382,
+      "learning_rate": 2.498707453708486e-05,
+      "loss": 0.9641,
+      "step": 33445
+    },
+    {
+      "epoch": 10.01,
+      "grad_norm": 1.2042300701141357,
+      "learning_rate": 2.4981199327603404e-05,
+      "loss": 1.1646,
+      "step": 33450
+    },
+    {
+      "epoch": 10.01,
+      "grad_norm": 2.661841869354248,
+      "learning_rate": 2.497532411916029e-05,
+      "loss": 1.1475,
+      "step": 33455
+    },
+    {
+      "epoch": 10.01,
+      "grad_norm": 2.6765615940093994,
+      "learning_rate": 2.4969448912079985e-05,
+      "loss": 0.9705,
+      "step": 33460
+    },
+    {
+      "epoch": 10.01,
+      "grad_norm": 2.9539201259613037,
+      "learning_rate": 2.4963573706686997e-05,
+      "loss": 0.7655,
+      "step": 33465
+    },
+    {
+      "epoch": 10.01,
+      "grad_norm": 6.744268894195557,
+      "learning_rate": 2.4957698503305786e-05,
+      "loss": 0.8524,
+      "step": 33470
+    },
+    {
+      "epoch": 10.02,
+      "grad_norm": 2.093243360519409,
+      "learning_rate": 2.495182330226085e-05,
+      "loss": 1.0871,
+      "step": 33475
+    },
+    {
+      "epoch": 10.02,
+      "grad_norm": 2.485497236251831,
+      "learning_rate": 2.494594810387666e-05,
+      "loss": 0.8595,
+      "step": 33480
+    },
+    {
+      "epoch": 10.02,
+      "grad_norm": 3.0233352184295654,
+      "learning_rate": 2.49400729084777e-05,
+      "loss": 0.8909,
+      "step": 33485
+    },
+    {
+      "epoch": 10.02,
+      "grad_norm": 2.09739351272583,
+      "learning_rate": 2.493419771638845e-05,
+      "loss": 1.0453,
+      "step": 33490
+    },
+    {
+      "epoch": 10.02,
+      "grad_norm": 2.0949864387512207,
+      "learning_rate": 2.4928322527933393e-05,
+      "loss": 0.9512,
+      "step": 33495
+    },
+    {
+      "epoch": 10.02,
+      "grad_norm": 2.8882389068603516,
+      "learning_rate": 2.4922447343437005e-05,
+      "loss": 1.0104,
+      "step": 33500
+    },
+    {
+      "epoch": 10.02,
+      "grad_norm": 1.7648423910140991,
+      "learning_rate": 2.4916572163223772e-05,
+      "loss": 0.9308,
+      "step": 33505
+    },
+    {
+      "epoch": 10.03,
+      "grad_norm": 16.295032501220703,
+      "learning_rate": 2.4910696987618174e-05,
+      "loss": 1.057,
+      "step": 33510
+    },
+    {
+      "epoch": 10.03,
+      "grad_norm": 4.960565090179443,
+      "learning_rate": 2.490482181694469e-05,
+      "loss": 1.0269,
+      "step": 33515
+    },
+    {
+      "epoch": 10.03,
+      "grad_norm": 1.606244683265686,
+      "learning_rate": 2.489894665152779e-05,
+      "loss": 1.1142,
+      "step": 33520
+    },
+    {
+      "epoch": 10.03,
+      "grad_norm": 2.571674108505249,
+      "learning_rate": 2.489307149169197e-05,
+      "loss": 1.1242,
+      "step": 33525
+    },
+    {
+      "epoch": 10.03,
+      "grad_norm": 1.8234943151474,
+      "learning_rate": 2.4887196337761693e-05,
+      "loss": 1.0145,
+      "step": 33530
+    },
+    {
+      "epoch": 10.03,
+      "grad_norm": 1.6958742141723633,
+      "learning_rate": 2.4881321190061453e-05,
+      "loss": 1.0754,
+      "step": 33535
+    },
+    {
+      "epoch": 10.03,
+      "grad_norm": 1.0637584924697876,
+      "learning_rate": 2.487544604891571e-05,
+      "loss": 1.028,
+      "step": 33540
+    },
+    {
+      "epoch": 10.04,
+      "grad_norm": 1.169573426246643,
+      "learning_rate": 2.4869570914648963e-05,
+      "loss": 1.0321,
+      "step": 33545
+    },
+    {
+      "epoch": 10.04,
+      "grad_norm": 1.4608122110366821,
+      "learning_rate": 2.4863695787585678e-05,
+      "loss": 1.1961,
+      "step": 33550
+    },
+    {
+      "epoch": 10.04,
+      "grad_norm": 2.158911943435669,
+      "learning_rate": 2.4857820668050324e-05,
+      "loss": 1.1969,
+      "step": 33555
+    },
+    {
+      "epoch": 10.04,
+      "grad_norm": 9.060139656066895,
+      "learning_rate": 2.4851945556367396e-05,
+      "loss": 0.7947,
+      "step": 33560
+    },
+    {
+      "epoch": 10.04,
+      "grad_norm": 4.058126449584961,
+      "learning_rate": 2.484607045286135e-05,
+      "loss": 0.8664,
+      "step": 33565
+    },
+    {
+      "epoch": 10.04,
+      "grad_norm": 3.4800760746002197,
+      "learning_rate": 2.4840195357856685e-05,
+      "loss": 1.0867,
+      "step": 33570
+    },
+    {
+      "epoch": 10.05,
+      "grad_norm": 2.5772833824157715,
+      "learning_rate": 2.483432027167785e-05,
+      "loss": 0.9432,
+      "step": 33575
+    },
+    {
+      "epoch": 10.05,
+      "grad_norm": 1.0311754941940308,
+      "learning_rate": 2.482844519464935e-05,
+      "loss": 1.0534,
+      "step": 33580
+    },
+    {
+      "epoch": 10.05,
+      "grad_norm": 3.5726497173309326,
+      "learning_rate": 2.4822570127095637e-05,
+      "loss": 0.9547,
+      "step": 33585
+    },
+    {
+      "epoch": 10.05,
+      "grad_norm": 3.7769217491149902,
+      "learning_rate": 2.4816695069341192e-05,
+      "loss": 1.01,
+      "step": 33590
+    },
+    {
+      "epoch": 10.05,
+      "grad_norm": 1.1994178295135498,
+      "learning_rate": 2.4810820021710486e-05,
+      "loss": 1.1168,
+      "step": 33595
+    },
+    {
+      "epoch": 10.05,
+      "grad_norm": 3.4825947284698486,
+      "learning_rate": 2.4804944984527995e-05,
+      "loss": 1.0242,
+      "step": 33600
+    },
+    {
+      "epoch": 10.05,
+      "grad_norm": 1.254137396812439,
+      "learning_rate": 2.4799069958118187e-05,
+      "loss": 0.9139,
+      "step": 33605
+    },
+    {
+      "epoch": 10.06,
+      "grad_norm": 5.904828071594238,
+      "learning_rate": 2.4793194942805545e-05,
+      "loss": 1.0005,
+      "step": 33610
+    },
+    {
+      "epoch": 10.06,
+      "grad_norm": 2.302828073501587,
+      "learning_rate": 2.478731993891452e-05,
+      "loss": 1.0739,
+      "step": 33615
+    },
+    {
+      "epoch": 10.06,
+      "grad_norm": 2.0300052165985107,
+      "learning_rate": 2.4781444946769603e-05,
+      "loss": 0.9196,
+      "step": 33620
+    },
+    {
+      "epoch": 10.06,
+      "grad_norm": 2.183711051940918,
+      "learning_rate": 2.4775569966695242e-05,
+      "loss": 1.069,
+      "step": 33625
+    },
+    {
+      "epoch": 10.06,
+      "grad_norm": 6.40975284576416,
+      "learning_rate": 2.476969499901593e-05,
+      "loss": 1.0779,
+      "step": 33630
+    },
+    {
+      "epoch": 10.06,
+      "grad_norm": 2.763566017150879,
+      "learning_rate": 2.4763820044056114e-05,
+      "loss": 1.191,
+      "step": 33635
+    },
+    {
+      "epoch": 10.06,
+      "grad_norm": 6.577224254608154,
+      "learning_rate": 2.4757945102140287e-05,
+      "loss": 0.9616,
+      "step": 33640
+    },
+    {
+      "epoch": 10.07,
+      "grad_norm": 2.686187267303467,
+      "learning_rate": 2.4752070173592895e-05,
+      "loss": 1.21,
+      "step": 33645
+    },
+    {
+      "epoch": 10.07,
+      "grad_norm": 1.8847886323928833,
+      "learning_rate": 2.47461952587384e-05,
+      "loss": 1.0332,
+      "step": 33650
+    },
+    {
+      "epoch": 10.07,
+      "grad_norm": 4.264788627624512,
+      "learning_rate": 2.4740320357901286e-05,
+      "loss": 0.9612,
+      "step": 33655
+    },
+    {
+      "epoch": 10.07,
+      "grad_norm": 3.13706111907959,
+      "learning_rate": 2.4734445471406e-05,
+      "loss": 1.1428,
+      "step": 33660
+    },
+    {
+      "epoch": 10.07,
+      "grad_norm": 3.5950639247894287,
+      "learning_rate": 2.472857059957703e-05,
+      "loss": 1.0232,
+      "step": 33665
+    },
+    {
+      "epoch": 10.07,
+      "grad_norm": 4.524314880371094,
+      "learning_rate": 2.4722695742738806e-05,
+      "loss": 0.8956,
+      "step": 33670
+    },
+    {
+      "epoch": 10.08,
+      "grad_norm": 2.5336546897888184,
+      "learning_rate": 2.471682090121582e-05,
+      "loss": 1.1582,
+      "step": 33675
+    },
+    {
+      "epoch": 10.08,
+      "grad_norm": 2.4728264808654785,
+      "learning_rate": 2.4710946075332515e-05,
+      "loss": 1.1319,
+      "step": 33680
+    },
+    {
+      "epoch": 10.08,
+      "grad_norm": 2.4756505489349365,
+      "learning_rate": 2.4705071265413355e-05,
+      "loss": 1.0484,
+      "step": 33685
+    },
+    {
+      "epoch": 10.08,
+      "grad_norm": 4.324848651885986,
+      "learning_rate": 2.469919647178281e-05,
+      "loss": 0.9865,
+      "step": 33690
+    },
+    {
+      "epoch": 10.08,
+      "grad_norm": 1.772365927696228,
+      "learning_rate": 2.4693321694765324e-05,
+      "loss": 1.1322,
+      "step": 33695
+    },
+    {
+      "epoch": 10.08,
+      "grad_norm": 4.808743476867676,
+      "learning_rate": 2.468744693468537e-05,
+      "loss": 0.9088,
+      "step": 33700
+    },
+    {
+      "epoch": 10.08,
+      "grad_norm": 2.5702743530273438,
+      "learning_rate": 2.46815721918674e-05,
+      "loss": 1.1035,
+      "step": 33705
+    },
+    {
+      "epoch": 10.09,
+      "grad_norm": 2.589160680770874,
+      "learning_rate": 2.4675697466635855e-05,
+      "loss": 1.2078,
+      "step": 33710
+    },
+    {
+      "epoch": 10.09,
+      "grad_norm": 1.3858650922775269,
+      "learning_rate": 2.466982275931521e-05,
+      "loss": 0.9329,
+      "step": 33715
+    },
+    {
+      "epoch": 10.09,
+      "grad_norm": 2.2143914699554443,
+      "learning_rate": 2.4663948070229905e-05,
+      "loss": 1.0403,
+      "step": 33720
+    },
+    {
+      "epoch": 10.09,
+      "grad_norm": 1.3419291973114014,
+      "learning_rate": 2.4658073399704405e-05,
+      "loss": 1.1443,
+      "step": 33725
+    },
+    {
+      "epoch": 10.09,
+      "grad_norm": 3.204826593399048,
+      "learning_rate": 2.4652198748063146e-05,
+      "loss": 1.0266,
+      "step": 33730
+    },
+    {
+      "epoch": 10.09,
+      "grad_norm": 2.098097085952759,
+      "learning_rate": 2.46463241156306e-05,
+      "loss": 1.1408,
+      "step": 33735
+    },
+    {
+      "epoch": 10.09,
+      "grad_norm": 3.6927952766418457,
+      "learning_rate": 2.4640449502731204e-05,
+      "loss": 1.0207,
+      "step": 33740
+    },
+    {
+      "epoch": 10.1,
+      "grad_norm": 4.6298828125,
+      "learning_rate": 2.46345749096894e-05,
+      "loss": 1.11,
+      "step": 33745
+    },
+    {
+      "epoch": 10.1,
+      "grad_norm": 1.5225332975387573,
+      "learning_rate": 2.4628700336829655e-05,
+      "loss": 1.0937,
+      "step": 33750
+    },
+    {
+      "epoch": 10.1,
+      "grad_norm": 2.734102964401245,
+      "learning_rate": 2.4622825784476392e-05,
+      "loss": 0.9719,
+      "step": 33755
+    },
+    {
+      "epoch": 10.1,
+      "grad_norm": 25.489742279052734,
+      "learning_rate": 2.4616951252954078e-05,
+      "loss": 1.1438,
+      "step": 33760
+    },
+    {
+      "epoch": 10.1,
+      "grad_norm": 1.8963426351547241,
+      "learning_rate": 2.4611076742587137e-05,
+      "loss": 1.106,
+      "step": 33765
+    },
+    {
+      "epoch": 10.1,
+      "grad_norm": 2.171485662460327,
+      "learning_rate": 2.4605202253700034e-05,
+      "loss": 1.0517,
+      "step": 33770
+    },
+    {
+      "epoch": 10.11,
+      "grad_norm": 3.205883502960205,
+      "learning_rate": 2.45993277866172e-05,
+      "loss": 1.1039,
+      "step": 33775
+    },
+    {
+      "epoch": 10.11,
+      "grad_norm": 3.306411027908325,
+      "learning_rate": 2.459345334166307e-05,
+      "loss": 1.0611,
+      "step": 33780
+    },
+    {
+      "epoch": 10.11,
+      "grad_norm": 2.3397672176361084,
+      "learning_rate": 2.4587578919162097e-05,
+      "loss": 1.1455,
+      "step": 33785
+    },
+    {
+      "epoch": 10.11,
+      "grad_norm": 3.6171796321868896,
+      "learning_rate": 2.45817045194387e-05,
+      "loss": 0.9492,
+      "step": 33790
+    },
+    {
+      "epoch": 10.11,
+      "grad_norm": 1.5914018154144287,
+      "learning_rate": 2.4575830142817342e-05,
+      "loss": 0.9019,
+      "step": 33795
+    },
+    {
+      "epoch": 10.11,
+      "grad_norm": 1.1069579124450684,
+      "learning_rate": 2.456995578962243e-05,
+      "loss": 1.1982,
+      "step": 33800
+    },
+    {
+      "epoch": 10.11,
+      "grad_norm": 4.241572856903076,
+      "learning_rate": 2.4564081460178427e-05,
+      "loss": 1.0414,
+      "step": 33805
+    },
+    {
+      "epoch": 10.12,
+      "grad_norm": 2.3619353771209717,
+      "learning_rate": 2.455820715480975e-05,
+      "loss": 1.16,
+      "step": 33810
+    },
+    {
+      "epoch": 10.12,
+      "grad_norm": 1.428951621055603,
+      "learning_rate": 2.4552332873840818e-05,
+      "loss": 1.1485,
+      "step": 33815
+    },
+    {
+      "epoch": 10.12,
+      "grad_norm": 1.82322359085083,
+      "learning_rate": 2.454645861759609e-05,
+      "loss": 1.0625,
+      "step": 33820
+    },
+    {
+      "epoch": 10.12,
+      "grad_norm": 2.700174570083618,
+      "learning_rate": 2.4540584386399974e-05,
+      "loss": 0.8932,
+      "step": 33825
+    },
+    {
+      "epoch": 10.12,
+      "grad_norm": 3.158874988555908,
+      "learning_rate": 2.4534710180576912e-05,
+      "loss": 1.1615,
+      "step": 33830
+    },
+    {
+      "epoch": 10.12,
+      "grad_norm": 6.1408820152282715,
+      "learning_rate": 2.4528836000451323e-05,
+      "loss": 1.0226,
+      "step": 33835
+    },
+    {
+      "epoch": 10.12,
+      "grad_norm": 1.1300007104873657,
+      "learning_rate": 2.452296184634762e-05,
+      "loss": 0.9967,
+      "step": 33840
+    },
+    {
+      "epoch": 10.13,
+      "grad_norm": 8.636886596679688,
+      "learning_rate": 2.4517087718590244e-05,
+      "loss": 0.9646,
+      "step": 33845
+    },
+    {
+      "epoch": 10.13,
+      "grad_norm": 4.390270233154297,
+      "learning_rate": 2.4511213617503616e-05,
+      "loss": 1.131,
+      "step": 33850
+    },
+    {
+      "epoch": 10.13,
+      "grad_norm": 4.6113409996032715,
+      "learning_rate": 2.4505339543412148e-05,
+      "loss": 1.1715,
+      "step": 33855
+    },
+    {
+      "epoch": 10.13,
+      "grad_norm": 2.2898902893066406,
+      "learning_rate": 2.449946549664026e-05,
+      "loss": 1.0189,
+      "step": 33860
+    },
+    {
+      "epoch": 10.13,
+      "grad_norm": 0.8719561100006104,
+      "learning_rate": 2.449359147751238e-05,
+      "loss": 1.1226,
+      "step": 33865
+    },
+    {
+      "epoch": 10.13,
+      "grad_norm": 3.1541287899017334,
+      "learning_rate": 2.4487717486352914e-05,
+      "loss": 0.9081,
+      "step": 33870
+    },
+    {
+      "epoch": 10.14,
+      "grad_norm": 4.781918048858643,
+      "learning_rate": 2.448184352348627e-05,
+      "loss": 1.2464,
+      "step": 33875
+    },
+    {
+      "epoch": 10.14,
+      "grad_norm": 3.1645240783691406,
+      "learning_rate": 2.4475969589236887e-05,
+      "loss": 0.847,
+      "step": 33880
+    },
+    {
+      "epoch": 10.14,
+      "grad_norm": 2.1200978755950928,
+      "learning_rate": 2.447009568392914e-05,
+      "loss": 1.1901,
+      "step": 33885
+    },
+    {
+      "epoch": 10.14,
+      "grad_norm": 3.1781980991363525,
+      "learning_rate": 2.446422180788747e-05,
+      "loss": 0.9578,
+      "step": 33890
+    },
+    {
+      "epoch": 10.14,
+      "grad_norm": 4.06983757019043,
+      "learning_rate": 2.4458347961436264e-05,
+      "loss": 0.9594,
+      "step": 33895
+    },
+    {
+      "epoch": 10.14,
+      "grad_norm": 13.415980339050293,
+      "learning_rate": 2.4452474144899947e-05,
+      "loss": 1.0795,
+      "step": 33900
+    },
+    {
+      "epoch": 10.14,
+      "grad_norm": 2.0610928535461426,
+      "learning_rate": 2.4446600358602915e-05,
+      "loss": 1.0463,
+      "step": 33905
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 1.7884243726730347,
+      "learning_rate": 2.4440726602869557e-05,
+      "loss": 1.1359,
+      "step": 33910
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 2.057011604309082,
+      "learning_rate": 2.44348528780243e-05,
+      "loss": 1.0827,
+      "step": 33915
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 3.1838924884796143,
+      "learning_rate": 2.442897918439152e-05,
+      "loss": 1.107,
+      "step": 33920
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 2.055171251296997,
+      "learning_rate": 2.4423105522295633e-05,
+      "loss": 0.9141,
+      "step": 33925
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 7.081864833831787,
+      "learning_rate": 2.441723189206102e-05,
+      "loss": 0.9043,
+      "step": 33930
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 1.6306078433990479,
+      "learning_rate": 2.44113582940121e-05,
+      "loss": 1.1316,
+      "step": 33935
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 3.7581050395965576,
+      "learning_rate": 2.440548472847324e-05,
+      "loss": 1.058,
+      "step": 33940
+    },
+    {
+      "epoch": 10.16,
+      "grad_norm": 2.7319552898406982,
+      "learning_rate": 2.4399611195768836e-05,
+      "loss": 1.2098,
+      "step": 33945
+    },
+    {
+      "epoch": 10.16,
+      "grad_norm": 4.5412068367004395,
+      "learning_rate": 2.4393737696223286e-05,
+      "loss": 1.0749,
+      "step": 33950
+    },
+    {
+      "epoch": 10.16,
+      "grad_norm": 3.6076760292053223,
+      "learning_rate": 2.4387864230160972e-05,
+      "loss": 1.0315,
+      "step": 33955
+    },
+    {
+      "epoch": 10.16,
+      "grad_norm": 1.9118307828903198,
+      "learning_rate": 2.4381990797906284e-05,
+      "loss": 1.0247,
+      "step": 33960
+    },
+    {
+      "epoch": 10.16,
+      "grad_norm": 2.3119821548461914,
+      "learning_rate": 2.4376117399783605e-05,
+      "loss": 1.0456,
+      "step": 33965
+    },
+    {
+      "epoch": 10.16,
+      "grad_norm": 1.257812261581421,
+      "learning_rate": 2.43702440361173e-05,
+      "loss": 1.1916,
+      "step": 33970
+    },
+    {
+      "epoch": 10.16,
+      "grad_norm": 3.263524055480957,
+      "learning_rate": 2.4364370707231777e-05,
+      "loss": 1.1353,
+      "step": 33975
+    },
+    {
+      "epoch": 10.17,
+      "grad_norm": 2.8173470497131348,
+      "learning_rate": 2.4358497413451383e-05,
+      "loss": 1.1754,
+      "step": 33980
+    },
+    {
+      "epoch": 10.17,
+      "grad_norm": 1.8419231176376343,
+      "learning_rate": 2.4352624155100525e-05,
+      "loss": 0.9914,
+      "step": 33985
+    },
+    {
+      "epoch": 10.17,
+      "grad_norm": 4.419125080108643,
+      "learning_rate": 2.434675093250355e-05,
+      "loss": 1.0154,
+      "step": 33990
+    },
+    {
+      "epoch": 10.17,
+      "grad_norm": 4.144635200500488,
+      "learning_rate": 2.4340877745984854e-05,
+      "loss": 1.1304,
+      "step": 33995
+    },
+    {
+      "epoch": 10.17,
+      "grad_norm": 2.172372341156006,
+      "learning_rate": 2.4335004595868794e-05,
+      "loss": 1.0345,
+      "step": 34000
+    },
+    {
+      "epoch": 10.17,
+      "grad_norm": 3.2084264755249023,
+      "learning_rate": 2.4329131482479727e-05,
+      "loss": 1.1348,
+      "step": 34005
+    },
+    {
+      "epoch": 10.18,
+      "grad_norm": 1.3004094362258911,
+      "learning_rate": 2.4323258406142042e-05,
+      "loss": 1.0589,
+      "step": 34010
+    },
+    {
+      "epoch": 10.18,
+      "grad_norm": 2.1280858516693115,
+      "learning_rate": 2.4317385367180083e-05,
+      "loss": 1.16,
+      "step": 34015
+    },
+    {
+      "epoch": 10.18,
+      "grad_norm": 3.7424776554107666,
+      "learning_rate": 2.4311512365918232e-05,
+      "loss": 1.0512,
+      "step": 34020
+    },
+    {
+      "epoch": 10.18,
+      "grad_norm": 2.389583110809326,
+      "learning_rate": 2.4305639402680825e-05,
+      "loss": 1.2194,
+      "step": 34025
+    },
+    {
+      "epoch": 10.18,
+      "grad_norm": 9.110445022583008,
+      "learning_rate": 2.4299766477792242e-05,
+      "loss": 1.1583,
+      "step": 34030
+    },
+    {
+      "epoch": 10.18,
+      "grad_norm": 2.022211790084839,
+      "learning_rate": 2.4293893591576825e-05,
+      "loss": 1.0861,
+      "step": 34035
+    },
+    {
+      "epoch": 10.18,
+      "grad_norm": 3.275707483291626,
+      "learning_rate": 2.428802074435893e-05,
+      "loss": 0.866,
+      "step": 34040
+    },
+    {
+      "epoch": 10.19,
+      "grad_norm": 1.1770823001861572,
+      "learning_rate": 2.428214793646291e-05,
+      "loss": 1.2091,
+      "step": 34045
+    },
+    {
+      "epoch": 10.19,
+      "grad_norm": 7.510256767272949,
+      "learning_rate": 2.4276275168213105e-05,
+      "loss": 1.1663,
+      "step": 34050
+    },
+    {
+      "epoch": 10.19,
+      "grad_norm": 2.775092840194702,
+      "learning_rate": 2.4270402439933886e-05,
+      "loss": 1.151,
+      "step": 34055
+    },
+    {
+      "epoch": 10.19,
+      "grad_norm": 2.5497207641601562,
+      "learning_rate": 2.4264529751949576e-05,
+      "loss": 1.1476,
+      "step": 34060
+    },
+    {
+      "epoch": 10.19,
+      "grad_norm": 1.3202245235443115,
+      "learning_rate": 2.4258657104584518e-05,
+      "loss": 1.0267,
+      "step": 34065
+    },
+    {
+      "epoch": 10.19,
+      "grad_norm": 1.2881629467010498,
+      "learning_rate": 2.4252784498163064e-05,
+      "loss": 1.0565,
+      "step": 34070
+    },
+    {
+      "epoch": 10.19,
+      "grad_norm": 2.9040749073028564,
+      "learning_rate": 2.4246911933009536e-05,
+      "loss": 1.1041,
+      "step": 34075
+    },
+    {
+      "epoch": 10.2,
+      "grad_norm": 1.2760438919067383,
+      "learning_rate": 2.424103940944829e-05,
+      "loss": 1.2469,
+      "step": 34080
+    },
+    {
+      "epoch": 10.2,
+      "grad_norm": 1.911608099937439,
+      "learning_rate": 2.423516692780364e-05,
+      "loss": 0.9625,
+      "step": 34085
+    },
+    {
+      "epoch": 10.2,
+      "grad_norm": 1.7194567918777466,
+      "learning_rate": 2.4229294488399935e-05,
+      "loss": 1.1776,
+      "step": 34090
+    },
+    {
+      "epoch": 10.2,
+      "grad_norm": 4.4640960693359375,
+      "learning_rate": 2.4223422091561493e-05,
+      "loss": 0.9402,
+      "step": 34095
+    },
+    {
+      "epoch": 10.2,
+      "grad_norm": 5.374719142913818,
+      "learning_rate": 2.4217549737612632e-05,
+      "loss": 1.0412,
+      "step": 34100
+    },
+    {
+      "epoch": 10.2,
+      "grad_norm": 5.335212230682373,
+      "learning_rate": 2.4211677426877698e-05,
+      "loss": 0.9229,
+      "step": 34105
+    },
+    {
+      "epoch": 10.21,
+      "grad_norm": 1.18242609500885,
+      "learning_rate": 2.4205805159680986e-05,
+      "loss": 0.9371,
+      "step": 34110
+    },
+    {
+      "epoch": 10.21,
+      "grad_norm": 2.225745916366577,
+      "learning_rate": 2.419993293634684e-05,
+      "loss": 1.1263,
+      "step": 34115
+    },
+    {
+      "epoch": 10.21,
+      "grad_norm": 4.692055702209473,
+      "learning_rate": 2.4194060757199557e-05,
+      "loss": 1.1487,
+      "step": 34120
+    },
+    {
+      "epoch": 10.21,
+      "grad_norm": 3.8176944255828857,
+      "learning_rate": 2.4188188622563474e-05,
+      "loss": 1.1546,
+      "step": 34125
+    },
+    {
+      "epoch": 10.21,
+      "grad_norm": 2.7893073558807373,
+      "learning_rate": 2.4182316532762884e-05,
+      "loss": 0.9161,
+      "step": 34130
+    },
+    {
+      "epoch": 10.21,
+      "grad_norm": 5.134355545043945,
+      "learning_rate": 2.417644448812209e-05,
+      "loss": 1.1189,
+      "step": 34135
+    },
+    {
+      "epoch": 10.21,
+      "grad_norm": 1.823405385017395,
+      "learning_rate": 2.4170572488965427e-05,
+      "loss": 1.0416,
+      "step": 34140
+    },
+    {
+      "epoch": 10.22,
+      "grad_norm": 2.36574125289917,
+      "learning_rate": 2.4164700535617174e-05,
+      "loss": 0.9979,
+      "step": 34145
+    },
+    {
+      "epoch": 10.22,
+      "grad_norm": 1.0665093660354614,
+      "learning_rate": 2.415882862840165e-05,
+      "loss": 0.937,
+      "step": 34150
+    },
+    {
+      "epoch": 10.22,
+      "grad_norm": 3.2020561695098877,
+      "learning_rate": 2.4152956767643138e-05,
+      "loss": 1.0047,
+      "step": 34155
+    },
+    {
+      "epoch": 10.22,
+      "grad_norm": 3.9763476848602295,
+      "learning_rate": 2.4147084953665953e-05,
+      "loss": 0.9047,
+      "step": 34160
+    },
+    {
+      "epoch": 10.22,
+      "grad_norm": 5.038437843322754,
+      "learning_rate": 2.4141213186794378e-05,
+      "loss": 0.9305,
+      "step": 34165
+    },
+    {
+      "epoch": 10.22,
+      "grad_norm": 2.591571092605591,
+      "learning_rate": 2.4135341467352697e-05,
+      "loss": 1.1079,
+      "step": 34170
+    },
+    {
+      "epoch": 10.22,
+      "grad_norm": 2.8258914947509766,
+      "learning_rate": 2.4129469795665215e-05,
+      "loss": 0.9622,
+      "step": 34175
+    },
+    {
+      "epoch": 10.23,
+      "grad_norm": 1.6246439218521118,
+      "learning_rate": 2.4123598172056205e-05,
+      "loss": 1.0708,
+      "step": 34180
+    },
+    {
+      "epoch": 10.23,
+      "grad_norm": 3.791229724884033,
+      "learning_rate": 2.4117726596849964e-05,
+      "loss": 1.185,
+      "step": 34185
+    },
+    {
+      "epoch": 10.23,
+      "grad_norm": 2.0742027759552,
+      "learning_rate": 2.411185507037077e-05,
+      "loss": 1.1087,
+      "step": 34190
+    },
+    {
+      "epoch": 10.23,
+      "grad_norm": 2.5888705253601074,
+      "learning_rate": 2.4105983592942886e-05,
+      "loss": 1.1562,
+      "step": 34195
+    },
+    {
+      "epoch": 10.23,
+      "grad_norm": 1.4645191431045532,
+      "learning_rate": 2.410011216489061e-05,
+      "loss": 1.0233,
+      "step": 34200
+    },
+    {
+      "epoch": 10.23,
+      "grad_norm": 5.13542366027832,
+      "learning_rate": 2.409424078653819e-05,
+      "loss": 1.0835,
+      "step": 34205
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 1.8826115131378174,
+      "learning_rate": 2.4088369458209916e-05,
+      "loss": 1.0035,
+      "step": 34210
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 2.7958590984344482,
+      "learning_rate": 2.408249818023005e-05,
+      "loss": 1.0768,
+      "step": 34215
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 1.63248872756958,
+      "learning_rate": 2.4076626952922857e-05,
+      "loss": 1.0479,
+      "step": 34220
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 3.1579835414886475,
+      "learning_rate": 2.4070755776612604e-05,
+      "loss": 1.1089,
+      "step": 34225
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 1.6841520071029663,
+      "learning_rate": 2.4064884651623527e-05,
+      "loss": 1.3328,
+      "step": 34230
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 1.0290015935897827,
+      "learning_rate": 2.4059013578279917e-05,
+      "loss": 1.1065,
+      "step": 34235
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 2.263608932495117,
+      "learning_rate": 2.4053142556905992e-05,
+      "loss": 1.0712,
+      "step": 34240
+    },
+    {
+      "epoch": 10.25,
+      "grad_norm": 3.4659013748168945,
+      "learning_rate": 2.4047271587826032e-05,
+      "loss": 0.9181,
+      "step": 34245
+    },
+    {
+      "epoch": 10.25,
+      "grad_norm": 3.7479326725006104,
+      "learning_rate": 2.404140067136426e-05,
+      "loss": 0.7427,
+      "step": 34250
+    },
+    {
+      "epoch": 10.25,
+      "grad_norm": 1.718619704246521,
+      "learning_rate": 2.4035529807844944e-05,
+      "loss": 1.0278,
+      "step": 34255
+    },
+    {
+      "epoch": 10.25,
+      "grad_norm": 1.7770065069198608,
+      "learning_rate": 2.4029658997592315e-05,
+      "loss": 1.1875,
+      "step": 34260
+    },
+    {
+      "epoch": 10.25,
+      "grad_norm": 5.215729236602783,
+      "learning_rate": 2.4023788240930603e-05,
+      "loss": 1.0645,
+      "step": 34265
+    },
+    {
+      "epoch": 10.25,
+      "grad_norm": 3.1575255393981934,
+      "learning_rate": 2.401791753818406e-05,
+      "loss": 1.1738,
+      "step": 34270
+    },
+    {
+      "epoch": 10.25,
+      "grad_norm": 1.1542024612426758,
+      "learning_rate": 2.4012046889676898e-05,
+      "loss": 1.002,
+      "step": 34275
+    },
+    {
+      "epoch": 10.26,
+      "grad_norm": 2.203456163406372,
+      "learning_rate": 2.400617629573337e-05,
+      "loss": 0.8584,
+      "step": 34280
+    },
+    {
+      "epoch": 10.26,
+      "grad_norm": 3.4563515186309814,
+      "learning_rate": 2.4000305756677685e-05,
+      "loss": 1.1988,
+      "step": 34285
+    },
+    {
+      "epoch": 10.26,
+      "grad_norm": 5.644460678100586,
+      "learning_rate": 2.399443527283408e-05,
+      "loss": 1.0035,
+      "step": 34290
+    },
+    {
+      "epoch": 10.26,
+      "grad_norm": 3.3462517261505127,
+      "learning_rate": 2.398856484452677e-05,
+      "loss": 1.2118,
+      "step": 34295
+    },
+    {
+      "epoch": 10.26,
+      "grad_norm": 2.617570161819458,
+      "learning_rate": 2.398269447207997e-05,
+      "loss": 1.1375,
+      "step": 34300
+    },
+    {
+      "epoch": 10.26,
+      "grad_norm": 3.606656789779663,
+      "learning_rate": 2.3976824155817894e-05,
+      "loss": 1.0468,
+      "step": 34305
+    },
+    {
+      "epoch": 10.27,
+      "grad_norm": 2.8307979106903076,
+      "learning_rate": 2.397095389606476e-05,
+      "loss": 1.089,
+      "step": 34310
+    },
+    {
+      "epoch": 10.27,
+      "grad_norm": 1.1224695444107056,
+      "learning_rate": 2.3965083693144773e-05,
+      "loss": 1.0908,
+      "step": 34315
+    },
+    {
+      "epoch": 10.27,
+      "grad_norm": 3.3736870288848877,
+      "learning_rate": 2.3959213547382138e-05,
+      "loss": 1.0837,
+      "step": 34320
+    },
+    {
+      "epoch": 10.27,
+      "grad_norm": 1.475428581237793,
+      "learning_rate": 2.395334345910105e-05,
+      "loss": 1.1278,
+      "step": 34325
+    },
+    {
+      "epoch": 10.27,
+      "grad_norm": 2.2437570095062256,
+      "learning_rate": 2.394747342862573e-05,
+      "loss": 1.1207,
+      "step": 34330
+    },
+    {
+      "epoch": 10.27,
+      "grad_norm": 1.757214903831482,
+      "learning_rate": 2.394160345628034e-05,
+      "loss": 1.0901,
+      "step": 34335
+    },
+    {
+      "epoch": 10.27,
+      "grad_norm": 1.8947856426239014,
+      "learning_rate": 2.3935733542389103e-05,
+      "loss": 1.164,
+      "step": 34340
+    },
+    {
+      "epoch": 10.28,
+      "grad_norm": 2.8285508155822754,
+      "learning_rate": 2.392986368727619e-05,
+      "loss": 0.9759,
+      "step": 34345
+    },
+    {
+      "epoch": 10.28,
+      "grad_norm": 1.154876470565796,
+      "learning_rate": 2.39239938912658e-05,
+      "loss": 1.1223,
+      "step": 34350
+    },
+    {
+      "epoch": 10.28,
+      "grad_norm": 2.4530434608459473,
+      "learning_rate": 2.391812415468211e-05,
+      "loss": 1.0409,
+      "step": 34355
+    },
+    {
+      "epoch": 10.28,
+      "grad_norm": 1.6551940441131592,
+      "learning_rate": 2.3912254477849286e-05,
+      "loss": 1.067,
+      "step": 34360
+    },
+    {
+      "epoch": 10.28,
+      "grad_norm": 2.834213972091675,
+      "learning_rate": 2.390638486109153e-05,
+      "loss": 0.9237,
+      "step": 34365
+    },
+    {
+      "epoch": 10.28,
+      "grad_norm": 3.8012025356292725,
+      "learning_rate": 2.390051530473299e-05,
+      "loss": 1.1377,
+      "step": 34370
+    },
+    {
+      "epoch": 10.28,
+      "grad_norm": 2.4281139373779297,
+      "learning_rate": 2.3894645809097858e-05,
+      "loss": 0.9931,
+      "step": 34375
+    },
+    {
+      "epoch": 10.29,
+      "grad_norm": 2.0580687522888184,
+      "learning_rate": 2.3888776374510273e-05,
+      "loss": 1.0118,
+      "step": 34380
+    },
+    {
+      "epoch": 10.29,
+      "grad_norm": 1.2598625421524048,
+      "learning_rate": 2.3882907001294433e-05,
+      "loss": 0.9545,
+      "step": 34385
+    },
+    {
+      "epoch": 10.29,
+      "grad_norm": 3.062026023864746,
+      "learning_rate": 2.3877037689774467e-05,
+      "loss": 1.0679,
+      "step": 34390
+    },
+    {
+      "epoch": 10.29,
+      "grad_norm": 3.9732203483581543,
+      "learning_rate": 2.3871168440274545e-05,
+      "loss": 1.1264,
+      "step": 34395
+    },
+    {
+      "epoch": 10.29,
+      "grad_norm": 8.932393074035645,
+      "learning_rate": 2.3865299253118816e-05,
+      "loss": 0.9523,
+      "step": 34400
+    },
+    {
+      "epoch": 10.29,
+      "grad_norm": 2.8433444499969482,
+      "learning_rate": 2.385943012863143e-05,
+      "loss": 0.9684,
+      "step": 34405
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 3.345172643661499,
+      "learning_rate": 2.385356106713653e-05,
+      "loss": 0.9367,
+      "step": 34410
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 16.16388702392578,
+      "learning_rate": 2.384769206895827e-05,
+      "loss": 0.9319,
+      "step": 34415
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 1.3967313766479492,
+      "learning_rate": 2.3841823134420767e-05,
+      "loss": 1.1586,
+      "step": 34420
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 3.0620460510253906,
+      "learning_rate": 2.3835954263848176e-05,
+      "loss": 1.1258,
+      "step": 34425
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 2.4952540397644043,
+      "learning_rate": 2.3830085457564613e-05,
+      "loss": 1.0679,
+      "step": 34430
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 2.0433554649353027,
+      "learning_rate": 2.3824216715894224e-05,
+      "loss": 1.2555,
+      "step": 34435
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 3.4012880325317383,
+      "learning_rate": 2.3818348039161115e-05,
+      "loss": 0.8585,
+      "step": 34440
+    },
+    {
+      "epoch": 10.31,
+      "grad_norm": 1.1092032194137573,
+      "learning_rate": 2.3812479427689424e-05,
+      "loss": 1.1172,
+      "step": 34445
+    },
+    {
+      "epoch": 10.31,
+      "grad_norm": 1.4429662227630615,
+      "learning_rate": 2.380661088180326e-05,
+      "loss": 1.2767,
+      "step": 34450
+    },
+    {
+      "epoch": 10.31,
+      "grad_norm": 8.600884437561035,
+      "learning_rate": 2.3800742401826727e-05,
+      "loss": 1.0851,
+      "step": 34455
+    },
+    {
+      "epoch": 10.31,
+      "grad_norm": 2.5165319442749023,
+      "learning_rate": 2.3794873988083954e-05,
+      "loss": 0.9119,
+      "step": 34460
+    },
+    {
+      "epoch": 10.31,
+      "grad_norm": 4.288184642791748,
+      "learning_rate": 2.378900564089903e-05,
+      "loss": 1.1302,
+      "step": 34465
+    },
+    {
+      "epoch": 10.31,
+      "grad_norm": 2.89193058013916,
+      "learning_rate": 2.3783137360596075e-05,
+      "loss": 1.1722,
+      "step": 34470
+    },
+    {
+      "epoch": 10.31,
+      "grad_norm": 4.590530872344971,
+      "learning_rate": 2.3777269147499165e-05,
+      "loss": 0.8362,
+      "step": 34475
+    },
+    {
+      "epoch": 10.32,
+      "grad_norm": 4.36327600479126,
+      "learning_rate": 2.377140100193242e-05,
+      "loss": 1.0983,
+      "step": 34480
+    },
+    {
+      "epoch": 10.32,
+      "grad_norm": 2.2678892612457275,
+      "learning_rate": 2.376553292421992e-05,
+      "loss": 1.1464,
+      "step": 34485
+    },
+    {
+      "epoch": 10.32,
+      "grad_norm": 3.2285358905792236,
+      "learning_rate": 2.3759664914685754e-05,
+      "loss": 1.1098,
+      "step": 34490
+    },
+    {
+      "epoch": 10.32,
+      "grad_norm": 4.26018762588501,
+      "learning_rate": 2.3753796973654e-05,
+      "loss": 1.0358,
+      "step": 34495
+    },
+    {
+      "epoch": 10.32,
+      "grad_norm": 1.6850954294204712,
+      "learning_rate": 2.374792910144874e-05,
+      "loss": 1.0134,
+      "step": 34500
+    },
+    {
+      "epoch": 10.32,
+      "grad_norm": 1.3236955404281616,
+      "learning_rate": 2.3742061298394065e-05,
+      "loss": 1.0113,
+      "step": 34505
+    },
+    {
+      "epoch": 10.32,
+      "grad_norm": 5.23323917388916,
+      "learning_rate": 2.373619356481403e-05,
+      "loss": 1.097,
+      "step": 34510
+    },
+    {
+      "epoch": 10.33,
+      "grad_norm": 1.9136748313903809,
+      "learning_rate": 2.3730325901032718e-05,
+      "loss": 1.0261,
+      "step": 34515
+    },
+    {
+      "epoch": 10.33,
+      "grad_norm": 4.667322635650635,
+      "learning_rate": 2.3724458307374187e-05,
+      "loss": 1.0971,
+      "step": 34520
+    },
+    {
+      "epoch": 10.33,
+      "grad_norm": 1.810530185699463,
+      "learning_rate": 2.3718590784162485e-05,
+      "loss": 1.0513,
+      "step": 34525
+    },
+    {
+      "epoch": 10.33,
+      "grad_norm": 5.116344451904297,
+      "learning_rate": 2.3712723331721698e-05,
+      "loss": 1.0967,
+      "step": 34530
+    },
+    {
+      "epoch": 10.33,
+      "grad_norm": 2.3734960556030273,
+      "learning_rate": 2.370685595037585e-05,
+      "loss": 1.1215,
+      "step": 34535
+    },
+    {
+      "epoch": 10.33,
+      "grad_norm": 1.920328974723816,
+      "learning_rate": 2.3700988640449014e-05,
+      "loss": 1.1294,
+      "step": 34540
+    },
+    {
+      "epoch": 10.34,
+      "grad_norm": 3.0604352951049805,
+      "learning_rate": 2.3695121402265224e-05,
+      "loss": 1.0668,
+      "step": 34545
+    },
+    {
+      "epoch": 10.34,
+      "grad_norm": 3.079564094543457,
+      "learning_rate": 2.3689254236148514e-05,
+      "loss": 0.9018,
+      "step": 34550
+    },
+    {
+      "epoch": 10.34,
+      "grad_norm": 2.8750357627868652,
+      "learning_rate": 2.368338714242294e-05,
+      "loss": 1.0537,
+      "step": 34555
+    },
+    {
+      "epoch": 10.34,
+      "grad_norm": 4.040964603424072,
+      "learning_rate": 2.3677520121412516e-05,
+      "loss": 0.8928,
+      "step": 34560
+    },
+    {
+      "epoch": 10.34,
+      "grad_norm": 4.141225814819336,
+      "learning_rate": 2.3671653173441292e-05,
+      "loss": 1.1755,
+      "step": 34565
+    },
+    {
+      "epoch": 10.34,
+      "grad_norm": 4.6155195236206055,
+      "learning_rate": 2.3665786298833266e-05,
+      "loss": 1.0279,
+      "step": 34570
+    },
+    {
+      "epoch": 10.34,
+      "grad_norm": 4.095025539398193,
+      "learning_rate": 2.3659919497912488e-05,
+      "loss": 1.0721,
+      "step": 34575
+    },
+    {
+      "epoch": 10.35,
+      "grad_norm": 2.4246129989624023,
+      "learning_rate": 2.3654052771002965e-05,
+      "loss": 1.0305,
+      "step": 34580
+    },
+    {
+      "epoch": 10.35,
+      "grad_norm": 1.700454592704773,
+      "learning_rate": 2.36481861184287e-05,
+      "loss": 1.3431,
+      "step": 34585
+    },
+    {
+      "epoch": 10.35,
+      "grad_norm": 2.0215277671813965,
+      "learning_rate": 2.364231954051372e-05,
+      "loss": 1.0706,
+      "step": 34590
+    },
+    {
+      "epoch": 10.35,
+      "grad_norm": 3.362088203430176,
+      "learning_rate": 2.363645303758201e-05,
+      "loss": 1.0435,
+      "step": 34595
+    },
+    {
+      "epoch": 10.35,
+      "grad_norm": 1.618765115737915,
+      "learning_rate": 2.3630586609957592e-05,
+      "loss": 1.0702,
+      "step": 34600
+    },
+    {
+      "epoch": 10.35,
+      "grad_norm": 2.0355236530303955,
+      "learning_rate": 2.3624720257964442e-05,
+      "loss": 1.0517,
+      "step": 34605
+    },
+    {
+      "epoch": 10.35,
+      "grad_norm": 4.993839263916016,
+      "learning_rate": 2.3618853981926573e-05,
+      "loss": 0.96,
+      "step": 34610
+    },
+    {
+      "epoch": 10.36,
+      "grad_norm": 2.558202028274536,
+      "learning_rate": 2.3612987782167964e-05,
+      "loss": 1.053,
+      "step": 34615
+    },
+    {
+      "epoch": 10.36,
+      "grad_norm": 1.7264127731323242,
+      "learning_rate": 2.3607121659012586e-05,
+      "loss": 1.0745,
+      "step": 34620
+    },
+    {
+      "epoch": 10.36,
+      "grad_norm": 6.849349021911621,
+      "learning_rate": 2.360125561278444e-05,
+      "loss": 1.0707,
+      "step": 34625
+    },
+    {
+      "epoch": 10.36,
+      "grad_norm": 2.509521245956421,
+      "learning_rate": 2.3595389643807488e-05,
+      "loss": 1.0879,
+      "step": 34630
+    },
+    {
+      "epoch": 10.36,
+      "grad_norm": 1.6693092584609985,
+      "learning_rate": 2.358952375240571e-05,
+      "loss": 0.9827,
+      "step": 34635
+    },
+    {
+      "epoch": 10.36,
+      "grad_norm": 10.258462905883789,
+      "learning_rate": 2.3583657938903057e-05,
+      "loss": 1.1256,
+      "step": 34640
+    },
+    {
+      "epoch": 10.37,
+      "grad_norm": 3.046783685684204,
+      "learning_rate": 2.3577792203623523e-05,
+      "loss": 1.0576,
+      "step": 34645
+    },
+    {
+      "epoch": 10.37,
+      "grad_norm": 4.163492202758789,
+      "learning_rate": 2.3571926546891042e-05,
+      "loss": 0.9902,
+      "step": 34650
+    },
+    {
+      "epoch": 10.37,
+      "grad_norm": 1.5861027240753174,
+      "learning_rate": 2.3566060969029563e-05,
+      "loss": 1.195,
+      "step": 34655
+    },
+    {
+      "epoch": 10.37,
+      "grad_norm": 2.5667498111724854,
+      "learning_rate": 2.356019547036305e-05,
+      "loss": 1.0898,
+      "step": 34660
+    },
+    {
+      "epoch": 10.37,
+      "grad_norm": 1.8456099033355713,
+      "learning_rate": 2.355433005121545e-05,
+      "loss": 0.9962,
+      "step": 34665
+    },
+    {
+      "epoch": 10.37,
+      "grad_norm": 3.5785045623779297,
+      "learning_rate": 2.3548464711910692e-05,
+      "loss": 1.1786,
+      "step": 34670
+    },
+    {
+      "epoch": 10.37,
+      "grad_norm": 2.0253660678863525,
+      "learning_rate": 2.354259945277273e-05,
+      "loss": 1.1772,
+      "step": 34675
+    },
+    {
+      "epoch": 10.38,
+      "grad_norm": 1.426957368850708,
+      "learning_rate": 2.353673427412547e-05,
+      "loss": 0.9349,
+      "step": 34680
+    },
+    {
+      "epoch": 10.38,
+      "grad_norm": 2.7187154293060303,
+      "learning_rate": 2.353086917629287e-05,
+      "loss": 1.0479,
+      "step": 34685
+    },
+    {
+      "epoch": 10.38,
+      "grad_norm": 7.200928211212158,
+      "learning_rate": 2.3525004159598822e-05,
+      "loss": 0.9289,
+      "step": 34690
+    },
+    {
+      "epoch": 10.38,
+      "grad_norm": 4.092981338500977,
+      "learning_rate": 2.3519139224367278e-05,
+      "loss": 0.9506,
+      "step": 34695
+    },
+    {
+      "epoch": 10.38,
+      "grad_norm": 2.0632410049438477,
+      "learning_rate": 2.351327437092212e-05,
+      "loss": 1.054,
+      "step": 34700
+    },
+    {
+      "epoch": 10.38,
+      "grad_norm": 2.467764377593994,
+      "learning_rate": 2.3507409599587287e-05,
+      "loss": 1.2235,
+      "step": 34705
+    },
+    {
+      "epoch": 10.38,
+      "grad_norm": 2.497389793395996,
+      "learning_rate": 2.350154491068667e-05,
+      "loss": 0.9353,
+      "step": 34710
+    },
+    {
+      "epoch": 10.39,
+      "grad_norm": 1.8514834642410278,
+      "learning_rate": 2.349568030454416e-05,
+      "loss": 1.1288,
+      "step": 34715
+    },
+    {
+      "epoch": 10.39,
+      "grad_norm": 1.1936990022659302,
+      "learning_rate": 2.348981578148367e-05,
+      "loss": 1.1242,
+      "step": 34720
+    },
+    {
+      "epoch": 10.39,
+      "grad_norm": 2.6927096843719482,
+      "learning_rate": 2.3483951341829077e-05,
+      "loss": 1.1503,
+      "step": 34725
+    },
+    {
+      "epoch": 10.39,
+      "grad_norm": 5.187263011932373,
+      "learning_rate": 2.3478086985904287e-05,
+      "loss": 1.0652,
+      "step": 34730
+    },
+    {
+      "epoch": 10.39,
+      "grad_norm": 3.8386361598968506,
+      "learning_rate": 2.3472222714033157e-05,
+      "loss": 1.0557,
+      "step": 34735
+    },
+    {
+      "epoch": 10.39,
+      "grad_norm": 1.5286484956741333,
+      "learning_rate": 2.346635852653959e-05,
+      "loss": 1.0264,
+      "step": 34740
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 2.502507209777832,
+      "learning_rate": 2.3460494423747443e-05,
+      "loss": 1.0286,
+      "step": 34745
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 2.476358652114868,
+      "learning_rate": 2.345463040598059e-05,
+      "loss": 0.9376,
+      "step": 34750
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 2.4512181282043457,
+      "learning_rate": 2.3448766473562892e-05,
+      "loss": 1.0785,
+      "step": 34755
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 3.614104986190796,
+      "learning_rate": 2.344290262681821e-05,
+      "loss": 1.0775,
+      "step": 34760
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 3.8133108615875244,
+      "learning_rate": 2.3437038866070396e-05,
+      "loss": 0.9405,
+      "step": 34765
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 3.1824207305908203,
+      "learning_rate": 2.3431175191643307e-05,
+      "loss": 0.924,
+      "step": 34770
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 4.146546840667725,
+      "learning_rate": 2.342531160386077e-05,
+      "loss": 1.0573,
+      "step": 34775
+    },
+    {
+      "epoch": 10.41,
+      "grad_norm": 3.6027698516845703,
+      "learning_rate": 2.341944810304665e-05,
+      "loss": 1.087,
+      "step": 34780
+    },
+    {
+      "epoch": 10.41,
+      "grad_norm": 4.211222171783447,
+      "learning_rate": 2.3413584689524753e-05,
+      "loss": 1.0654,
+      "step": 34785
+    },
+    {
+      "epoch": 10.41,
+      "grad_norm": 3.0907680988311768,
+      "learning_rate": 2.3407721363618935e-05,
+      "loss": 1.1095,
+      "step": 34790
+    },
+    {
+      "epoch": 10.41,
+      "grad_norm": 3.797797679901123,
+      "learning_rate": 2.3401858125653006e-05,
+      "loss": 0.9867,
+      "step": 34795
+    },
+    {
+      "epoch": 10.41,
+      "grad_norm": 3.5584876537323,
+      "learning_rate": 2.33959949759508e-05,
+      "loss": 1.0036,
+      "step": 34800
+    },
+    {
+      "epoch": 10.41,
+      "grad_norm": 2.268932580947876,
+      "learning_rate": 2.3390131914836122e-05,
+      "loss": 1.1396,
+      "step": 34805
+    },
+    {
+      "epoch": 10.41,
+      "grad_norm": 1.9871071577072144,
+      "learning_rate": 2.3384268942632777e-05,
+      "loss": 1.049,
+      "step": 34810
+    },
+    {
+      "epoch": 10.42,
+      "grad_norm": 5.160571098327637,
+      "learning_rate": 2.3378406059664587e-05,
+      "loss": 0.8987,
+      "step": 34815
+    },
+    {
+      "epoch": 10.42,
+      "grad_norm": 3.7017822265625,
+      "learning_rate": 2.3372543266255335e-05,
+      "loss": 1.0438,
+      "step": 34820
+    },
+    {
+      "epoch": 10.42,
+      "grad_norm": 2.0446062088012695,
+      "learning_rate": 2.336668056272884e-05,
+      "loss": 1.1254,
+      "step": 34825
+    },
+    {
+      "epoch": 10.42,
+      "grad_norm": 3.492748260498047,
+      "learning_rate": 2.3360817949408864e-05,
+      "loss": 0.9717,
+      "step": 34830
+    },
+    {
+      "epoch": 10.42,
+      "grad_norm": 4.558749198913574,
+      "learning_rate": 2.3354955426619222e-05,
+      "loss": 1.0145,
+      "step": 34835
+    },
+    {
+      "epoch": 10.42,
+      "grad_norm": 2.7081844806671143,
+      "learning_rate": 2.3349092994683676e-05,
+      "loss": 1.0516,
+      "step": 34840
+    },
+    {
+      "epoch": 10.43,
+      "grad_norm": NaN,
+      "learning_rate": 2.334440311476777e-05,
+      "loss": 0.975,
+      "step": 34845
+    },
+    {
+      "epoch": 10.43,
+      "grad_norm": 3.3851234912872314,
+      "learning_rate": 2.3338540847185518e-05,
+      "loss": 1.0871,
+      "step": 34850
+    },
+    {
+      "epoch": 10.43,
+      "grad_norm": 1.13603675365448,
+      "learning_rate": 2.3332678671363932e-05,
+      "loss": 1.0211,
+      "step": 34855
+    },
+    {
+      "epoch": 10.43,
+      "grad_norm": 1.5273590087890625,
+      "learning_rate": 2.3326816587626775e-05,
+      "loss": 1.0605,
+      "step": 34860
+    },
+    {
+      "epoch": 10.43,
+      "grad_norm": 2.7484490871429443,
+      "learning_rate": 2.3320954596297788e-05,
+      "loss": 1.193,
+      "step": 34865
+    },
+    {
+      "epoch": 10.43,
+      "grad_norm": 6.367726802825928,
+      "learning_rate": 2.3315092697700742e-05,
+      "loss": 1.1871,
+      "step": 34870
+    },
+    {
+      "epoch": 10.43,
+      "grad_norm": 2.484698534011841,
+      "learning_rate": 2.3309230892159365e-05,
+      "loss": 1.1412,
+      "step": 34875
+    },
+    {
+      "epoch": 10.44,
+      "grad_norm": 2.0732154846191406,
+      "learning_rate": 2.3303369179997418e-05,
+      "loss": 1.114,
+      "step": 34880
+    },
+    {
+      "epoch": 10.44,
+      "grad_norm": 2.8750739097595215,
+      "learning_rate": 2.329750756153862e-05,
+      "loss": 1.0455,
+      "step": 34885
+    },
+    {
+      "epoch": 10.44,
+      "grad_norm": 1.7856959104537964,
+      "learning_rate": 2.329164603710672e-05,
+      "loss": 1.0768,
+      "step": 34890
+    },
+    {
+      "epoch": 10.44,
+      "grad_norm": 8.524049758911133,
+      "learning_rate": 2.328578460702543e-05,
+      "loss": 1.1066,
+      "step": 34895
+    },
+    {
+      "epoch": 10.44,
+      "grad_norm": 3.2621653079986572,
+      "learning_rate": 2.3279923271618465e-05,
+      "loss": 1.3073,
+      "step": 34900
+    },
+    {
+      "epoch": 10.44,
+      "grad_norm": 1.9567291736602783,
+      "learning_rate": 2.3274062031209563e-05,
+      "loss": 1.0333,
+      "step": 34905
+    },
+    {
+      "epoch": 10.44,
+      "grad_norm": 3.069105386734009,
+      "learning_rate": 2.326820088612241e-05,
+      "loss": 1.0295,
+      "step": 34910
+    },
+    {
+      "epoch": 10.45,
+      "grad_norm": 1.8950694799423218,
+      "learning_rate": 2.3262339836680727e-05,
+      "loss": 1.0614,
+      "step": 34915
+    },
+    {
+      "epoch": 10.45,
+      "grad_norm": 3.2250144481658936,
+      "learning_rate": 2.3256478883208206e-05,
+      "loss": 1.0405,
+      "step": 34920
+    },
+    {
+      "epoch": 10.45,
+      "grad_norm": 4.422776699066162,
+      "learning_rate": 2.3250618026028543e-05,
+      "loss": 1.1056,
+      "step": 34925
+    },
+    {
+      "epoch": 10.45,
+      "grad_norm": 1.4227927923202515,
+      "learning_rate": 2.3244757265465435e-05,
+      "loss": 1.0868,
+      "step": 34930
+    },
+    {
+      "epoch": 10.45,
+      "grad_norm": 2.4411497116088867,
+      "learning_rate": 2.3238896601842548e-05,
+      "loss": 0.9554,
+      "step": 34935
+    },
+    {
+      "epoch": 10.45,
+      "grad_norm": 2.067897081375122,
+      "learning_rate": 2.3233036035483587e-05,
+      "loss": 1.1577,
+      "step": 34940
+    },
+    {
+      "epoch": 10.46,
+      "grad_norm": 2.7568440437316895,
+      "learning_rate": 2.322717556671219e-05,
+      "loss": 1.0572,
+      "step": 34945
+    },
+    {
+      "epoch": 10.46,
+      "grad_norm": 4.667929649353027,
+      "learning_rate": 2.3221315195852058e-05,
+      "loss": 0.9588,
+      "step": 34950
+    },
+    {
+      "epoch": 10.46,
+      "grad_norm": 2.4807698726654053,
+      "learning_rate": 2.321545492322684e-05,
+      "loss": 1.0147,
+      "step": 34955
+    },
+    {
+      "epoch": 10.46,
+      "grad_norm": 15.608848571777344,
+      "learning_rate": 2.320959474916018e-05,
+      "loss": 1.2589,
+      "step": 34960
+    },
+    {
+      "epoch": 10.46,
+      "grad_norm": 2.8121330738067627,
+      "learning_rate": 2.3203734673975753e-05,
+      "loss": 1.0937,
+      "step": 34965
+    },
+    {
+      "epoch": 10.46,
+      "grad_norm": 2.748537540435791,
+      "learning_rate": 2.319787469799718e-05,
+      "loss": 0.9264,
+      "step": 34970
+    },
+    {
+      "epoch": 10.46,
+      "grad_norm": 3.233227014541626,
+      "learning_rate": 2.3192014821548127e-05,
+      "loss": 1.2414,
+      "step": 34975
+    },
+    {
+      "epoch": 10.47,
+      "grad_norm": 1.753865361213684,
+      "learning_rate": 2.3186155044952203e-05,
+      "loss": 1.0152,
+      "step": 34980
+    },
+    {
+      "epoch": 10.47,
+      "grad_norm": 3.4801855087280273,
+      "learning_rate": 2.3180295368533063e-05,
+      "loss": 0.9391,
+      "step": 34985
+    },
+    {
+      "epoch": 10.47,
+      "grad_norm": 2.416815757751465,
+      "learning_rate": 2.3174435792614318e-05,
+      "loss": 1.2012,
+      "step": 34990
+    },
+    {
+      "epoch": 10.47,
+      "grad_norm": 8.600541114807129,
+      "learning_rate": 2.316857631751958e-05,
+      "loss": 1.1342,
+      "step": 34995
+    },
+    {
+      "epoch": 10.47,
+      "grad_norm": 3.0248448848724365,
+      "learning_rate": 2.3162716943572465e-05,
+      "loss": 1.0246,
+      "step": 35000
+    },
+    {
+      "epoch": 10.47,
+      "grad_norm": 2.2905187606811523,
+      "learning_rate": 2.3156857671096592e-05,
+      "loss": 0.9875,
+      "step": 35005
+    },
+    {
+      "epoch": 10.47,
+      "grad_norm": 5.242259979248047,
+      "learning_rate": 2.3150998500415546e-05,
+      "loss": 0.9466,
+      "step": 35010
+    },
+    {
+      "epoch": 10.48,
+      "grad_norm": 2.770148992538452,
+      "learning_rate": 2.3145139431852934e-05,
+      "loss": 1.0718,
+      "step": 35015
+    },
+    {
+      "epoch": 10.48,
+      "grad_norm": 5.146824836730957,
+      "learning_rate": 2.3139280465732348e-05,
+      "loss": 1.1159,
+      "step": 35020
+    },
+    {
+      "epoch": 10.48,
+      "grad_norm": 1.9731031656265259,
+      "learning_rate": 2.3133421602377366e-05,
+      "loss": 0.8097,
+      "step": 35025
+    },
+    {
+      "epoch": 10.48,
+      "grad_norm": 1.8398514986038208,
+      "learning_rate": 2.3127562842111565e-05,
+      "loss": 1.1045,
+      "step": 35030
+    },
+    {
+      "epoch": 10.48,
+      "grad_norm": 3.5004403591156006,
+      "learning_rate": 2.3121704185258527e-05,
+      "loss": 1.0801,
+      "step": 35035
+    },
+    {
+      "epoch": 10.48,
+      "grad_norm": 3.4678022861480713,
+      "learning_rate": 2.3115845632141806e-05,
+      "loss": 1.1383,
+      "step": 35040
+    },
+    {
+      "epoch": 10.49,
+      "grad_norm": 2.484443426132202,
+      "learning_rate": 2.310998718308498e-05,
+      "loss": 1.0761,
+      "step": 35045
+    },
+    {
+      "epoch": 10.49,
+      "grad_norm": 3.0486910343170166,
+      "learning_rate": 2.31041288384116e-05,
+      "loss": 1.2576,
+      "step": 35050
+    },
+    {
+      "epoch": 10.49,
+      "grad_norm": 2.7550971508026123,
+      "learning_rate": 2.3098270598445203e-05,
+      "loss": 0.8973,
+      "step": 35055
+    },
+    {
+      "epoch": 10.49,
+      "grad_norm": 1.8515464067459106,
+      "learning_rate": 2.3092412463509357e-05,
+      "loss": 1.0277,
+      "step": 35060
+    },
+    {
+      "epoch": 10.49,
+      "grad_norm": 1.7236946821212769,
+      "learning_rate": 2.3086554433927573e-05,
+      "loss": 1.1393,
+      "step": 35065
+    },
+    {
+      "epoch": 10.49,
+      "grad_norm": 3.1678953170776367,
+      "learning_rate": 2.308069651002341e-05,
+      "loss": 1.0378,
+      "step": 35070
+    },
+    {
+      "epoch": 10.49,
+      "grad_norm": 1.7023390531539917,
+      "learning_rate": 2.3074838692120378e-05,
+      "loss": 1.0801,
+      "step": 35075
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 3.274125814437866,
+      "learning_rate": 2.306898098054201e-05,
+      "loss": 1.027,
+      "step": 35080
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 4.002904415130615,
+      "learning_rate": 2.306312337561181e-05,
+      "loss": 0.9389,
+      "step": 35085
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 3.656179904937744,
+      "learning_rate": 2.305726587765329e-05,
+      "loss": 1.0733,
+      "step": 35090
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 5.379166126251221,
+      "learning_rate": 2.305140848698996e-05,
+      "loss": 1.0789,
+      "step": 35095
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 1.2176295518875122,
+      "learning_rate": 2.3045551203945314e-05,
+      "loss": 1.0208,
+      "step": 35100
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 2.7093119621276855,
+      "learning_rate": 2.3039694028842847e-05,
+      "loss": 0.9681,
+      "step": 35105
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 9.889016151428223,
+      "learning_rate": 2.3033836962006033e-05,
+      "loss": 0.8443,
+      "step": 35110
+    },
+    {
+      "epoch": 10.51,
+      "grad_norm": 4.702127456665039,
+      "learning_rate": 2.3027980003758366e-05,
+      "loss": 0.9901,
+      "step": 35115
+    },
+    {
+      "epoch": 10.51,
+      "grad_norm": 5.113590240478516,
+      "learning_rate": 2.3022123154423316e-05,
+      "loss": 1.1291,
+      "step": 35120
+    },
+    {
+      "epoch": 10.51,
+      "grad_norm": 1.5303608179092407,
+      "learning_rate": 2.301626641432434e-05,
+      "loss": 1.1448,
+      "step": 35125
+    },
+    {
+      "epoch": 10.51,
+      "grad_norm": 2.7038862705230713,
+      "learning_rate": 2.3010409783784913e-05,
+      "loss": 1.049,
+      "step": 35130
+    },
+    {
+      "epoch": 10.51,
+      "grad_norm": 3.5487146377563477,
+      "learning_rate": 2.3004553263128483e-05,
+      "loss": 1.0263,
+      "step": 35135
+    },
+    {
+      "epoch": 10.51,
+      "grad_norm": 1.9222893714904785,
+      "learning_rate": 2.299869685267851e-05,
+      "loss": 1.0709,
+      "step": 35140
+    },
+    {
+      "epoch": 10.51,
+      "grad_norm": 3.621490716934204,
+      "learning_rate": 2.2992840552758428e-05,
+      "loss": 1.1768,
+      "step": 35145
+    },
+    {
+      "epoch": 10.52,
+      "grad_norm": 1.7279472351074219,
+      "learning_rate": 2.2986984363691663e-05,
+      "loss": 1.0183,
+      "step": 35150
+    },
+    {
+      "epoch": 10.52,
+      "grad_norm": 1.8752796649932861,
+      "learning_rate": 2.2981128285801672e-05,
+      "loss": 1.0267,
+      "step": 35155
+    },
+    {
+      "epoch": 10.52,
+      "grad_norm": 2.7973175048828125,
+      "learning_rate": 2.297527231941186e-05,
+      "loss": 1.1971,
+      "step": 35160
+    },
+    {
+      "epoch": 10.52,
+      "grad_norm": 1.7248725891113281,
+      "learning_rate": 2.296941646484566e-05,
+      "loss": 0.9569,
+      "step": 35165
+    },
+    {
+      "epoch": 10.52,
+      "grad_norm": 2.4975392818450928,
+      "learning_rate": 2.2963560722426468e-05,
+      "loss": 1.111,
+      "step": 35170
+    },
+    {
+      "epoch": 10.52,
+      "grad_norm": 3.2511773109436035,
+      "learning_rate": 2.295770509247771e-05,
+      "loss": 1.066,
+      "step": 35175
+    },
+    {
+      "epoch": 10.53,
+      "grad_norm": 2.9692437648773193,
+      "learning_rate": 2.2951849575322772e-05,
+      "loss": 0.99,
+      "step": 35180
+    },
+    {
+      "epoch": 10.53,
+      "grad_norm": 2.788581371307373,
+      "learning_rate": 2.2945994171285058e-05,
+      "loss": 1.1557,
+      "step": 35185
+    },
+    {
+      "epoch": 10.53,
+      "grad_norm": 3.816265106201172,
+      "learning_rate": 2.2940138880687946e-05,
+      "loss": 1.0066,
+      "step": 35190
+    },
+    {
+      "epoch": 10.53,
+      "grad_norm": 5.253087520599365,
+      "learning_rate": 2.2934283703854823e-05,
+      "loss": 0.9963,
+      "step": 35195
+    },
+    {
+      "epoch": 10.53,
+      "grad_norm": 2.2299044132232666,
+      "learning_rate": 2.2928428641109065e-05,
+      "loss": 1.1243,
+      "step": 35200
+    },
+    {
+      "epoch": 10.53,
+      "grad_norm": 4.349503993988037,
+      "learning_rate": 2.2922573692774034e-05,
+      "loss": 1.1135,
+      "step": 35205
+    },
+    {
+      "epoch": 10.53,
+      "grad_norm": 1.9277504682540894,
+      "learning_rate": 2.2916718859173108e-05,
+      "loss": 1.0192,
+      "step": 35210
+    },
+    {
+      "epoch": 10.54,
+      "grad_norm": 2.2060022354125977,
+      "learning_rate": 2.2910864140629634e-05,
+      "loss": 1.1672,
+      "step": 35215
+    },
+    {
+      "epoch": 10.54,
+      "grad_norm": 3.025515556335449,
+      "learning_rate": 2.2905009537466955e-05,
+      "loss": 0.9311,
+      "step": 35220
+    },
+    {
+      "epoch": 10.54,
+      "grad_norm": 3.500619649887085,
+      "learning_rate": 2.2899155050008428e-05,
+      "loss": 1.1652,
+      "step": 35225
+    },
+    {
+      "epoch": 10.54,
+      "grad_norm": 8.94033145904541,
+      "learning_rate": 2.2893300678577376e-05,
+      "loss": 1.251,
+      "step": 35230
+    },
+    {
+      "epoch": 10.54,
+      "grad_norm": 3.2193968296051025,
+      "learning_rate": 2.2887446423497147e-05,
+      "loss": 0.9508,
+      "step": 35235
+    },
+    {
+      "epoch": 10.54,
+      "grad_norm": 3.0045528411865234,
+      "learning_rate": 2.2881592285091045e-05,
+      "loss": 0.9492,
+      "step": 35240
+    },
+    {
+      "epoch": 10.54,
+      "grad_norm": 3.3813045024871826,
+      "learning_rate": 2.2875738263682413e-05,
+      "loss": 0.9394,
+      "step": 35245
+    },
+    {
+      "epoch": 10.55,
+      "grad_norm": 1.6937832832336426,
+      "learning_rate": 2.2869884359594545e-05,
+      "loss": 1.1169,
+      "step": 35250
+    },
+    {
+      "epoch": 10.55,
+      "grad_norm": 1.0461424589157104,
+      "learning_rate": 2.2864030573150738e-05,
+      "loss": 1.2219,
+      "step": 35255
+    },
+    {
+      "epoch": 10.55,
+      "grad_norm": 2.934732675552368,
+      "learning_rate": 2.2858176904674317e-05,
+      "loss": 1.14,
+      "step": 35260
+    },
+    {
+      "epoch": 10.55,
+      "grad_norm": 4.22146463394165,
+      "learning_rate": 2.2852323354488548e-05,
+      "loss": 1.103,
+      "step": 35265
+    },
+    {
+      "epoch": 10.55,
+      "grad_norm": 3.387679100036621,
+      "learning_rate": 2.2846469922916736e-05,
+      "loss": 1.0281,
+      "step": 35270
+    },
+    {
+      "epoch": 10.55,
+      "grad_norm": 2.5129520893096924,
+      "learning_rate": 2.284061661028215e-05,
+      "loss": 0.9338,
+      "step": 35275
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 1.3689956665039062,
+      "learning_rate": 2.2834763416908057e-05,
+      "loss": 1.1322,
+      "step": 35280
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 1.68153977394104,
+      "learning_rate": 2.2828910343117734e-05,
+      "loss": 1.1103,
+      "step": 35285
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 3.7177228927612305,
+      "learning_rate": 2.2823057389234432e-05,
+      "loss": 1.1346,
+      "step": 35290
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 1.6719677448272705,
+      "learning_rate": 2.281720455558142e-05,
+      "loss": 1.0287,
+      "step": 35295
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 2.740884780883789,
+      "learning_rate": 2.2811351842481916e-05,
+      "loss": 1.0737,
+      "step": 35300
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 3.8541574478149414,
+      "learning_rate": 2.280549925025919e-05,
+      "loss": 1.045,
+      "step": 35305
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 3.7161831855773926,
+      "learning_rate": 2.2799646779236454e-05,
+      "loss": 1.1965,
+      "step": 35310
+    },
+    {
+      "epoch": 10.57,
+      "grad_norm": 4.742947578430176,
+      "learning_rate": 2.2793794429736933e-05,
+      "loss": 1.0853,
+      "step": 35315
+    },
+    {
+      "epoch": 10.57,
+      "grad_norm": 1.507947564125061,
+      "learning_rate": 2.278794220208386e-05,
+      "loss": 0.9959,
+      "step": 35320
+    },
+    {
+      "epoch": 10.57,
+      "grad_norm": 1.6198567152023315,
+      "learning_rate": 2.2782090096600434e-05,
+      "loss": 1.0109,
+      "step": 35325
+    },
+    {
+      "epoch": 10.57,
+      "grad_norm": 1.7230312824249268,
+      "learning_rate": 2.277623811360987e-05,
+      "loss": 1.1351,
+      "step": 35330
+    },
+    {
+      "epoch": 10.57,
+      "grad_norm": 4.703332901000977,
+      "learning_rate": 2.2770386253435358e-05,
+      "loss": 1.0449,
+      "step": 35335
+    },
+    {
+      "epoch": 10.57,
+      "grad_norm": 2.2942113876342773,
+      "learning_rate": 2.2764534516400106e-05,
+      "loss": 1.1787,
+      "step": 35340
+    },
+    {
+      "epoch": 10.57,
+      "grad_norm": 4.3071513175964355,
+      "learning_rate": 2.275868290282729e-05,
+      "loss": 1.092,
+      "step": 35345
+    },
+    {
+      "epoch": 10.58,
+      "grad_norm": 2.7554285526275635,
+      "learning_rate": 2.2752831413040074e-05,
+      "loss": 1.1289,
+      "step": 35350
+    },
+    {
+      "epoch": 10.58,
+      "grad_norm": 2.5249757766723633,
+      "learning_rate": 2.2746980047361654e-05,
+      "loss": 1.0906,
+      "step": 35355
+    },
+    {
+      "epoch": 10.58,
+      "grad_norm": 2.719829559326172,
+      "learning_rate": 2.2741128806115176e-05,
+      "loss": 0.9871,
+      "step": 35360
+    },
+    {
+      "epoch": 10.58,
+      "grad_norm": 2.8020482063293457,
+      "learning_rate": 2.273527768962381e-05,
+      "loss": 1.1126,
+      "step": 35365
+    },
+    {
+      "epoch": 10.58,
+      "grad_norm": 2.6041548252105713,
+      "learning_rate": 2.2729426698210703e-05,
+      "loss": 1.1024,
+      "step": 35370
+    },
+    {
+      "epoch": 10.58,
+      "grad_norm": 3.7162394523620605,
+      "learning_rate": 2.2723575832198997e-05,
+      "loss": 0.9589,
+      "step": 35375
+    },
+    {
+      "epoch": 10.59,
+      "grad_norm": 25.685956954956055,
+      "learning_rate": 2.2717725091911843e-05,
+      "loss": 0.9937,
+      "step": 35380
+    },
+    {
+      "epoch": 10.59,
+      "grad_norm": 4.040263652801514,
+      "learning_rate": 2.2711874477672342e-05,
+      "loss": 0.9,
+      "step": 35385
+    },
+    {
+      "epoch": 10.59,
+      "grad_norm": 2.358093023300171,
+      "learning_rate": 2.2706023989803653e-05,
+      "loss": 1.068,
+      "step": 35390
+    },
+    {
+      "epoch": 10.59,
+      "grad_norm": 3.6727354526519775,
+      "learning_rate": 2.270017362862886e-05,
+      "loss": 0.9494,
+      "step": 35395
+    },
+    {
+      "epoch": 10.59,
+      "grad_norm": 2.7660393714904785,
+      "learning_rate": 2.2694323394471097e-05,
+      "loss": 0.9925,
+      "step": 35400
+    },
+    {
+      "epoch": 10.59,
+      "grad_norm": 1.4063373804092407,
+      "learning_rate": 2.2688473287653457e-05,
+      "loss": 1.2022,
+      "step": 35405
+    },
+    {
+      "epoch": 10.59,
+      "grad_norm": 2.7881920337677,
+      "learning_rate": 2.2682623308499023e-05,
+      "loss": 0.9778,
+      "step": 35410
+    },
+    {
+      "epoch": 10.6,
+      "grad_norm": 2.657111644744873,
+      "learning_rate": 2.267677345733091e-05,
+      "loss": 1.0118,
+      "step": 35415
+    },
+    {
+      "epoch": 10.6,
+      "grad_norm": 2.8737306594848633,
+      "learning_rate": 2.267092373447217e-05,
+      "loss": 0.9643,
+      "step": 35420
+    },
+    {
+      "epoch": 10.6,
+      "grad_norm": 3.328007459640503,
+      "learning_rate": 2.26650741402459e-05,
+      "loss": 1.0657,
+      "step": 35425
+    },
+    {
+      "epoch": 10.6,
+      "grad_norm": 4.150506019592285,
+      "learning_rate": 2.265922467497515e-05,
+      "loss": 1.0437,
+      "step": 35430
+    },
+    {
+      "epoch": 10.6,
+      "grad_norm": 1.4019544124603271,
+      "learning_rate": 2.2653375338983e-05,
+      "loss": 1.1049,
+      "step": 35435
+    },
+    {
+      "epoch": 10.6,
+      "grad_norm": 2.7139785289764404,
+      "learning_rate": 2.264752613259249e-05,
+      "loss": 1.0097,
+      "step": 35440
+    },
+    {
+      "epoch": 10.6,
+      "grad_norm": 1.9748849868774414,
+      "learning_rate": 2.2641677056126654e-05,
+      "loss": 1.1295,
+      "step": 35445
+    },
+    {
+      "epoch": 10.61,
+      "grad_norm": 3.615483283996582,
+      "learning_rate": 2.263582810990855e-05,
+      "loss": 1.0361,
+      "step": 35450
+    },
+    {
+      "epoch": 10.61,
+      "grad_norm": 2.8796679973602295,
+      "learning_rate": 2.26299792942612e-05,
+      "loss": 1.1339,
+      "step": 35455
+    },
+    {
+      "epoch": 10.61,
+      "grad_norm": 1.1939923763275146,
+      "learning_rate": 2.262413060950763e-05,
+      "loss": 1.0356,
+      "step": 35460
+    },
+    {
+      "epoch": 10.61,
+      "grad_norm": 7.1389994621276855,
+      "learning_rate": 2.261828205597086e-05,
+      "loss": 1.0418,
+      "step": 35465
+    },
+    {
+      "epoch": 10.61,
+      "grad_norm": 1.4644900560379028,
+      "learning_rate": 2.2612433633973896e-05,
+      "loss": 1.0545,
+      "step": 35470
+    },
+    {
+      "epoch": 10.61,
+      "grad_norm": 1.3660115003585815,
+      "learning_rate": 2.2606585343839744e-05,
+      "loss": 1.0775,
+      "step": 35475
+    },
+    {
+      "epoch": 10.62,
+      "grad_norm": 2.621532440185547,
+      "learning_rate": 2.2600737185891385e-05,
+      "loss": 1.1644,
+      "step": 35480
+    },
+    {
+      "epoch": 10.62,
+      "grad_norm": 6.380326271057129,
+      "learning_rate": 2.2594889160451828e-05,
+      "loss": 1.1348,
+      "step": 35485
+    },
+    {
+      "epoch": 10.62,
+      "grad_norm": 2.7322096824645996,
+      "learning_rate": 2.2589041267844034e-05,
+      "loss": 1.097,
+      "step": 35490
+    },
+    {
+      "epoch": 10.62,
+      "grad_norm": 3.6032140254974365,
+      "learning_rate": 2.2583193508390993e-05,
+      "loss": 1.0526,
+      "step": 35495
+    },
+    {
+      "epoch": 10.62,
+      "grad_norm": 2.900298833847046,
+      "learning_rate": 2.2577345882415663e-05,
+      "loss": 0.9409,
+      "step": 35500
+    },
+    {
+      "epoch": 10.62,
+      "grad_norm": 1.8747291564941406,
+      "learning_rate": 2.257149839024099e-05,
+      "loss": 1.1623,
+      "step": 35505
+    },
+    {
+      "epoch": 10.62,
+      "grad_norm": 2.2065136432647705,
+      "learning_rate": 2.2565651032189948e-05,
+      "loss": 0.9874,
+      "step": 35510
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 2.0098252296447754,
+      "learning_rate": 2.255980380858546e-05,
+      "loss": 1.1739,
+      "step": 35515
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 3.0822856426239014,
+      "learning_rate": 2.2553956719750483e-05,
+      "loss": 1.1192,
+      "step": 35520
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 2.1216416358947754,
+      "learning_rate": 2.254810976600792e-05,
+      "loss": 0.9649,
+      "step": 35525
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 1.3982808589935303,
+      "learning_rate": 2.254226294768072e-05,
+      "loss": 1.0809,
+      "step": 35530
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 1.7782407999038696,
+      "learning_rate": 2.2536416265091775e-05,
+      "loss": 1.1739,
+      "step": 35535
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 3.666782855987549,
+      "learning_rate": 2.2530569718563998e-05,
+      "loss": 1.1696,
+      "step": 35540
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 2.729365587234497,
+      "learning_rate": 2.252472330842029e-05,
+      "loss": 1.0824,
+      "step": 35545
+    },
+    {
+      "epoch": 10.64,
+      "grad_norm": 2.964341163635254,
+      "learning_rate": 2.251887703498354e-05,
+      "loss": 1.2195,
+      "step": 35550
+    },
+    {
+      "epoch": 10.64,
+      "grad_norm": 3.209306240081787,
+      "learning_rate": 2.2513030898576635e-05,
+      "loss": 1.1367,
+      "step": 35555
+    },
+    {
+      "epoch": 10.64,
+      "grad_norm": 3.5836918354034424,
+      "learning_rate": 2.2507184899522447e-05,
+      "loss": 1.142,
+      "step": 35560
+    },
+    {
+      "epoch": 10.64,
+      "grad_norm": 1.1011484861373901,
+      "learning_rate": 2.2501339038143843e-05,
+      "loss": 1.11,
+      "step": 35565
+    },
+    {
+      "epoch": 10.64,
+      "grad_norm": 1.9480361938476562,
+      "learning_rate": 2.2495493314763697e-05,
+      "loss": 1.0384,
+      "step": 35570
+    },
+    {
+      "epoch": 10.64,
+      "grad_norm": 1.9894143342971802,
+      "learning_rate": 2.2489647729704838e-05,
+      "loss": 1.0765,
+      "step": 35575
+    },
+    {
+      "epoch": 10.65,
+      "grad_norm": 2.3497085571289062,
+      "learning_rate": 2.2483802283290137e-05,
+      "loss": 1.0708,
+      "step": 35580
+    },
+    {
+      "epoch": 10.65,
+      "grad_norm": 2.3375086784362793,
+      "learning_rate": 2.2477956975842407e-05,
+      "loss": 1.041,
+      "step": 35585
+    },
+    {
+      "epoch": 10.65,
+      "grad_norm": 4.431266784667969,
+      "learning_rate": 2.2472111807684507e-05,
+      "loss": 1.0047,
+      "step": 35590
+    },
+    {
+      "epoch": 10.65,
+      "grad_norm": 6.045180797576904,
+      "learning_rate": 2.246626677913923e-05,
+      "loss": 1.0029,
+      "step": 35595
+    },
+    {
+      "epoch": 10.65,
+      "grad_norm": 2.7329964637756348,
+      "learning_rate": 2.2460421890529417e-05,
+      "loss": 1.0755,
+      "step": 35600
+    },
+    {
+      "epoch": 10.65,
+      "grad_norm": 3.588158130645752,
+      "learning_rate": 2.2454577142177865e-05,
+      "loss": 0.8856,
+      "step": 35605
+    },
+    {
+      "epoch": 10.65,
+      "grad_norm": 2.5146617889404297,
+      "learning_rate": 2.244873253440736e-05,
+      "loss": 1.2,
+      "step": 35610
+    },
+    {
+      "epoch": 10.66,
+      "grad_norm": 5.354367733001709,
+      "learning_rate": 2.2442888067540715e-05,
+      "loss": 1.0349,
+      "step": 35615
+    },
+    {
+      "epoch": 10.66,
+      "grad_norm": 1.597602367401123,
+      "learning_rate": 2.243704374190069e-05,
+      "loss": 1.1549,
+      "step": 35620
+    },
+    {
+      "epoch": 10.66,
+      "grad_norm": 1.7936272621154785,
+      "learning_rate": 2.2431199557810092e-05,
+      "loss": 1.0453,
+      "step": 35625
+    },
+    {
+      "epoch": 10.66,
+      "grad_norm": 1.9876726865768433,
+      "learning_rate": 2.2425355515591666e-05,
+      "loss": 1.0696,
+      "step": 35630
+    },
+    {
+      "epoch": 10.66,
+      "grad_norm": 3.0657265186309814,
+      "learning_rate": 2.241951161556818e-05,
+      "loss": 0.9361,
+      "step": 35635
+    },
+    {
+      "epoch": 10.66,
+      "grad_norm": 4.501842021942139,
+      "learning_rate": 2.241366785806238e-05,
+      "loss": 1.1468,
+      "step": 35640
+    },
+    {
+      "epoch": 10.66,
+      "grad_norm": 3.099733829498291,
+      "learning_rate": 2.240782424339702e-05,
+      "loss": 0.9717,
+      "step": 35645
+    },
+    {
+      "epoch": 10.67,
+      "grad_norm": 4.878641605377197,
+      "learning_rate": 2.2401980771894828e-05,
+      "loss": 0.9771,
+      "step": 35650
+    },
+    {
+      "epoch": 10.67,
+      "grad_norm": 3.61327862739563,
+      "learning_rate": 2.2396137443878534e-05,
+      "loss": 1.0749,
+      "step": 35655
+    },
+    {
+      "epoch": 10.67,
+      "grad_norm": 5.230576038360596,
+      "learning_rate": 2.2390294259670877e-05,
+      "loss": 1.1493,
+      "step": 35660
+    },
+    {
+      "epoch": 10.67,
+      "grad_norm": 1.4460947513580322,
+      "learning_rate": 2.238445121959455e-05,
+      "loss": 1.2674,
+      "step": 35665
+    },
+    {
+      "epoch": 10.67,
+      "grad_norm": 2.8116650581359863,
+      "learning_rate": 2.2378608323972255e-05,
+      "loss": 1.0535,
+      "step": 35670
+    },
+    {
+      "epoch": 10.67,
+      "grad_norm": 3.138775587081909,
+      "learning_rate": 2.2372765573126712e-05,
+      "loss": 0.8282,
+      "step": 35675
+    },
+    {
+      "epoch": 10.68,
+      "grad_norm": 2.2588982582092285,
+      "learning_rate": 2.236692296738058e-05,
+      "loss": 1.1188,
+      "step": 35680
+    },
+    {
+      "epoch": 10.68,
+      "grad_norm": 2.0676910877227783,
+      "learning_rate": 2.2361080507056565e-05,
+      "loss": 1.1425,
+      "step": 35685
+    },
+    {
+      "epoch": 10.68,
+      "grad_norm": 7.246330738067627,
+      "learning_rate": 2.2355238192477324e-05,
+      "loss": 0.9489,
+      "step": 35690
+    },
+    {
+      "epoch": 10.68,
+      "grad_norm": 3.171942710876465,
+      "learning_rate": 2.234939602396554e-05,
+      "loss": 1.1224,
+      "step": 35695
+    },
+    {
+      "epoch": 10.68,
+      "grad_norm": 2.2110137939453125,
+      "learning_rate": 2.2343554001843857e-05,
+      "loss": 1.1013,
+      "step": 35700
+    },
+    {
+      "epoch": 10.68,
+      "grad_norm": 2.53434157371521,
+      "learning_rate": 2.233771212643491e-05,
+      "loss": 0.9745,
+      "step": 35705
+    },
+    {
+      "epoch": 10.68,
+      "grad_norm": 2.9125607013702393,
+      "learning_rate": 2.2331870398061372e-05,
+      "loss": 1.3203,
+      "step": 35710
+    },
+    {
+      "epoch": 10.69,
+      "grad_norm": 3.6007089614868164,
+      "learning_rate": 2.2326028817045842e-05,
+      "loss": 1.0354,
+      "step": 35715
+    },
+    {
+      "epoch": 10.69,
+      "grad_norm": 4.191182613372803,
+      "learning_rate": 2.2320187383710978e-05,
+      "loss": 1.0592,
+      "step": 35720
+    },
+    {
+      "epoch": 10.69,
+      "grad_norm": 6.260267734527588,
+      "learning_rate": 2.2314346098379367e-05,
+      "loss": 1.1301,
+      "step": 35725
+    },
+    {
+      "epoch": 10.69,
+      "grad_norm": 2.3227128982543945,
+      "learning_rate": 2.230850496137363e-05,
+      "loss": 0.998,
+      "step": 35730
+    },
+    {
+      "epoch": 10.69,
+      "grad_norm": 2.7904467582702637,
+      "learning_rate": 2.2302663973016374e-05,
+      "loss": 1.0603,
+      "step": 35735
+    },
+    {
+      "epoch": 10.69,
+      "grad_norm": 5.104443550109863,
+      "learning_rate": 2.2296823133630174e-05,
+      "loss": 0.9873,
+      "step": 35740
+    },
+    {
+      "epoch": 10.69,
+      "grad_norm": 5.6274094581604,
+      "learning_rate": 2.2290982443537633e-05,
+      "loss": 0.9872,
+      "step": 35745
+    },
+    {
+      "epoch": 10.7,
+      "grad_norm": 1.1891549825668335,
+      "learning_rate": 2.2285141903061304e-05,
+      "loss": 1.0641,
+      "step": 35750
+    },
+    {
+      "epoch": 10.7,
+      "grad_norm": 2.298877000808716,
+      "learning_rate": 2.2279301512523778e-05,
+      "loss": 1.1714,
+      "step": 35755
+    },
+    {
+      "epoch": 10.7,
+      "grad_norm": 2.9934897422790527,
+      "learning_rate": 2.22734612722476e-05,
+      "loss": 0.8837,
+      "step": 35760
+    },
+    {
+      "epoch": 10.7,
+      "grad_norm": 3.266824722290039,
+      "learning_rate": 2.2267621182555313e-05,
+      "loss": 0.9443,
+      "step": 35765
+    },
+    {
+      "epoch": 10.7,
+      "grad_norm": 6.283466815948486,
+      "learning_rate": 2.2261781243769478e-05,
+      "loss": 1.1074,
+      "step": 35770
+    },
+    {
+      "epoch": 10.7,
+      "grad_norm": 1.5854196548461914,
+      "learning_rate": 2.2255941456212606e-05,
+      "loss": 0.9736,
+      "step": 35775
+    },
+    {
+      "epoch": 10.7,
+      "grad_norm": 4.49138069152832,
+      "learning_rate": 2.2250101820207246e-05,
+      "loss": 1.1243,
+      "step": 35780
+    },
+    {
+      "epoch": 10.71,
+      "grad_norm": 3.611423969268799,
+      "learning_rate": 2.2244262336075896e-05,
+      "loss": 1.1206,
+      "step": 35785
+    },
+    {
+      "epoch": 10.71,
+      "grad_norm": 2.668792247772217,
+      "learning_rate": 2.2238423004141083e-05,
+      "loss": 1.3554,
+      "step": 35790
+    },
+    {
+      "epoch": 10.71,
+      "grad_norm": 1.8896316289901733,
+      "learning_rate": 2.2232583824725296e-05,
+      "loss": 1.036,
+      "step": 35795
+    },
+    {
+      "epoch": 10.71,
+      "grad_norm": 2.4683837890625,
+      "learning_rate": 2.2226744798151017e-05,
+      "loss": 1.2378,
+      "step": 35800
+    },
+    {
+      "epoch": 10.71,
+      "grad_norm": 19.00712776184082,
+      "learning_rate": 2.2220905924740752e-05,
+      "loss": 0.9128,
+      "step": 35805
+    },
+    {
+      "epoch": 10.71,
+      "grad_norm": 3.103806972503662,
+      "learning_rate": 2.221506720481695e-05,
+      "loss": 1.1218,
+      "step": 35810
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 5.2492756843566895,
+      "learning_rate": 2.2209228638702098e-05,
+      "loss": 1.0806,
+      "step": 35815
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 2.3563156127929688,
+      "learning_rate": 2.2203390226718652e-05,
+      "loss": 1.1438,
+      "step": 35820
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 1.5582448244094849,
+      "learning_rate": 2.2197551969189052e-05,
+      "loss": 1.194,
+      "step": 35825
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 3.394693613052368,
+      "learning_rate": 2.219171386643575e-05,
+      "loss": 1.1377,
+      "step": 35830
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 2.7966670989990234,
+      "learning_rate": 2.2185875918781163e-05,
+      "loss": 1.0573,
+      "step": 35835
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 12.529913902282715,
+      "learning_rate": 2.218003812654773e-05,
+      "loss": 0.8754,
+      "step": 35840
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 4.540731430053711,
+      "learning_rate": 2.2174200490057856e-05,
+      "loss": 0.9227,
+      "step": 35845
+    },
+    {
+      "epoch": 10.73,
+      "grad_norm": 2.8158795833587646,
+      "learning_rate": 2.2168363009633958e-05,
+      "loss": 0.9375,
+      "step": 35850
+    },
+    {
+      "epoch": 10.73,
+      "grad_norm": 3.6659469604492188,
+      "learning_rate": 2.216252568559843e-05,
+      "loss": 1.2302,
+      "step": 35855
+    },
+    {
+      "epoch": 10.73,
+      "grad_norm": 2.3665411472320557,
+      "learning_rate": 2.2156688518273648e-05,
+      "loss": 0.997,
+      "step": 35860
+    },
+    {
+      "epoch": 10.73,
+      "grad_norm": 3.4320127964019775,
+      "learning_rate": 2.2150851507982014e-05,
+      "loss": 1.1298,
+      "step": 35865
+    },
+    {
+      "epoch": 10.73,
+      "grad_norm": 2.856066942214966,
+      "learning_rate": 2.2145014655045876e-05,
+      "loss": 1.2628,
+      "step": 35870
+    },
+    {
+      "epoch": 10.73,
+      "grad_norm": 2.302837610244751,
+      "learning_rate": 2.2139177959787625e-05,
+      "loss": 0.997,
+      "step": 35875
+    },
+    {
+      "epoch": 10.73,
+      "grad_norm": 3.4538989067077637,
+      "learning_rate": 2.213334142252959e-05,
+      "loss": 0.9814,
+      "step": 35880
+    },
+    {
+      "epoch": 10.74,
+      "grad_norm": 1.6373672485351562,
+      "learning_rate": 2.212750504359414e-05,
+      "loss": 1.069,
+      "step": 35885
+    },
+    {
+      "epoch": 10.74,
+      "grad_norm": 1.9250200986862183,
+      "learning_rate": 2.2121668823303595e-05,
+      "loss": 1.0603,
+      "step": 35890
+    },
+    {
+      "epoch": 10.74,
+      "grad_norm": 2.419745445251465,
+      "learning_rate": 2.211583276198029e-05,
+      "loss": 1.1043,
+      "step": 35895
+    },
+    {
+      "epoch": 10.74,
+      "grad_norm": 1.965515375137329,
+      "learning_rate": 2.210999685994654e-05,
+      "loss": 0.9786,
+      "step": 35900
+    },
+    {
+      "epoch": 10.74,
+      "grad_norm": 2.6983680725097656,
+      "learning_rate": 2.2104161117524664e-05,
+      "loss": 0.9368,
+      "step": 35905
+    },
+    {
+      "epoch": 10.74,
+      "grad_norm": 2.2785370349884033,
+      "learning_rate": 2.2098325535036957e-05,
+      "loss": 1.1098,
+      "step": 35910
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 2.5998971462249756,
+      "learning_rate": 2.2092490112805715e-05,
+      "loss": 0.9755,
+      "step": 35915
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 2.66782546043396,
+      "learning_rate": 2.208665485115322e-05,
+      "loss": 1.2246,
+      "step": 35920
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 2.6380605697631836,
+      "learning_rate": 2.2080819750401757e-05,
+      "loss": 1.1043,
+      "step": 35925
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 1.2431594133377075,
+      "learning_rate": 2.2074984810873572e-05,
+      "loss": 0.9766,
+      "step": 35930
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 1.7657440900802612,
+      "learning_rate": 2.2069150032890952e-05,
+      "loss": 1.0666,
+      "step": 35935
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 1.1442104578018188,
+      "learning_rate": 2.2064482327034192e-05,
+      "loss": 0.958,
+      "step": 35940
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 4.242776393890381,
+      "learning_rate": 2.2058647840645615e-05,
+      "loss": 1.0115,
+      "step": 35945
+    },
+    {
+      "epoch": 10.76,
+      "grad_norm": 2.7217624187469482,
+      "learning_rate": 2.2052813516704852e-05,
+      "loss": 1.0799,
+      "step": 35950
+    },
+    {
+      "epoch": 10.76,
+      "grad_norm": 2.8261914253234863,
+      "learning_rate": 2.204697935553414e-05,
+      "loss": 1.1698,
+      "step": 35955
+    },
+    {
+      "epoch": 10.76,
+      "grad_norm": 1.2735272645950317,
+      "learning_rate": 2.2041145357455684e-05,
+      "loss": 0.9582,
+      "step": 35960
+    },
+    {
+      "epoch": 10.76,
+      "grad_norm": 2.095346212387085,
+      "learning_rate": 2.2035311522791705e-05,
+      "loss": 1.0745,
+      "step": 35965
+    },
+    {
+      "epoch": 10.76,
+      "grad_norm": 3.746208906173706,
+      "learning_rate": 2.2029477851864376e-05,
+      "loss": 0.9504,
+      "step": 35970
+    },
+    {
+      "epoch": 10.76,
+      "grad_norm": 32.70151138305664,
+      "learning_rate": 2.2023644344995907e-05,
+      "loss": 1.1023,
+      "step": 35975
+    },
+    {
+      "epoch": 10.76,
+      "grad_norm": 2.321796417236328,
+      "learning_rate": 2.2017811002508466e-05,
+      "loss": 1.1198,
+      "step": 35980
+    },
+    {
+      "epoch": 10.77,
+      "grad_norm": 2.214085102081299,
+      "learning_rate": 2.201197782472422e-05,
+      "loss": 1.0783,
+      "step": 35985
+    },
+    {
+      "epoch": 10.77,
+      "grad_norm": 2.9954833984375,
+      "learning_rate": 2.2006144811965336e-05,
+      "loss": 1.059,
+      "step": 35990
+    },
+    {
+      "epoch": 10.77,
+      "grad_norm": 1.0794235467910767,
+      "learning_rate": 2.2000311964553954e-05,
+      "loss": 1.1247,
+      "step": 35995
+    },
+    {
+      "epoch": 10.77,
+      "grad_norm": 5.492161750793457,
+      "learning_rate": 2.1994479282812236e-05,
+      "loss": 0.9969,
+      "step": 36000
+    },
+    {
+      "epoch": 10.77,
+      "grad_norm": 1.0977524518966675,
+      "learning_rate": 2.19886467670623e-05,
+      "loss": 0.9352,
+      "step": 36005
+    },
+    {
+      "epoch": 10.77,
+      "grad_norm": 2.7051050662994385,
+      "learning_rate": 2.1982814417626265e-05,
+      "loss": 0.7758,
+      "step": 36010
+    },
+    {
+      "epoch": 10.78,
+      "grad_norm": 2.0435142517089844,
+      "learning_rate": 2.1976982234826265e-05,
+      "loss": 1.0474,
+      "step": 36015
+    },
+    {
+      "epoch": 10.78,
+      "grad_norm": 3.4940507411956787,
+      "learning_rate": 2.197115021898438e-05,
+      "loss": 1.0514,
+      "step": 36020
+    },
+    {
+      "epoch": 10.78,
+      "grad_norm": 2.082819700241089,
+      "learning_rate": 2.1965318370422735e-05,
+      "loss": 1.0285,
+      "step": 36025
+    },
+    {
+      "epoch": 10.78,
+      "grad_norm": 2.880784511566162,
+      "learning_rate": 2.195948668946339e-05,
+      "loss": 1.2136,
+      "step": 36030
+    },
+    {
+      "epoch": 10.78,
+      "grad_norm": 3.054262161254883,
+      "learning_rate": 2.1953655176428445e-05,
+      "loss": 1.0397,
+      "step": 36035
+    },
+    {
+      "epoch": 10.78,
+      "grad_norm": 3.0019352436065674,
+      "learning_rate": 2.194782383163996e-05,
+      "loss": 1.11,
+      "step": 36040
+    },
+    {
+      "epoch": 10.78,
+      "grad_norm": 0.9828662276268005,
+      "learning_rate": 2.194199265541998e-05,
+      "loss": 1.0822,
+      "step": 36045
+    },
+    {
+      "epoch": 10.79,
+      "grad_norm": 3.027097702026367,
+      "learning_rate": 2.1936161648090575e-05,
+      "loss": 0.9884,
+      "step": 36050
+    },
+    {
+      "epoch": 10.79,
+      "grad_norm": 6.978221893310547,
+      "learning_rate": 2.193033080997377e-05,
+      "loss": 1.0059,
+      "step": 36055
+    },
+    {
+      "epoch": 10.79,
+      "grad_norm": 3.2999072074890137,
+      "learning_rate": 2.192450014139161e-05,
+      "loss": 1.2139,
+      "step": 36060
+    },
+    {
+      "epoch": 10.79,
+      "grad_norm": 2.107088804244995,
+      "learning_rate": 2.1918669642666106e-05,
+      "loss": 0.9479,
+      "step": 36065
+    },
+    {
+      "epoch": 10.79,
+      "grad_norm": 1.4788484573364258,
+      "learning_rate": 2.1912839314119276e-05,
+      "loss": 1.0025,
+      "step": 36070
+    },
+    {
+      "epoch": 10.79,
+      "grad_norm": 1.6675289869308472,
+      "learning_rate": 2.190700915607313e-05,
+      "loss": 1.1289,
+      "step": 36075
+    },
+    {
+      "epoch": 10.79,
+      "grad_norm": 4.08076286315918,
+      "learning_rate": 2.190117916884964e-05,
+      "loss": 1.1822,
+      "step": 36080
+    },
+    {
+      "epoch": 10.8,
+      "grad_norm": 9.089180946350098,
+      "learning_rate": 2.1895349352770816e-05,
+      "loss": 1.0968,
+      "step": 36085
+    },
+    {
+      "epoch": 10.8,
+      "grad_norm": 4.459921360015869,
+      "learning_rate": 2.188951970815861e-05,
+      "loss": 0.8567,
+      "step": 36090
+    },
+    {
+      "epoch": 10.8,
+      "grad_norm": 3.4034626483917236,
+      "learning_rate": 2.1883690235335004e-05,
+      "loss": 1.05,
+      "step": 36095
+    },
+    {
+      "epoch": 10.8,
+      "grad_norm": 4.486004829406738,
+      "learning_rate": 2.187786093462194e-05,
+      "loss": 1.04,
+      "step": 36100
+    },
+    {
+      "epoch": 10.8,
+      "grad_norm": 2.139437675476074,
+      "learning_rate": 2.187203180634138e-05,
+      "loss": 0.9213,
+      "step": 36105
+    },
+    {
+      "epoch": 10.8,
+      "grad_norm": 2.6758251190185547,
+      "learning_rate": 2.1866202850815254e-05,
+      "loss": 1.0313,
+      "step": 36110
+    },
+    {
+      "epoch": 10.81,
+      "grad_norm": 2.4901866912841797,
+      "learning_rate": 2.1860374068365473e-05,
+      "loss": 1.0219,
+      "step": 36115
+    },
+    {
+      "epoch": 10.81,
+      "grad_norm": 4.980169296264648,
+      "learning_rate": 2.1854545459313985e-05,
+      "loss": 1.1223,
+      "step": 36120
+    },
+    {
+      "epoch": 10.81,
+      "grad_norm": 2.8188419342041016,
+      "learning_rate": 2.1848717023982667e-05,
+      "loss": 1.0865,
+      "step": 36125
+    },
+    {
+      "epoch": 10.81,
+      "grad_norm": 6.227662086486816,
+      "learning_rate": 2.1842888762693444e-05,
+      "loss": 1.1196,
+      "step": 36130
+    },
+    {
+      "epoch": 10.81,
+      "grad_norm": 2.4889774322509766,
+      "learning_rate": 2.1837060675768196e-05,
+      "loss": 1.0085,
+      "step": 36135
+    },
+    {
+      "epoch": 10.81,
+      "grad_norm": 2.1352288722991943,
+      "learning_rate": 2.1831232763528783e-05,
+      "loss": 1.0865,
+      "step": 36140
+    },
+    {
+      "epoch": 10.81,
+      "grad_norm": 1.6987781524658203,
+      "learning_rate": 2.1825405026297103e-05,
+      "loss": 1.1085,
+      "step": 36145
+    },
+    {
+      "epoch": 10.82,
+      "grad_norm": 2.239659309387207,
+      "learning_rate": 2.1819577464394992e-05,
+      "loss": 1.2226,
+      "step": 36150
+    },
+    {
+      "epoch": 10.82,
+      "grad_norm": 2.6719276905059814,
+      "learning_rate": 2.181375007814432e-05,
+      "loss": 1.2125,
+      "step": 36155
+    },
+    {
+      "epoch": 10.82,
+      "grad_norm": 1.225275993347168,
+      "learning_rate": 2.180792286786692e-05,
+      "loss": 1.0033,
+      "step": 36160
+    },
+    {
+      "epoch": 10.82,
+      "grad_norm": 4.794998645782471,
+      "learning_rate": 2.180209583388462e-05,
+      "loss": 1.0673,
+      "step": 36165
+    },
+    {
+      "epoch": 10.82,
+      "grad_norm": 2.080230951309204,
+      "learning_rate": 2.179626897651925e-05,
+      "loss": 1.067,
+      "step": 36170
+    },
+    {
+      "epoch": 10.82,
+      "grad_norm": 1.528211236000061,
+      "learning_rate": 2.17904422960926e-05,
+      "loss": 1.1119,
+      "step": 36175
+    },
+    {
+      "epoch": 10.82,
+      "grad_norm": 4.195923328399658,
+      "learning_rate": 2.17846157929265e-05,
+      "loss": 1.2904,
+      "step": 36180
+    },
+    {
+      "epoch": 10.83,
+      "grad_norm": 3.0591914653778076,
+      "learning_rate": 2.1778789467342713e-05,
+      "loss": 1.1747,
+      "step": 36185
+    },
+    {
+      "epoch": 10.83,
+      "grad_norm": 9.022075653076172,
+      "learning_rate": 2.177296331966305e-05,
+      "loss": 1.0482,
+      "step": 36190
+    },
+    {
+      "epoch": 10.83,
+      "grad_norm": 3.723809003829956,
+      "learning_rate": 2.1767137350209253e-05,
+      "loss": 1.0859,
+      "step": 36195
+    },
+    {
+      "epoch": 10.83,
+      "grad_norm": 2.5620508193969727,
+      "learning_rate": 2.1761311559303117e-05,
+      "loss": 0.9736,
+      "step": 36200
+    },
+    {
+      "epoch": 10.83,
+      "grad_norm": 3.8153581619262695,
+      "learning_rate": 2.1755485947266375e-05,
+      "loss": 1.2724,
+      "step": 36205
+    },
+    {
+      "epoch": 10.83,
+      "grad_norm": 1.3778871297836304,
+      "learning_rate": 2.174966051442076e-05,
+      "loss": 1.1058,
+      "step": 36210
+    },
+    {
+      "epoch": 10.84,
+      "grad_norm": 3.073227643966675,
+      "learning_rate": 2.174383526108803e-05,
+      "loss": 1.0846,
+      "step": 36215
+    },
+    {
+      "epoch": 10.84,
+      "grad_norm": 1.9298650026321411,
+      "learning_rate": 2.1738010187589878e-05,
+      "loss": 0.9319,
+      "step": 36220
+    },
+    {
+      "epoch": 10.84,
+      "grad_norm": 1.1409422159194946,
+      "learning_rate": 2.1732185294248045e-05,
+      "loss": 1.1355,
+      "step": 36225
+    },
+    {
+      "epoch": 10.84,
+      "grad_norm": 2.380774974822998,
+      "learning_rate": 2.1726360581384218e-05,
+      "loss": 1.1526,
+      "step": 36230
+    },
+    {
+      "epoch": 10.84,
+      "grad_norm": 1.3581106662750244,
+      "learning_rate": 2.172053604932009e-05,
+      "loss": 1.0837,
+      "step": 36235
+    },
+    {
+      "epoch": 10.84,
+      "grad_norm": 4.142109394073486,
+      "learning_rate": 2.1714711698377348e-05,
+      "loss": 1.0323,
+      "step": 36240
+    },
+    {
+      "epoch": 10.84,
+      "grad_norm": 2.672729015350342,
+      "learning_rate": 2.1708887528877668e-05,
+      "loss": 1.0175,
+      "step": 36245
+    },
+    {
+      "epoch": 10.85,
+      "grad_norm": 1.8875662088394165,
+      "learning_rate": 2.1703063541142703e-05,
+      "loss": 1.191,
+      "step": 36250
+    },
+    {
+      "epoch": 10.85,
+      "grad_norm": 1.7792450189590454,
+      "learning_rate": 2.1697239735494117e-05,
+      "loss": 0.9895,
+      "step": 36255
+    },
+    {
+      "epoch": 10.85,
+      "grad_norm": 2.7931838035583496,
+      "learning_rate": 2.1691416112253547e-05,
+      "loss": 1.0206,
+      "step": 36260
+    },
+    {
+      "epoch": 10.85,
+      "grad_norm": 3.638197660446167,
+      "learning_rate": 2.1685592671742626e-05,
+      "loss": 1.1491,
+      "step": 36265
+    },
+    {
+      "epoch": 10.85,
+      "grad_norm": 1.3871898651123047,
+      "learning_rate": 2.167976941428297e-05,
+      "loss": 1.109,
+      "step": 36270
+    },
+    {
+      "epoch": 10.85,
+      "grad_norm": 2.8166041374206543,
+      "learning_rate": 2.167394634019621e-05,
+      "loss": 1.0765,
+      "step": 36275
+    },
+    {
+      "epoch": 10.85,
+      "grad_norm": 2.0707552433013916,
+      "learning_rate": 2.1668123449803924e-05,
+      "loss": 1.0585,
+      "step": 36280
+    },
+    {
+      "epoch": 10.86,
+      "grad_norm": 3.0661659240722656,
+      "learning_rate": 2.1662300743427727e-05,
+      "loss": 1.0604,
+      "step": 36285
+    },
+    {
+      "epoch": 10.86,
+      "grad_norm": 2.1380584239959717,
+      "learning_rate": 2.1656478221389183e-05,
+      "loss": 1.0168,
+      "step": 36290
+    },
+    {
+      "epoch": 10.86,
+      "grad_norm": 2.518028974533081,
+      "learning_rate": 2.1650655884009875e-05,
+      "loss": 1.2193,
+      "step": 36295
+    },
+    {
+      "epoch": 10.86,
+      "grad_norm": 1.8819745779037476,
+      "learning_rate": 2.1644833731611366e-05,
+      "loss": 0.9329,
+      "step": 36300
+    },
+    {
+      "epoch": 10.86,
+      "grad_norm": 3.2668797969818115,
+      "learning_rate": 2.163901176451519e-05,
+      "loss": 0.9986,
+      "step": 36305
+    },
+    {
+      "epoch": 10.86,
+      "grad_norm": 2.280186653137207,
+      "learning_rate": 2.1633189983042916e-05,
+      "loss": 0.8617,
+      "step": 36310
+    },
+    {
+      "epoch": 10.87,
+      "grad_norm": 1.871981143951416,
+      "learning_rate": 2.1627368387516043e-05,
+      "loss": 0.9601,
+      "step": 36315
+    },
+    {
+      "epoch": 10.87,
+      "grad_norm": 2.4408321380615234,
+      "learning_rate": 2.162154697825612e-05,
+      "loss": 1.0367,
+      "step": 36320
+    },
+    {
+      "epoch": 10.87,
+      "grad_norm": 8.324357986450195,
+      "learning_rate": 2.1615725755584636e-05,
+      "loss": 1.1349,
+      "step": 36325
+    },
+    {
+      "epoch": 10.87,
+      "grad_norm": 3.0855600833892822,
+      "learning_rate": 2.1609904719823116e-05,
+      "loss": 1.0233,
+      "step": 36330
+    },
+    {
+      "epoch": 10.87,
+      "grad_norm": 1.6718279123306274,
+      "learning_rate": 2.1604083871293023e-05,
+      "loss": 1.1736,
+      "step": 36335
+    },
+    {
+      "epoch": 10.87,
+      "grad_norm": 2.869089365005493,
+      "learning_rate": 2.159826321031585e-05,
+      "loss": 1.0822,
+      "step": 36340
+    },
+    {
+      "epoch": 10.87,
+      "grad_norm": 6.643579483032227,
+      "learning_rate": 2.1592442737213064e-05,
+      "loss": 0.9564,
+      "step": 36345
+    },
+    {
+      "epoch": 10.88,
+      "grad_norm": 1.4432473182678223,
+      "learning_rate": 2.1586622452306125e-05,
+      "loss": 1.0499,
+      "step": 36350
+    },
+    {
+      "epoch": 10.88,
+      "grad_norm": 3.8422656059265137,
+      "learning_rate": 2.158080235591648e-05,
+      "loss": 0.7816,
+      "step": 36355
+    },
+    {
+      "epoch": 10.88,
+      "grad_norm": 2.3019320964813232,
+      "learning_rate": 2.1574982448365568e-05,
+      "loss": 1.1791,
+      "step": 36360
+    },
+    {
+      "epoch": 10.88,
+      "grad_norm": 2.5562939643859863,
+      "learning_rate": 2.1569162729974806e-05,
+      "loss": 1.1605,
+      "step": 36365
+    },
+    {
+      "epoch": 10.88,
+      "grad_norm": 6.145001411437988,
+      "learning_rate": 2.1563343201065633e-05,
+      "loss": 1.1362,
+      "step": 36370
+    },
+    {
+      "epoch": 10.88,
+      "grad_norm": 3.0975069999694824,
+      "learning_rate": 2.1557523861959434e-05,
+      "loss": 0.8223,
+      "step": 36375
+    },
+    {
+      "epoch": 10.88,
+      "grad_norm": 3.5439906120300293,
+      "learning_rate": 2.1551704712977623e-05,
+      "loss": 1.0676,
+      "step": 36380
+    },
+    {
+      "epoch": 10.89,
+      "grad_norm": 1.0710362195968628,
+      "learning_rate": 2.1545885754441562e-05,
+      "loss": 1.1736,
+      "step": 36385
+    },
+    {
+      "epoch": 10.89,
+      "grad_norm": 8.471335411071777,
+      "learning_rate": 2.1540066986672655e-05,
+      "loss": 0.8974,
+      "step": 36390
+    },
+    {
+      "epoch": 10.89,
+      "grad_norm": 3.0269336700439453,
+      "learning_rate": 2.153424840999225e-05,
+      "loss": 1.0499,
+      "step": 36395
+    },
+    {
+      "epoch": 10.89,
+      "grad_norm": 3.6303861141204834,
+      "learning_rate": 2.1528430024721694e-05,
+      "loss": 1.0851,
+      "step": 36400
+    },
+    {
+      "epoch": 10.89,
+      "grad_norm": 4.8882293701171875,
+      "learning_rate": 2.1522611831182348e-05,
+      "loss": 0.8858,
+      "step": 36405
+    },
+    {
+      "epoch": 10.89,
+      "grad_norm": 2.3631794452667236,
+      "learning_rate": 2.1516793829695526e-05,
+      "loss": 1.1782,
+      "step": 36410
+    },
+    {
+      "epoch": 10.89,
+      "grad_norm": 3.9514851570129395,
+      "learning_rate": 2.151097602058257e-05,
+      "loss": 1.195,
+      "step": 36415
+    },
+    {
+      "epoch": 10.9,
+      "grad_norm": 2.633768320083618,
+      "learning_rate": 2.1505158404164773e-05,
+      "loss": 1.1729,
+      "step": 36420
+    },
+    {
+      "epoch": 10.9,
+      "grad_norm": 6.990433692932129,
+      "learning_rate": 2.1499340980763456e-05,
+      "loss": 0.9894,
+      "step": 36425
+    },
+    {
+      "epoch": 10.9,
+      "grad_norm": 2.7082440853118896,
+      "learning_rate": 2.149352375069989e-05,
+      "loss": 1.0238,
+      "step": 36430
+    },
+    {
+      "epoch": 10.9,
+      "grad_norm": 3.5102274417877197,
+      "learning_rate": 2.1487706714295357e-05,
+      "loss": 0.9571,
+      "step": 36435
+    },
+    {
+      "epoch": 10.9,
+      "grad_norm": 2.7834293842315674,
+      "learning_rate": 2.1481889871871143e-05,
+      "loss": 1.1934,
+      "step": 36440
+    },
+    {
+      "epoch": 10.9,
+      "grad_norm": 1.7707253694534302,
+      "learning_rate": 2.1476073223748485e-05,
+      "loss": 1.0634,
+      "step": 36445
+    },
+    {
+      "epoch": 10.91,
+      "grad_norm": 1.9651144742965698,
+      "learning_rate": 2.147025677024865e-05,
+      "loss": 1.3354,
+      "step": 36450
+    },
+    {
+      "epoch": 10.91,
+      "grad_norm": 4.304943084716797,
+      "learning_rate": 2.146444051169285e-05,
+      "loss": 0.9292,
+      "step": 36455
+    },
+    {
+      "epoch": 10.91,
+      "grad_norm": 3.199054479598999,
+      "learning_rate": 2.1458624448402343e-05,
+      "loss": 0.9946,
+      "step": 36460
+    },
+    {
+      "epoch": 10.91,
+      "grad_norm": 1.4956828355789185,
+      "learning_rate": 2.1452808580698325e-05,
+      "loss": 0.9661,
+      "step": 36465
+    },
+    {
+      "epoch": 10.91,
+      "grad_norm": 1.1398528814315796,
+      "learning_rate": 2.144699290890199e-05,
+      "loss": 1.0639,
+      "step": 36470
+    },
+    {
+      "epoch": 10.91,
+      "grad_norm": 7.341798782348633,
+      "learning_rate": 2.1441177433334552e-05,
+      "loss": 0.8947,
+      "step": 36475
+    },
+    {
+      "epoch": 10.91,
+      "grad_norm": 3.4453506469726562,
+      "learning_rate": 2.1435362154317176e-05,
+      "loss": 1.0649,
+      "step": 36480
+    },
+    {
+      "epoch": 10.92,
+      "grad_norm": 0.8624611496925354,
+      "learning_rate": 2.1429547072171057e-05,
+      "loss": 1.1476,
+      "step": 36485
+    },
+    {
+      "epoch": 10.92,
+      "grad_norm": 5.505244255065918,
+      "learning_rate": 2.1423732187217338e-05,
+      "loss": 0.9781,
+      "step": 36490
+    },
+    {
+      "epoch": 10.92,
+      "grad_norm": 3.857640504837036,
+      "learning_rate": 2.1417917499777164e-05,
+      "loss": 1.0794,
+      "step": 36495
+    },
+    {
+      "epoch": 10.92,
+      "grad_norm": 3.2706289291381836,
+      "learning_rate": 2.1412103010171693e-05,
+      "loss": 1.0661,
+      "step": 36500
+    },
+    {
+      "epoch": 10.92,
+      "grad_norm": 2.5430166721343994,
+      "learning_rate": 2.140628871872203e-05,
+      "loss": 1.0644,
+      "step": 36505
+    },
+    {
+      "epoch": 10.92,
+      "grad_norm": 3.8892323970794678,
+      "learning_rate": 2.140047462574932e-05,
+      "loss": 0.9569,
+      "step": 36510
+    },
+    {
+      "epoch": 10.92,
+      "grad_norm": 5.289989948272705,
+      "learning_rate": 2.1394660731574643e-05,
+      "loss": 1.04,
+      "step": 36515
+    },
+    {
+      "epoch": 10.93,
+      "grad_norm": 1.7710543870925903,
+      "learning_rate": 2.138884703651911e-05,
+      "loss": 1.024,
+      "step": 36520
+    },
+    {
+      "epoch": 10.93,
+      "grad_norm": 7.825911998748779,
+      "learning_rate": 2.138303354090381e-05,
+      "loss": 0.9066,
+      "step": 36525
+    },
+    {
+      "epoch": 10.93,
+      "grad_norm": 1.9190192222595215,
+      "learning_rate": 2.1377220245049793e-05,
+      "loss": 0.8708,
+      "step": 36530
+    },
+    {
+      "epoch": 10.93,
+      "grad_norm": 2.441214084625244,
+      "learning_rate": 2.1371407149278152e-05,
+      "loss": 0.9889,
+      "step": 36535
+    },
+    {
+      "epoch": 10.93,
+      "grad_norm": 5.316892623901367,
+      "learning_rate": 2.136559425390991e-05,
+      "loss": 1.0597,
+      "step": 36540
+    },
+    {
+      "epoch": 10.93,
+      "grad_norm": 5.2226667404174805,
+      "learning_rate": 2.135978155926613e-05,
+      "loss": 1.0606,
+      "step": 36545
+    },
+    {
+      "epoch": 10.94,
+      "grad_norm": 3.5756514072418213,
+      "learning_rate": 2.135396906566782e-05,
+      "loss": 1.1334,
+      "step": 36550
+    },
+    {
+      "epoch": 10.94,
+      "grad_norm": 1.187839150428772,
+      "learning_rate": 2.134815677343602e-05,
+      "loss": 1.2043,
+      "step": 36555
+    },
+    {
+      "epoch": 10.94,
+      "grad_norm": 5.102145671844482,
+      "learning_rate": 2.1342344682891722e-05,
+      "loss": 0.7688,
+      "step": 36560
+    },
+    {
+      "epoch": 10.94,
+      "grad_norm": 4.2323527336120605,
+      "learning_rate": 2.133653279435592e-05,
+      "loss": 1.158,
+      "step": 36565
+    },
+    {
+      "epoch": 10.94,
+      "grad_norm": 2.2554049491882324,
+      "learning_rate": 2.1330721108149614e-05,
+      "loss": 0.8822,
+      "step": 36570
+    },
+    {
+      "epoch": 10.94,
+      "grad_norm": 2.766228437423706,
+      "learning_rate": 2.132490962459375e-05,
+      "loss": 1.0252,
+      "step": 36575
+    },
+    {
+      "epoch": 10.94,
+      "grad_norm": 5.992697238922119,
+      "learning_rate": 2.1319098344009325e-05,
+      "loss": 1.0864,
+      "step": 36580
+    },
+    {
+      "epoch": 10.95,
+      "grad_norm": 1.6674883365631104,
+      "learning_rate": 2.1313287266717267e-05,
+      "loss": 0.9115,
+      "step": 36585
+    },
+    {
+      "epoch": 10.95,
+      "grad_norm": 1.9288809299468994,
+      "learning_rate": 2.1307476393038517e-05,
+      "loss": 1.0394,
+      "step": 36590
+    },
+    {
+      "epoch": 10.95,
+      "grad_norm": 1.3914613723754883,
+      "learning_rate": 2.1301665723294008e-05,
+      "loss": 0.9237,
+      "step": 36595
+    },
+    {
+      "epoch": 10.95,
+      "grad_norm": 2.7078752517700195,
+      "learning_rate": 2.129585525780466e-05,
+      "loss": 0.9573,
+      "step": 36600
+    },
+    {
+      "epoch": 10.95,
+      "grad_norm": 1.410092830657959,
+      "learning_rate": 2.1290044996891373e-05,
+      "loss": 0.9738,
+      "step": 36605
+    },
+    {
+      "epoch": 10.95,
+      "grad_norm": 2.085670232772827,
+      "learning_rate": 2.1284234940875045e-05,
+      "loss": 1.0609,
+      "step": 36610
+    },
+    {
+      "epoch": 10.95,
+      "grad_norm": 3.979780435562134,
+      "learning_rate": 2.1278425090076558e-05,
+      "loss": 1.1589,
+      "step": 36615
+    },
+    {
+      "epoch": 10.96,
+      "grad_norm": 4.352084636688232,
+      "learning_rate": 2.1272615444816792e-05,
+      "loss": 1.0897,
+      "step": 36620
+    },
+    {
+      "epoch": 10.96,
+      "grad_norm": 3.6647934913635254,
+      "learning_rate": 2.126680600541659e-05,
+      "loss": 1.0315,
+      "step": 36625
+    },
+    {
+      "epoch": 10.96,
+      "grad_norm": 4.659989833831787,
+      "learning_rate": 2.126099677219682e-05,
+      "loss": 0.9488,
+      "step": 36630
+    },
+    {
+      "epoch": 10.96,
+      "grad_norm": 4.262448310852051,
+      "learning_rate": 2.1255187745478305e-05,
+      "loss": 1.18,
+      "step": 36635
+    },
+    {
+      "epoch": 10.96,
+      "grad_norm": 7.49601936340332,
+      "learning_rate": 2.124937892558189e-05,
+      "loss": 1.1229,
+      "step": 36640
+    },
+    {
+      "epoch": 10.96,
+      "grad_norm": 1.145183801651001,
+      "learning_rate": 2.1243570312828365e-05,
+      "loss": 0.9284,
+      "step": 36645
+    },
+    {
+      "epoch": 10.97,
+      "grad_norm": 2.557342290878296,
+      "learning_rate": 2.1237761907538556e-05,
+      "loss": 0.982,
+      "step": 36650
+    },
+    {
+      "epoch": 10.97,
+      "grad_norm": 2.6082074642181396,
+      "learning_rate": 2.1231953710033246e-05,
+      "loss": 1.0863,
+      "step": 36655
+    },
+    {
+      "epoch": 10.97,
+      "grad_norm": 3.32584810256958,
+      "learning_rate": 2.122614572063321e-05,
+      "loss": 1.0423,
+      "step": 36660
+    },
+    {
+      "epoch": 10.97,
+      "grad_norm": 2.8699169158935547,
+      "learning_rate": 2.122033793965923e-05,
+      "loss": 1.0971,
+      "step": 36665
+    },
+    {
+      "epoch": 10.97,
+      "grad_norm": 3.991990566253662,
+      "learning_rate": 2.1214530367432047e-05,
+      "loss": 1.1467,
+      "step": 36670
+    },
+    {
+      "epoch": 10.97,
+      "grad_norm": 2.663698673248291,
+      "learning_rate": 2.120872300427243e-05,
+      "loss": 0.9862,
+      "step": 36675
+    },
+    {
+      "epoch": 10.97,
+      "grad_norm": 3.4537861347198486,
+      "learning_rate": 2.1202915850501087e-05,
+      "loss": 1.1518,
+      "step": 36680
+    },
+    {
+      "epoch": 10.98,
+      "grad_norm": 3.4938299655914307,
+      "learning_rate": 2.1197108906438765e-05,
+      "loss": 1.1402,
+      "step": 36685
+    },
+    {
+      "epoch": 10.98,
+      "grad_norm": 2.6186256408691406,
+      "learning_rate": 2.119130217240616e-05,
+      "loss": 1.2347,
+      "step": 36690
+    },
+    {
+      "epoch": 10.98,
+      "grad_norm": 2.723661422729492,
+      "learning_rate": 2.118549564872398e-05,
+      "loss": 1.0232,
+      "step": 36695
+    },
+    {
+      "epoch": 10.98,
+      "grad_norm": 4.313886642456055,
+      "learning_rate": 2.1179689335712906e-05,
+      "loss": 1.2257,
+      "step": 36700
+    },
+    {
+      "epoch": 10.98,
+      "grad_norm": 1.8701528310775757,
+      "learning_rate": 2.1173883233693623e-05,
+      "loss": 1.207,
+      "step": 36705
+    },
+    {
+      "epoch": 10.98,
+      "grad_norm": 1.5927581787109375,
+      "learning_rate": 2.1168077342986793e-05,
+      "loss": 1.0767,
+      "step": 36710
+    },
+    {
+      "epoch": 10.98,
+      "grad_norm": 5.454744338989258,
+      "learning_rate": 2.116227166391307e-05,
+      "loss": 0.9561,
+      "step": 36715
+    },
+    {
+      "epoch": 10.99,
+      "grad_norm": 0.9375532269477844,
+      "learning_rate": 2.1156466196793086e-05,
+      "loss": 1.1743,
+      "step": 36720
+    },
+    {
+      "epoch": 10.99,
+      "grad_norm": 5.345798015594482,
+      "learning_rate": 2.1150660941947486e-05,
+      "loss": 1.0879,
+      "step": 36725
+    },
+    {
+      "epoch": 10.99,
+      "grad_norm": 2.450878381729126,
+      "learning_rate": 2.1144855899696873e-05,
+      "loss": 0.8149,
+      "step": 36730
+    },
+    {
+      "epoch": 10.99,
+      "grad_norm": 2.603236436843872,
+      "learning_rate": 2.113905107036187e-05,
+      "loss": 1.1418,
+      "step": 36735
+    },
+    {
+      "epoch": 10.99,
+      "grad_norm": 3.163844347000122,
+      "learning_rate": 2.1133246454263053e-05,
+      "loss": 1.0855,
+      "step": 36740
+    },
+    {
+      "epoch": 10.99,
+      "grad_norm": 2.926191568374634,
+      "learning_rate": 2.1127442051721026e-05,
+      "loss": 1.1282,
+      "step": 36745
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 4.042135238647461,
+      "learning_rate": 2.112163786305635e-05,
+      "loss": 1.0874,
+      "step": 36750
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 4.720858097076416,
+      "learning_rate": 2.1115833888589575e-05,
+      "loss": 1.306,
+      "step": 36755
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 5.201600074768066,
+      "learning_rate": 2.1110030128641264e-05,
+      "loss": 1.0244,
+      "step": 36760
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 4.97416353225708,
+      "learning_rate": 2.1104226583531936e-05,
+      "loss": 1.1876,
+      "step": 36765
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 1.4578512907028198,
+      "learning_rate": 2.1098423253582136e-05,
+      "loss": 1.1209,
+      "step": 36770
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 2.909250259399414,
+      "learning_rate": 2.109262013911235e-05,
+      "loss": 0.8941,
+      "step": 36775
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 1.240099549293518,
+      "learning_rate": 2.108681724044311e-05,
+      "loss": 0.9776,
+      "step": 36780
+    },
+    {
+      "epoch": 11.01,
+      "grad_norm": 3.794100284576416,
+      "learning_rate": 2.1081014557894875e-05,
+      "loss": 1.0358,
+      "step": 36785
+    },
+    {
+      "epoch": 11.01,
+      "grad_norm": 1.7050538063049316,
+      "learning_rate": 2.1075212091788133e-05,
+      "loss": 0.8003,
+      "step": 36790
+    },
+    {
+      "epoch": 11.01,
+      "grad_norm": 1.4368036985397339,
+      "learning_rate": 2.106940984244335e-05,
+      "loss": 1.1668,
+      "step": 36795
+    },
+    {
+      "epoch": 11.01,
+      "grad_norm": 4.024328708648682,
+      "learning_rate": 2.1063607810180975e-05,
+      "loss": 1.0022,
+      "step": 36800
+    },
+    {
+      "epoch": 11.01,
+      "grad_norm": 2.9807326793670654,
+      "learning_rate": 2.1057805995321447e-05,
+      "loss": 0.9635,
+      "step": 36805
+    },
+    {
+      "epoch": 11.01,
+      "grad_norm": 2.2452774047851562,
+      "learning_rate": 2.1052004398185193e-05,
+      "loss": 1.0438,
+      "step": 36810
+    },
+    {
+      "epoch": 11.01,
+      "grad_norm": 2.678302764892578,
+      "learning_rate": 2.104620301909264e-05,
+      "loss": 0.9814,
+      "step": 36815
+    },
+    {
+      "epoch": 11.02,
+      "grad_norm": 1.3954975605010986,
+      "learning_rate": 2.104040185836419e-05,
+      "loss": 1.003,
+      "step": 36820
+    },
+    {
+      "epoch": 11.02,
+      "grad_norm": 1.7821259498596191,
+      "learning_rate": 2.1034600916320212e-05,
+      "loss": 1.1241,
+      "step": 36825
+    },
+    {
+      "epoch": 11.02,
+      "grad_norm": 2.254903793334961,
+      "learning_rate": 2.102880019328112e-05,
+      "loss": 1.0537,
+      "step": 36830
+    },
+    {
+      "epoch": 11.02,
+      "grad_norm": 3.2131004333496094,
+      "learning_rate": 2.102299968956725e-05,
+      "loss": 1.0781,
+      "step": 36835
+    },
+    {
+      "epoch": 11.02,
+      "grad_norm": 2.1577985286712646,
+      "learning_rate": 2.101719940549899e-05,
+      "loss": 0.9388,
+      "step": 36840
+    },
+    {
+      "epoch": 11.02,
+      "grad_norm": 3.0737853050231934,
+      "learning_rate": 2.1011399341396664e-05,
+      "loss": 0.9643,
+      "step": 36845
+    },
+    {
+      "epoch": 11.03,
+      "grad_norm": 3.187286853790283,
+      "learning_rate": 2.1005599497580596e-05,
+      "loss": 1.0258,
+      "step": 36850
+    },
+    {
+      "epoch": 11.03,
+      "grad_norm": 1.8270454406738281,
+      "learning_rate": 2.0999799874371124e-05,
+      "loss": 1.0375,
+      "step": 36855
+    },
+    {
+      "epoch": 11.03,
+      "grad_norm": 3.7559213638305664,
+      "learning_rate": 2.0994000472088537e-05,
+      "loss": 1.117,
+      "step": 36860
+    },
+    {
+      "epoch": 11.03,
+      "grad_norm": 2.0919082164764404,
+      "learning_rate": 2.0988201291053154e-05,
+      "loss": 1.0887,
+      "step": 36865
+    },
+    {
+      "epoch": 11.03,
+      "grad_norm": 2.3934216499328613,
+      "learning_rate": 2.098240233158523e-05,
+      "loss": 1.1399,
+      "step": 36870
+    },
+    {
+      "epoch": 11.03,
+      "grad_norm": 1.183286190032959,
+      "learning_rate": 2.0976603594005063e-05,
+      "loss": 1.0758,
+      "step": 36875
+    },
+    {
+      "epoch": 11.03,
+      "grad_norm": 2.7067835330963135,
+      "learning_rate": 2.0970805078632887e-05,
+      "loss": 1.2898,
+      "step": 36880
+    },
+    {
+      "epoch": 11.04,
+      "grad_norm": 2.4200856685638428,
+      "learning_rate": 2.0965006785788958e-05,
+      "loss": 0.9024,
+      "step": 36885
+    },
+    {
+      "epoch": 11.04,
+      "grad_norm": 1.7485568523406982,
+      "learning_rate": 2.0959208715793516e-05,
+      "loss": 1.0622,
+      "step": 36890
+    },
+    {
+      "epoch": 11.04,
+      "grad_norm": 1.9122002124786377,
+      "learning_rate": 2.095341086896677e-05,
+      "loss": 0.8777,
+      "step": 36895
+    },
+    {
+      "epoch": 11.04,
+      "grad_norm": 2.849827289581299,
+      "learning_rate": 2.0947613245628944e-05,
+      "loss": 0.9607,
+      "step": 36900
+    },
+    {
+      "epoch": 11.04,
+      "grad_norm": 2.4549102783203125,
+      "learning_rate": 2.0941815846100216e-05,
+      "loss": 0.8974,
+      "step": 36905
+    },
+    {
+      "epoch": 11.04,
+      "grad_norm": 6.144309997558594,
+      "learning_rate": 2.093601867070079e-05,
+      "loss": 1.0058,
+      "step": 36910
+    },
+    {
+      "epoch": 11.04,
+      "grad_norm": 2.7253663539886475,
+      "learning_rate": 2.093022171975083e-05,
+      "loss": 1.0684,
+      "step": 36915
+    },
+    {
+      "epoch": 11.05,
+      "grad_norm": 7.483120918273926,
+      "learning_rate": 2.0924424993570485e-05,
+      "loss": 0.8457,
+      "step": 36920
+    },
+    {
+      "epoch": 11.05,
+      "grad_norm": 2.1432487964630127,
+      "learning_rate": 2.091862849247992e-05,
+      "loss": 0.9498,
+      "step": 36925
+    },
+    {
+      "epoch": 11.05,
+      "grad_norm": 1.383862853050232,
+      "learning_rate": 2.091283221679925e-05,
+      "loss": 1.0247,
+      "step": 36930
+    },
+    {
+      "epoch": 11.05,
+      "grad_norm": 2.251617431640625,
+      "learning_rate": 2.090703616684862e-05,
+      "loss": 1.0731,
+      "step": 36935
+    },
+    {
+      "epoch": 11.05,
+      "grad_norm": 2.736708879470825,
+      "learning_rate": 2.0901240342948128e-05,
+      "loss": 1.0303,
+      "step": 36940
+    },
+    {
+      "epoch": 11.05,
+      "grad_norm": 8.472616195678711,
+      "learning_rate": 2.089544474541786e-05,
+      "loss": 1.1477,
+      "step": 36945
+    },
+    {
+      "epoch": 11.06,
+      "grad_norm": 2.7550952434539795,
+      "learning_rate": 2.0889649374577923e-05,
+      "loss": 0.7887,
+      "step": 36950
+    },
+    {
+      "epoch": 11.06,
+      "grad_norm": 4.178884506225586,
+      "learning_rate": 2.088385423074837e-05,
+      "loss": 1.0598,
+      "step": 36955
+    },
+    {
+      "epoch": 11.06,
+      "grad_norm": 2.964043140411377,
+      "learning_rate": 2.087805931424927e-05,
+      "loss": 1.1823,
+      "step": 36960
+    },
+    {
+      "epoch": 11.06,
+      "grad_norm": 1.5841584205627441,
+      "learning_rate": 2.0872264625400673e-05,
+      "loss": 0.8966,
+      "step": 36965
+    },
+    {
+      "epoch": 11.06,
+      "grad_norm": 3.2998929023742676,
+      "learning_rate": 2.086647016452261e-05,
+      "loss": 1.2444,
+      "step": 36970
+    },
+    {
+      "epoch": 11.06,
+      "grad_norm": 9.047067642211914,
+      "learning_rate": 2.086067593193511e-05,
+      "loss": 1.1063,
+      "step": 36975
+    },
+    {
+      "epoch": 11.06,
+      "grad_norm": 4.006898403167725,
+      "learning_rate": 2.085488192795816e-05,
+      "loss": 1.0788,
+      "step": 36980
+    },
+    {
+      "epoch": 11.07,
+      "grad_norm": 1.3867641687393188,
+      "learning_rate": 2.084908815291179e-05,
+      "loss": 1.0392,
+      "step": 36985
+    },
+    {
+      "epoch": 11.07,
+      "grad_norm": 2.389448404312134,
+      "learning_rate": 2.084329460711595e-05,
+      "loss": 1.169,
+      "step": 36990
+    },
+    {
+      "epoch": 11.07,
+      "grad_norm": 6.467934608459473,
+      "learning_rate": 2.0837501290890644e-05,
+      "loss": 1.1412,
+      "step": 36995
+    },
+    {
+      "epoch": 11.07,
+      "grad_norm": 1.3220041990280151,
+      "learning_rate": 2.08317082045558e-05,
+      "loss": 1.0507,
+      "step": 37000
+    },
+    {
+      "epoch": 11.07,
+      "grad_norm": 2.8691465854644775,
+      "learning_rate": 2.082591534843139e-05,
+      "loss": 1.0901,
+      "step": 37005
+    },
+    {
+      "epoch": 11.07,
+      "grad_norm": 1.4073899984359741,
+      "learning_rate": 2.082012272283734e-05,
+      "loss": 1.1085,
+      "step": 37010
+    },
+    {
+      "epoch": 11.07,
+      "grad_norm": 1.7630438804626465,
+      "learning_rate": 2.0814330328093557e-05,
+      "loss": 1.072,
+      "step": 37015
+    },
+    {
+      "epoch": 11.08,
+      "grad_norm": 3.4774584770202637,
+      "learning_rate": 2.0808538164519965e-05,
+      "loss": 0.9849,
+      "step": 37020
+    },
+    {
+      "epoch": 11.08,
+      "grad_norm": 1.4681512117385864,
+      "learning_rate": 2.0802746232436445e-05,
+      "loss": 1.0581,
+      "step": 37025
+    },
+    {
+      "epoch": 11.08,
+      "grad_norm": 1.4620872735977173,
+      "learning_rate": 2.0796954532162898e-05,
+      "loss": 1.1402,
+      "step": 37030
+    },
+    {
+      "epoch": 11.08,
+      "grad_norm": 4.764907360076904,
+      "learning_rate": 2.0791163064019174e-05,
+      "loss": 0.9688,
+      "step": 37035
+    },
+    {
+      "epoch": 11.08,
+      "grad_norm": 1.992462396621704,
+      "learning_rate": 2.078537182832515e-05,
+      "loss": 0.8545,
+      "step": 37040
+    },
+    {
+      "epoch": 11.08,
+      "grad_norm": 2.524420976638794,
+      "learning_rate": 2.0779580825400653e-05,
+      "loss": 0.9803,
+      "step": 37045
+    },
+    {
+      "epoch": 11.08,
+      "grad_norm": 4.6814351081848145,
+      "learning_rate": 2.077379005556552e-05,
+      "loss": 1.0973,
+      "step": 37050
+    },
+    {
+      "epoch": 11.09,
+      "grad_norm": 1.9848724603652954,
+      "learning_rate": 2.076799951913957e-05,
+      "loss": 1.0625,
+      "step": 37055
+    },
+    {
+      "epoch": 11.09,
+      "grad_norm": 5.070384502410889,
+      "learning_rate": 2.0762209216442607e-05,
+      "loss": 1.1657,
+      "step": 37060
+    },
+    {
+      "epoch": 11.09,
+      "grad_norm": 3.4245998859405518,
+      "learning_rate": 2.0756419147794427e-05,
+      "loss": 0.7992,
+      "step": 37065
+    },
+    {
+      "epoch": 11.09,
+      "grad_norm": 3.5566506385803223,
+      "learning_rate": 2.075062931351481e-05,
+      "loss": 0.98,
+      "step": 37070
+    },
+    {
+      "epoch": 11.09,
+      "grad_norm": 1.4207732677459717,
+      "learning_rate": 2.0744839713923503e-05,
+      "loss": 1.1456,
+      "step": 37075
+    },
+    {
+      "epoch": 11.09,
+      "grad_norm": 1.3465211391448975,
+      "learning_rate": 2.073905034934029e-05,
+      "loss": 1.0347,
+      "step": 37080
+    },
+    {
+      "epoch": 11.1,
+      "grad_norm": 1.9427791833877563,
+      "learning_rate": 2.0733261220084886e-05,
+      "loss": 1.0611,
+      "step": 37085
+    },
+    {
+      "epoch": 11.1,
+      "grad_norm": 1.6937408447265625,
+      "learning_rate": 2.0727472326477043e-05,
+      "loss": 1.0662,
+      "step": 37090
+    },
+    {
+      "epoch": 11.1,
+      "grad_norm": 3.2311325073242188,
+      "learning_rate": 2.072168366883645e-05,
+      "loss": 1.085,
+      "step": 37095
+    },
+    {
+      "epoch": 11.1,
+      "grad_norm": 1.9753808975219727,
+      "learning_rate": 2.0715895247482833e-05,
+      "loss": 0.9925,
+      "step": 37100
+    },
+    {
+      "epoch": 11.1,
+      "grad_norm": 1.8392963409423828,
+      "learning_rate": 2.0710107062735867e-05,
+      "loss": 0.9266,
+      "step": 37105
+    },
+    {
+      "epoch": 11.1,
+      "grad_norm": 1.352810025215149,
+      "learning_rate": 2.0704319114915218e-05,
+      "loss": 1.0309,
+      "step": 37110
+    },
+    {
+      "epoch": 11.1,
+      "grad_norm": 2.322559118270874,
+      "learning_rate": 2.0698531404340573e-05,
+      "loss": 1.0988,
+      "step": 37115
+    },
+    {
+      "epoch": 11.11,
+      "grad_norm": 6.6522650718688965,
+      "learning_rate": 2.0692743931331554e-05,
+      "loss": 0.7917,
+      "step": 37120
+    },
+    {
+      "epoch": 11.11,
+      "grad_norm": 5.853671550750732,
+      "learning_rate": 2.068695669620782e-05,
+      "loss": 1.034,
+      "step": 37125
+    },
+    {
+      "epoch": 11.11,
+      "grad_norm": 2.2540881633758545,
+      "learning_rate": 2.0681169699288974e-05,
+      "loss": 1.0728,
+      "step": 37130
+    },
+    {
+      "epoch": 11.11,
+      "grad_norm": 3.5193233489990234,
+      "learning_rate": 2.067538294089465e-05,
+      "loss": 1.0295,
+      "step": 37135
+    },
+    {
+      "epoch": 11.11,
+      "grad_norm": 3.8422839641571045,
+      "learning_rate": 2.0669596421344422e-05,
+      "loss": 1.1198,
+      "step": 37140
+    },
+    {
+      "epoch": 11.11,
+      "grad_norm": 2.048434019088745,
+      "learning_rate": 2.0663810140957884e-05,
+      "loss": 1.0309,
+      "step": 37145
+    },
+    {
+      "epoch": 11.11,
+      "grad_norm": 1.556140661239624,
+      "learning_rate": 2.0658024100054608e-05,
+      "loss": 0.9682,
+      "step": 37150
+    },
+    {
+      "epoch": 11.12,
+      "grad_norm": 2.3426733016967773,
+      "learning_rate": 2.0652238298954142e-05,
+      "loss": 1.0461,
+      "step": 37155
+    },
+    {
+      "epoch": 11.12,
+      "grad_norm": 2.597688913345337,
+      "learning_rate": 2.0646452737976037e-05,
+      "loss": 1.1395,
+      "step": 37160
+    },
+    {
+      "epoch": 11.12,
+      "grad_norm": 2.909790277481079,
+      "learning_rate": 2.0640667417439826e-05,
+      "loss": 1.0784,
+      "step": 37165
+    },
+    {
+      "epoch": 11.12,
+      "grad_norm": 4.795456409454346,
+      "learning_rate": 2.0634882337665018e-05,
+      "loss": 0.9231,
+      "step": 37170
+    },
+    {
+      "epoch": 11.12,
+      "grad_norm": 4.892722129821777,
+      "learning_rate": 2.0629097498971128e-05,
+      "loss": 1.0418,
+      "step": 37175
+    },
+    {
+      "epoch": 11.12,
+      "grad_norm": 2.8806307315826416,
+      "learning_rate": 2.062331290167763e-05,
+      "loss": 1.1341,
+      "step": 37180
+    },
+    {
+      "epoch": 11.13,
+      "grad_norm": 2.0100202560424805,
+      "learning_rate": 2.061752854610402e-05,
+      "loss": 1.0645,
+      "step": 37185
+    },
+    {
+      "epoch": 11.13,
+      "grad_norm": 3.8369128704071045,
+      "learning_rate": 2.0611744432569743e-05,
+      "loss": 0.8733,
+      "step": 37190
+    },
+    {
+      "epoch": 11.13,
+      "grad_norm": 1.7996447086334229,
+      "learning_rate": 2.060596056139427e-05,
+      "loss": 1.0846,
+      "step": 37195
+    },
+    {
+      "epoch": 11.13,
+      "grad_norm": 2.4281742572784424,
+      "learning_rate": 2.060017693289703e-05,
+      "loss": 0.9427,
+      "step": 37200
+    },
+    {
+      "epoch": 11.13,
+      "grad_norm": 3.646941900253296,
+      "learning_rate": 2.0594393547397432e-05,
+      "loss": 0.9663,
+      "step": 37205
+    },
+    {
+      "epoch": 11.13,
+      "grad_norm": 2.000669240951538,
+      "learning_rate": 2.058861040521491e-05,
+      "loss": 1.1134,
+      "step": 37210
+    },
+    {
+      "epoch": 11.13,
+      "grad_norm": 3.394282579421997,
+      "learning_rate": 2.058282750666884e-05,
+      "loss": 1.0419,
+      "step": 37215
+    },
+    {
+      "epoch": 11.14,
+      "grad_norm": 1.8753750324249268,
+      "learning_rate": 2.0577044852078624e-05,
+      "loss": 0.9423,
+      "step": 37220
+    },
+    {
+      "epoch": 11.14,
+      "grad_norm": 1.0014708042144775,
+      "learning_rate": 2.0571262441763613e-05,
+      "loss": 1.1235,
+      "step": 37225
+    },
+    {
+      "epoch": 11.14,
+      "grad_norm": 6.966102600097656,
+      "learning_rate": 2.0565480276043186e-05,
+      "loss": 1.0036,
+      "step": 37230
+    },
+    {
+      "epoch": 11.14,
+      "grad_norm": 3.9602415561676025,
+      "learning_rate": 2.055969835523667e-05,
+      "loss": 1.0308,
+      "step": 37235
+    },
+    {
+      "epoch": 11.14,
+      "grad_norm": 13.549233436584473,
+      "learning_rate": 2.0553916679663394e-05,
+      "loss": 1.1082,
+      "step": 37240
+    },
+    {
+      "epoch": 11.14,
+      "grad_norm": 3.726292610168457,
+      "learning_rate": 2.0548135249642683e-05,
+      "loss": 1.1484,
+      "step": 37245
+    },
+    {
+      "epoch": 11.14,
+      "grad_norm": 2.3705708980560303,
+      "learning_rate": 2.054235406549383e-05,
+      "loss": 1.0761,
+      "step": 37250
+    },
+    {
+      "epoch": 11.15,
+      "grad_norm": 2.891065835952759,
+      "learning_rate": 2.0536573127536133e-05,
+      "loss": 1.0185,
+      "step": 37255
+    },
+    {
+      "epoch": 11.15,
+      "grad_norm": 1.9550832509994507,
+      "learning_rate": 2.053079243608886e-05,
+      "loss": 1.1753,
+      "step": 37260
+    },
+    {
+      "epoch": 11.15,
+      "grad_norm": 4.056787490844727,
+      "learning_rate": 2.052501199147128e-05,
+      "loss": 0.965,
+      "step": 37265
+    },
+    {
+      "epoch": 11.15,
+      "grad_norm": 4.4339494705200195,
+      "learning_rate": 2.0519231794002637e-05,
+      "loss": 1.0662,
+      "step": 37270
+    },
+    {
+      "epoch": 11.15,
+      "grad_norm": 1.7741667032241821,
+      "learning_rate": 2.0513451844002154e-05,
+      "loss": 1.0956,
+      "step": 37275
+    },
+    {
+      "epoch": 11.15,
+      "grad_norm": 7.996791839599609,
+      "learning_rate": 2.0507672141789074e-05,
+      "loss": 0.7328,
+      "step": 37280
+    },
+    {
+      "epoch": 11.16,
+      "grad_norm": 1.697172999382019,
+      "learning_rate": 2.050189268768258e-05,
+      "loss": 1.0675,
+      "step": 37285
+    },
+    {
+      "epoch": 11.16,
+      "grad_norm": 1.4075742959976196,
+      "learning_rate": 2.049611348200189e-05,
+      "loss": 1.0553,
+      "step": 37290
+    },
+    {
+      "epoch": 11.16,
+      "grad_norm": 2.2331416606903076,
+      "learning_rate": 2.0490334525066172e-05,
+      "loss": 0.9881,
+      "step": 37295
+    },
+    {
+      "epoch": 11.16,
+      "grad_norm": 1.8239549398422241,
+      "learning_rate": 2.0484555817194576e-05,
+      "loss": 1.1081,
+      "step": 37300
+    },
+    {
+      "epoch": 11.16,
+      "grad_norm": 3.33955454826355,
+      "learning_rate": 2.047877735870628e-05,
+      "loss": 1.0621,
+      "step": 37305
+    },
+    {
+      "epoch": 11.16,
+      "grad_norm": 4.439455509185791,
+      "learning_rate": 2.0472999149920403e-05,
+      "loss": 0.8967,
+      "step": 37310
+    },
+    {
+      "epoch": 11.16,
+      "grad_norm": 1.3203307390213013,
+      "learning_rate": 2.0467221191156085e-05,
+      "loss": 1.1234,
+      "step": 37315
+    },
+    {
+      "epoch": 11.17,
+      "grad_norm": 1.8145140409469604,
+      "learning_rate": 2.046144348273242e-05,
+      "loss": 0.8947,
+      "step": 37320
+    },
+    {
+      "epoch": 11.17,
+      "grad_norm": 1.8895070552825928,
+      "learning_rate": 2.0455666024968527e-05,
+      "loss": 0.8826,
+      "step": 37325
+    },
+    {
+      "epoch": 11.17,
+      "grad_norm": 2.4296302795410156,
+      "learning_rate": 2.0449888818183465e-05,
+      "loss": 1.1869,
+      "step": 37330
+    },
+    {
+      "epoch": 11.17,
+      "grad_norm": 6.826413154602051,
+      "learning_rate": 2.0444111862696314e-05,
+      "loss": 0.9804,
+      "step": 37335
+    },
+    {
+      "epoch": 11.17,
+      "grad_norm": 3.0793893337249756,
+      "learning_rate": 2.0438335158826134e-05,
+      "loss": 1.1888,
+      "step": 37340
+    },
+    {
+      "epoch": 11.17,
+      "grad_norm": 5.991076469421387,
+      "learning_rate": 2.043255870689196e-05,
+      "loss": 1.0324,
+      "step": 37345
+    },
+    {
+      "epoch": 11.17,
+      "grad_norm": 4.768267631530762,
+      "learning_rate": 2.0426782507212822e-05,
+      "loss": 0.9743,
+      "step": 37350
+    },
+    {
+      "epoch": 11.18,
+      "grad_norm": 2.971740245819092,
+      "learning_rate": 2.0421006560107726e-05,
+      "loss": 1.0149,
+      "step": 37355
+    },
+    {
+      "epoch": 11.18,
+      "grad_norm": 4.107237339019775,
+      "learning_rate": 2.041523086589569e-05,
+      "loss": 1.147,
+      "step": 37360
+    },
+    {
+      "epoch": 11.18,
+      "grad_norm": 2.612445116043091,
+      "learning_rate": 2.0409455424895686e-05,
+      "loss": 0.9139,
+      "step": 37365
+    },
+    {
+      "epoch": 11.18,
+      "grad_norm": 1.6917109489440918,
+      "learning_rate": 2.0403680237426677e-05,
+      "loss": 1.003,
+      "step": 37370
+    },
+    {
+      "epoch": 11.18,
+      "grad_norm": 24.12595558166504,
+      "learning_rate": 2.0397905303807642e-05,
+      "loss": 0.9988,
+      "step": 37375
+    },
+    {
+      "epoch": 11.18,
+      "grad_norm": 3.371530532836914,
+      "learning_rate": 2.0392130624357502e-05,
+      "loss": 0.9194,
+      "step": 37380
+    },
+    {
+      "epoch": 11.19,
+      "grad_norm": 1.867092490196228,
+      "learning_rate": 2.038635619939521e-05,
+      "loss": 1.015,
+      "step": 37385
+    },
+    {
+      "epoch": 11.19,
+      "grad_norm": 1.5033526420593262,
+      "learning_rate": 2.0380582029239655e-05,
+      "loss": 1.0523,
+      "step": 37390
+    },
+    {
+      "epoch": 11.19,
+      "grad_norm": 1.303634524345398,
+      "learning_rate": 2.0374808114209767e-05,
+      "loss": 0.9923,
+      "step": 37395
+    },
+    {
+      "epoch": 11.19,
+      "grad_norm": 0.9598073363304138,
+      "learning_rate": 2.036903445462442e-05,
+      "loss": 0.8324,
+      "step": 37400
+    },
+    {
+      "epoch": 11.19,
+      "grad_norm": 2.9755938053131104,
+      "learning_rate": 2.0363261050802473e-05,
+      "loss": 1.0322,
+      "step": 37405
+    },
+    {
+      "epoch": 11.19,
+      "grad_norm": 2.0213468074798584,
+      "learning_rate": 2.0357487903062805e-05,
+      "loss": 1.1045,
+      "step": 37410
+    },
+    {
+      "epoch": 11.19,
+      "grad_norm": 2.2053709030151367,
+      "learning_rate": 2.0351715011724255e-05,
+      "loss": 1.0586,
+      "step": 37415
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 2.988892078399658,
+      "learning_rate": 2.0345942377105654e-05,
+      "loss": 0.9125,
+      "step": 37420
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 1.9951380491256714,
+      "learning_rate": 2.034016999952582e-05,
+      "loss": 1.0415,
+      "step": 37425
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 2.1068739891052246,
+      "learning_rate": 2.0334397879303545e-05,
+      "loss": 1.0635,
+      "step": 37430
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 1.8847640752792358,
+      "learning_rate": 2.032862601675764e-05,
+      "loss": 1.1524,
+      "step": 37435
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 2.446730375289917,
+      "learning_rate": 2.0322854412206848e-05,
+      "loss": 1.1736,
+      "step": 37440
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 3.1144163608551025,
+      "learning_rate": 2.0317083065969957e-05,
+      "loss": 1.0448,
+      "step": 37445
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 1.5927165746688843,
+      "learning_rate": 2.0311311978365694e-05,
+      "loss": 0.8874,
+      "step": 37450
+    },
+    {
+      "epoch": 11.21,
+      "grad_norm": 2.248284101486206,
+      "learning_rate": 2.0305541149712802e-05,
+      "loss": 1.134,
+      "step": 37455
+    },
+    {
+      "epoch": 11.21,
+      "grad_norm": 5.157928943634033,
+      "learning_rate": 2.0299770580329997e-05,
+      "loss": 1.1917,
+      "step": 37460
+    },
+    {
+      "epoch": 11.21,
+      "grad_norm": 2.8743419647216797,
+      "learning_rate": 2.0294000270535963e-05,
+      "loss": 1.0266,
+      "step": 37465
+    },
+    {
+      "epoch": 11.21,
+      "grad_norm": 2.73984956741333,
+      "learning_rate": 2.0288230220649415e-05,
+      "loss": 0.9329,
+      "step": 37470
+    },
+    {
+      "epoch": 11.21,
+      "grad_norm": 1.6899389028549194,
+      "learning_rate": 2.0282460430989003e-05,
+      "loss": 0.9871,
+      "step": 37475
+    },
+    {
+      "epoch": 11.21,
+      "grad_norm": 3.168191432952881,
+      "learning_rate": 2.027669090187341e-05,
+      "loss": 1.0343,
+      "step": 37480
+    },
+    {
+      "epoch": 11.22,
+      "grad_norm": 4.179646968841553,
+      "learning_rate": 2.0270921633621255e-05,
+      "loss": 0.9998,
+      "step": 37485
+    },
+    {
+      "epoch": 11.22,
+      "grad_norm": 2.1786577701568604,
+      "learning_rate": 2.0265152626551195e-05,
+      "loss": 1.2523,
+      "step": 37490
+    },
+    {
+      "epoch": 11.22,
+      "grad_norm": 2.7645535469055176,
+      "learning_rate": 2.025938388098183e-05,
+      "loss": 1.078,
+      "step": 37495
+    },
+    {
+      "epoch": 11.22,
+      "grad_norm": 4.356101036071777,
+      "learning_rate": 2.0253615397231764e-05,
+      "loss": 1.0382,
+      "step": 37500
+    },
+    {
+      "epoch": 11.22,
+      "grad_norm": 1.3754960298538208,
+      "learning_rate": 2.0247847175619584e-05,
+      "loss": 1.1546,
+      "step": 37505
+    },
+    {
+      "epoch": 11.22,
+      "grad_norm": 1.7943824529647827,
+      "learning_rate": 2.024207921646387e-05,
+      "loss": 1.0802,
+      "step": 37510
+    },
+    {
+      "epoch": 11.22,
+      "grad_norm": 2.5919864177703857,
+      "learning_rate": 2.023631152008317e-05,
+      "loss": 1.0181,
+      "step": 37515
+    },
+    {
+      "epoch": 11.23,
+      "grad_norm": 3.6458067893981934,
+      "learning_rate": 2.0230544086796045e-05,
+      "loss": 0.9573,
+      "step": 37520
+    },
+    {
+      "epoch": 11.23,
+      "grad_norm": 3.077378273010254,
+      "learning_rate": 2.0224776916920996e-05,
+      "loss": 1.2146,
+      "step": 37525
+    },
+    {
+      "epoch": 11.23,
+      "grad_norm": 3.0590150356292725,
+      "learning_rate": 2.0219010010776568e-05,
+      "loss": 0.8816,
+      "step": 37530
+    },
+    {
+      "epoch": 11.23,
+      "grad_norm": 2.068225145339966,
+      "learning_rate": 2.0213243368681234e-05,
+      "loss": 0.929,
+      "step": 37535
+    },
+    {
+      "epoch": 11.23,
+      "grad_norm": 3.8140151500701904,
+      "learning_rate": 2.0207476990953505e-05,
+      "loss": 0.8558,
+      "step": 37540
+    },
+    {
+      "epoch": 11.23,
+      "grad_norm": 2.6358039379119873,
+      "learning_rate": 2.0201710877911832e-05,
+      "loss": 1.1409,
+      "step": 37545
+    },
+    {
+      "epoch": 11.23,
+      "grad_norm": 3.132976531982422,
+      "learning_rate": 2.0195945029874687e-05,
+      "loss": 0.9641,
+      "step": 37550
+    },
+    {
+      "epoch": 11.24,
+      "grad_norm": 5.017673015594482,
+      "learning_rate": 2.0190179447160505e-05,
+      "loss": 1.0201,
+      "step": 37555
+    },
+    {
+      "epoch": 11.24,
+      "grad_norm": 2.018449306488037,
+      "learning_rate": 2.0184414130087704e-05,
+      "loss": 1.1206,
+      "step": 37560
+    },
+    {
+      "epoch": 11.24,
+      "grad_norm": 5.14389705657959,
+      "learning_rate": 2.0178649078974714e-05,
+      "loss": 1.0431,
+      "step": 37565
+    },
+    {
+      "epoch": 11.24,
+      "grad_norm": 1.725806713104248,
+      "learning_rate": 2.0172884294139917e-05,
+      "loss": 0.9678,
+      "step": 37570
+    },
+    {
+      "epoch": 11.24,
+      "grad_norm": 1.3185760974884033,
+      "learning_rate": 2.0167119775901706e-05,
+      "loss": 0.8791,
+      "step": 37575
+    },
+    {
+      "epoch": 11.24,
+      "grad_norm": 2.4314846992492676,
+      "learning_rate": 2.016135552457844e-05,
+      "loss": 0.9986,
+      "step": 37580
+    },
+    {
+      "epoch": 11.24,
+      "grad_norm": 3.166346311569214,
+      "learning_rate": 2.015559154048849e-05,
+      "loss": 1.1143,
+      "step": 37585
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 2.213935136795044,
+      "learning_rate": 2.014982782395018e-05,
+      "loss": 0.9987,
+      "step": 37590
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 1.993200659751892,
+      "learning_rate": 2.0144064375281834e-05,
+      "loss": 1.0445,
+      "step": 37595
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 3.238537073135376,
+      "learning_rate": 2.0138301194801768e-05,
+      "loss": 1.1212,
+      "step": 37600
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 5.660292625427246,
+      "learning_rate": 2.0132538282828273e-05,
+      "loss": 1.1843,
+      "step": 37605
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 2.940760374069214,
+      "learning_rate": 2.012677563967963e-05,
+      "loss": 1.1591,
+      "step": 37610
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 2.0081498622894287,
+      "learning_rate": 2.0121013265674095e-05,
+      "loss": 1.0123,
+      "step": 37615
+    },
+    {
+      "epoch": 11.26,
+      "grad_norm": 2.4187111854553223,
+      "learning_rate": 2.011525116112994e-05,
+      "loss": 1.019,
+      "step": 37620
+    },
+    {
+      "epoch": 11.26,
+      "grad_norm": 2.3297410011291504,
+      "learning_rate": 2.0109489326365384e-05,
+      "loss": 1.0562,
+      "step": 37625
+    },
+    {
+      "epoch": 11.26,
+      "grad_norm": 4.753705024719238,
+      "learning_rate": 2.0103727761698636e-05,
+      "loss": 1.0245,
+      "step": 37630
+    },
+    {
+      "epoch": 11.26,
+      "grad_norm": 2.0567519664764404,
+      "learning_rate": 2.0097966467447926e-05,
+      "loss": 1.1675,
+      "step": 37635
+    },
+    {
+      "epoch": 11.26,
+      "grad_norm": 4.220091342926025,
+      "learning_rate": 2.0092205443931422e-05,
+      "loss": 0.7716,
+      "step": 37640
+    },
+    {
+      "epoch": 11.26,
+      "grad_norm": 4.092503547668457,
+      "learning_rate": 2.0086444691467325e-05,
+      "loss": 1.083,
+      "step": 37645
+    },
+    {
+      "epoch": 11.26,
+      "grad_norm": 2.5490052700042725,
+      "learning_rate": 2.0080684210373777e-05,
+      "loss": 1.3257,
+      "step": 37650
+    },
+    {
+      "epoch": 11.27,
+      "grad_norm": 2.8553218841552734,
+      "learning_rate": 2.0074924000968913e-05,
+      "loss": 0.8709,
+      "step": 37655
+    },
+    {
+      "epoch": 11.27,
+      "grad_norm": 12.186079978942871,
+      "learning_rate": 2.0069164063570896e-05,
+      "loss": 1.0308,
+      "step": 37660
+    },
+    {
+      "epoch": 11.27,
+      "grad_norm": 2.453186511993408,
+      "learning_rate": 2.0063404398497803e-05,
+      "loss": 1.2019,
+      "step": 37665
+    },
+    {
+      "epoch": 11.27,
+      "grad_norm": 1.8696728944778442,
+      "learning_rate": 2.0057645006067767e-05,
+      "loss": 0.9861,
+      "step": 37670
+    },
+    {
+      "epoch": 11.27,
+      "grad_norm": 2.764322519302368,
+      "learning_rate": 2.0051885886598855e-05,
+      "loss": 1.0721,
+      "step": 37675
+    },
+    {
+      "epoch": 11.27,
+      "grad_norm": 3.057974338531494,
+      "learning_rate": 2.0046127040409145e-05,
+      "loss": 1.0609,
+      "step": 37680
+    },
+    {
+      "epoch": 11.27,
+      "grad_norm": 1.6020811796188354,
+      "learning_rate": 2.0040368467816688e-05,
+      "loss": 0.8825,
+      "step": 37685
+    },
+    {
+      "epoch": 11.28,
+      "grad_norm": 1.326548457145691,
+      "learning_rate": 2.0034610169139527e-05,
+      "loss": 1.1104,
+      "step": 37690
+    },
+    {
+      "epoch": 11.28,
+      "grad_norm": 3.1496689319610596,
+      "learning_rate": 2.002885214469568e-05,
+      "loss": 1.106,
+      "step": 37695
+    },
+    {
+      "epoch": 11.28,
+      "grad_norm": 2.7148725986480713,
+      "learning_rate": 2.002309439480316e-05,
+      "loss": 0.9303,
+      "step": 37700
+    },
+    {
+      "epoch": 11.28,
+      "grad_norm": 2.3015408515930176,
+      "learning_rate": 2.0017336919779973e-05,
+      "loss": 0.9726,
+      "step": 37705
+    },
+    {
+      "epoch": 11.28,
+      "grad_norm": 6.830209732055664,
+      "learning_rate": 2.001157971994408e-05,
+      "loss": 1.008,
+      "step": 37710
+    },
+    {
+      "epoch": 11.28,
+      "grad_norm": 6.348620891571045,
+      "learning_rate": 2.000582279561346e-05,
+      "loss": 0.8794,
+      "step": 37715
+    },
+    {
+      "epoch": 11.29,
+      "grad_norm": 2.9782297611236572,
+      "learning_rate": 2.000006614710606e-05,
+      "loss": 0.96,
+      "step": 37720
+    },
+    {
+      "epoch": 11.29,
+      "grad_norm": 3.9869275093078613,
+      "learning_rate": 1.9994309774739797e-05,
+      "loss": 1.0065,
+      "step": 37725
+    },
+    {
+      "epoch": 11.29,
+      "grad_norm": 9.005064010620117,
+      "learning_rate": 1.9988553678832612e-05,
+      "loss": 1.1601,
+      "step": 37730
+    },
+    {
+      "epoch": 11.29,
+      "grad_norm": 3.4525146484375,
+      "learning_rate": 1.998279785970239e-05,
+      "loss": 1.2198,
+      "step": 37735
+    },
+    {
+      "epoch": 11.29,
+      "grad_norm": 3.2368991374969482,
+      "learning_rate": 1.997704231766704e-05,
+      "loss": 1.0094,
+      "step": 37740
+    },
+    {
+      "epoch": 11.29,
+      "grad_norm": 2.183854341506958,
+      "learning_rate": 1.9971287053044406e-05,
+      "loss": 1.1691,
+      "step": 37745
+    },
+    {
+      "epoch": 11.29,
+      "grad_norm": 1.5113462209701538,
+      "learning_rate": 1.9965532066152373e-05,
+      "loss": 1.0431,
+      "step": 37750
+    },
+    {
+      "epoch": 11.3,
+      "grad_norm": 3.4166901111602783,
+      "learning_rate": 1.9959777357308772e-05,
+      "loss": 0.9979,
+      "step": 37755
+    },
+    {
+      "epoch": 11.3,
+      "grad_norm": 1.2922513484954834,
+      "learning_rate": 1.9954022926831416e-05,
+      "loss": 1.0215,
+      "step": 37760
+    },
+    {
+      "epoch": 11.3,
+      "grad_norm": 3.349341630935669,
+      "learning_rate": 1.994826877503814e-05,
+      "loss": 0.9885,
+      "step": 37765
+    },
+    {
+      "epoch": 11.3,
+      "grad_norm": 9.634007453918457,
+      "learning_rate": 1.994251490224672e-05,
+      "loss": 0.9638,
+      "step": 37770
+    },
+    {
+      "epoch": 11.3,
+      "grad_norm": 2.8767809867858887,
+      "learning_rate": 1.993676130877495e-05,
+      "loss": 1.0608,
+      "step": 37775
+    },
+    {
+      "epoch": 11.3,
+      "grad_norm": 1.4088268280029297,
+      "learning_rate": 1.9931007994940592e-05,
+      "loss": 0.9073,
+      "step": 37780
+    },
+    {
+      "epoch": 11.3,
+      "grad_norm": 3.421081066131592,
+      "learning_rate": 1.9925254961061385e-05,
+      "loss": 1.1258,
+      "step": 37785
+    },
+    {
+      "epoch": 11.31,
+      "grad_norm": 3.029031991958618,
+      "learning_rate": 1.9919502207455083e-05,
+      "loss": 1.053,
+      "step": 37790
+    },
+    {
+      "epoch": 11.31,
+      "grad_norm": 3.345547676086426,
+      "learning_rate": 1.991374973443938e-05,
+      "loss": 1.0845,
+      "step": 37795
+    },
+    {
+      "epoch": 11.31,
+      "grad_norm": 3.7649970054626465,
+      "learning_rate": 1.9907997542332006e-05,
+      "loss": 0.9521,
+      "step": 37800
+    },
+    {
+      "epoch": 11.31,
+      "grad_norm": 5.937236785888672,
+      "learning_rate": 1.990224563145062e-05,
+      "loss": 1.0654,
+      "step": 37805
+    },
+    {
+      "epoch": 11.31,
+      "grad_norm": 3.4234955310821533,
+      "learning_rate": 1.9896494002112926e-05,
+      "loss": 0.981,
+      "step": 37810
+    },
+    {
+      "epoch": 11.31,
+      "grad_norm": 1.7007039785385132,
+      "learning_rate": 1.989074265463656e-05,
+      "loss": 0.8819,
+      "step": 37815
+    },
+    {
+      "epoch": 11.32,
+      "grad_norm": 1.6560670137405396,
+      "learning_rate": 1.9884991589339157e-05,
+      "loss": 0.943,
+      "step": 37820
+    },
+    {
+      "epoch": 11.32,
+      "grad_norm": 5.352659225463867,
+      "learning_rate": 1.987924080653836e-05,
+      "loss": 0.8836,
+      "step": 37825
+    },
+    {
+      "epoch": 11.32,
+      "grad_norm": 3.438709259033203,
+      "learning_rate": 1.9873490306551762e-05,
+      "loss": 1.086,
+      "step": 37830
+    },
+    {
+      "epoch": 11.32,
+      "grad_norm": 1.2173386812210083,
+      "learning_rate": 1.9867740089696976e-05,
+      "loss": 1.1148,
+      "step": 37835
+    },
+    {
+      "epoch": 11.32,
+      "grad_norm": 1.3468738794326782,
+      "learning_rate": 1.986199015629156e-05,
+      "loss": 0.9861,
+      "step": 37840
+    },
+    {
+      "epoch": 11.32,
+      "grad_norm": 3.865586280822754,
+      "learning_rate": 1.9856240506653097e-05,
+      "loss": 0.9005,
+      "step": 37845
+    },
+    {
+      "epoch": 11.32,
+      "grad_norm": 3.671863079071045,
+      "learning_rate": 1.9850491141099122e-05,
+      "loss": 1.0731,
+      "step": 37850
+    },
+    {
+      "epoch": 11.33,
+      "grad_norm": 3.499868154525757,
+      "learning_rate": 1.984474205994716e-05,
+      "loss": 1.0681,
+      "step": 37855
+    },
+    {
+      "epoch": 11.33,
+      "grad_norm": 3.519331216812134,
+      "learning_rate": 1.983899326351474e-05,
+      "loss": 1.1132,
+      "step": 37860
+    },
+    {
+      "epoch": 11.33,
+      "grad_norm": 1.4864459037780762,
+      "learning_rate": 1.983324475211936e-05,
+      "loss": 1.0481,
+      "step": 37865
+    },
+    {
+      "epoch": 11.33,
+      "grad_norm": 3.8415820598602295,
+      "learning_rate": 1.9827496526078498e-05,
+      "loss": 0.9927,
+      "step": 37870
+    },
+    {
+      "epoch": 11.33,
+      "grad_norm": 3.5681967735290527,
+      "learning_rate": 1.9821748585709634e-05,
+      "loss": 0.9556,
+      "step": 37875
+    },
+    {
+      "epoch": 11.33,
+      "grad_norm": 2.1792047023773193,
+      "learning_rate": 1.9816000931330203e-05,
+      "loss": 0.9622,
+      "step": 37880
+    },
+    {
+      "epoch": 11.33,
+      "grad_norm": 1.3530371189117432,
+      "learning_rate": 1.9810253563257662e-05,
+      "loss": 1.1555,
+      "step": 37885
+    },
+    {
+      "epoch": 11.34,
+      "grad_norm": 1.9766597747802734,
+      "learning_rate": 1.980450648180941e-05,
+      "loss": 1.093,
+      "step": 37890
+    },
+    {
+      "epoch": 11.34,
+      "grad_norm": 3.305300712585449,
+      "learning_rate": 1.9798759687302874e-05,
+      "loss": 0.9854,
+      "step": 37895
+    },
+    {
+      "epoch": 11.34,
+      "grad_norm": 1.5727300643920898,
+      "learning_rate": 1.9793013180055427e-05,
+      "loss": 0.9959,
+      "step": 37900
+    },
+    {
+      "epoch": 11.34,
+      "grad_norm": 2.8745787143707275,
+      "learning_rate": 1.9787266960384458e-05,
+      "loss": 1.2233,
+      "step": 37905
+    },
+    {
+      "epoch": 11.34,
+      "grad_norm": 1.7028470039367676,
+      "learning_rate": 1.978152102860732e-05,
+      "loss": 1.04,
+      "step": 37910
+    },
+    {
+      "epoch": 11.34,
+      "grad_norm": 7.1427717208862305,
+      "learning_rate": 1.977577538504134e-05,
+      "loss": 0.9223,
+      "step": 37915
+    },
+    {
+      "epoch": 11.35,
+      "grad_norm": 2.86995530128479,
+      "learning_rate": 1.9770030030003863e-05,
+      "loss": 1.1699,
+      "step": 37920
+    },
+    {
+      "epoch": 11.35,
+      "grad_norm": 1.6554863452911377,
+      "learning_rate": 1.9764284963812183e-05,
+      "loss": 1.1282,
+      "step": 37925
+    },
+    {
+      "epoch": 11.35,
+      "grad_norm": 2.9643630981445312,
+      "learning_rate": 1.9758540186783613e-05,
+      "loss": 0.883,
+      "step": 37930
+    },
+    {
+      "epoch": 11.35,
+      "grad_norm": 2.1681020259857178,
+      "learning_rate": 1.9752795699235406e-05,
+      "loss": 1.1171,
+      "step": 37935
+    },
+    {
+      "epoch": 11.35,
+      "grad_norm": 3.2414727210998535,
+      "learning_rate": 1.9747051501484852e-05,
+      "loss": 1.1585,
+      "step": 37940
+    },
+    {
+      "epoch": 11.35,
+      "grad_norm": 1.7487002611160278,
+      "learning_rate": 1.9741307593849178e-05,
+      "loss": 1.011,
+      "step": 37945
+    },
+    {
+      "epoch": 11.35,
+      "grad_norm": 1.7904525995254517,
+      "learning_rate": 1.9735563976645617e-05,
+      "loss": 1.0142,
+      "step": 37950
+    },
+    {
+      "epoch": 11.36,
+      "grad_norm": 3.4145519733428955,
+      "learning_rate": 1.9729820650191388e-05,
+      "loss": 1.0456,
+      "step": 37955
+    },
+    {
+      "epoch": 11.36,
+      "grad_norm": 1.48880934715271,
+      "learning_rate": 1.9724077614803686e-05,
+      "loss": 1.0409,
+      "step": 37960
+    },
+    {
+      "epoch": 11.36,
+      "grad_norm": 2.0806167125701904,
+      "learning_rate": 1.9718334870799693e-05,
+      "loss": 1.024,
+      "step": 37965
+    },
+    {
+      "epoch": 11.36,
+      "grad_norm": 3.4798760414123535,
+      "learning_rate": 1.9712592418496577e-05,
+      "loss": 1.042,
+      "step": 37970
+    },
+    {
+      "epoch": 11.36,
+      "grad_norm": 2.2736198902130127,
+      "learning_rate": 1.9706850258211485e-05,
+      "loss": 1.0028,
+      "step": 37975
+    },
+    {
+      "epoch": 11.36,
+      "grad_norm": 4.118747711181641,
+      "learning_rate": 1.9701108390261556e-05,
+      "loss": 1.0363,
+      "step": 37980
+    },
+    {
+      "epoch": 11.36,
+      "grad_norm": 26.541406631469727,
+      "learning_rate": 1.969536681496389e-05,
+      "loss": 1.1795,
+      "step": 37985
+    },
+    {
+      "epoch": 11.37,
+      "grad_norm": 2.652695417404175,
+      "learning_rate": 1.9689625532635615e-05,
+      "loss": 1.1951,
+      "step": 37990
+    },
+    {
+      "epoch": 11.37,
+      "grad_norm": 1.8068537712097168,
+      "learning_rate": 1.9683884543593788e-05,
+      "loss": 0.9577,
+      "step": 37995
+    },
+    {
+      "epoch": 11.37,
+      "grad_norm": 1.585398554801941,
+      "learning_rate": 1.9678143848155505e-05,
+      "loss": 1.0057,
+      "step": 38000
+    },
+    {
+      "epoch": 11.37,
+      "grad_norm": 4.244786739349365,
+      "learning_rate": 1.9672403446637806e-05,
+      "loss": 1.1152,
+      "step": 38005
+    },
+    {
+      "epoch": 11.37,
+      "grad_norm": 4.891877174377441,
+      "learning_rate": 1.9666663339357714e-05,
+      "loss": 1.1257,
+      "step": 38010
+    },
+    {
+      "epoch": 11.37,
+      "grad_norm": 3.488956928253174,
+      "learning_rate": 1.9660923526632275e-05,
+      "loss": 0.9834,
+      "step": 38015
+    },
+    {
+      "epoch": 11.38,
+      "grad_norm": 1.9256796836853027,
+      "learning_rate": 1.9655184008778467e-05,
+      "loss": 0.9027,
+      "step": 38020
+    },
+    {
+      "epoch": 11.38,
+      "grad_norm": 2.239374876022339,
+      "learning_rate": 1.9649444786113303e-05,
+      "loss": 1.1056,
+      "step": 38025
+    },
+    {
+      "epoch": 11.38,
+      "grad_norm": 2.374089241027832,
+      "learning_rate": 1.964370585895373e-05,
+      "loss": 0.9732,
+      "step": 38030
+    },
+    {
+      "epoch": 11.38,
+      "grad_norm": 1.6094075441360474,
+      "learning_rate": 1.9637967227616723e-05,
+      "loss": 1.1684,
+      "step": 38035
+    },
+    {
+      "epoch": 11.38,
+      "grad_norm": 2.5619702339172363,
+      "learning_rate": 1.9632228892419214e-05,
+      "loss": 1.1079,
+      "step": 38040
+    },
+    {
+      "epoch": 11.38,
+      "grad_norm": 2.8484513759613037,
+      "learning_rate": 1.962649085367812e-05,
+      "loss": 0.9438,
+      "step": 38045
+    },
+    {
+      "epoch": 11.38,
+      "grad_norm": 2.0517704486846924,
+      "learning_rate": 1.962075311171035e-05,
+      "loss": 1.1127,
+      "step": 38050
+    },
+    {
+      "epoch": 11.39,
+      "grad_norm": 4.042993545532227,
+      "learning_rate": 1.9615015666832795e-05,
+      "loss": 0.776,
+      "step": 38055
+    },
+    {
+      "epoch": 11.39,
+      "grad_norm": 3.7668004035949707,
+      "learning_rate": 1.9609278519362326e-05,
+      "loss": 1.0291,
+      "step": 38060
+    },
+    {
+      "epoch": 11.39,
+      "grad_norm": 4.582650184631348,
+      "learning_rate": 1.9603541669615796e-05,
+      "loss": 0.9049,
+      "step": 38065
+    },
+    {
+      "epoch": 11.39,
+      "grad_norm": 2.7174317836761475,
+      "learning_rate": 1.959780511791006e-05,
+      "loss": 1.0892,
+      "step": 38070
+    },
+    {
+      "epoch": 11.39,
+      "grad_norm": 3.4516286849975586,
+      "learning_rate": 1.9592068864561936e-05,
+      "loss": 1.0218,
+      "step": 38075
+    },
+    {
+      "epoch": 11.39,
+      "grad_norm": 1.96125328540802,
+      "learning_rate": 1.9586332909888216e-05,
+      "loss": 1.028,
+      "step": 38080
+    },
+    {
+      "epoch": 11.39,
+      "grad_norm": 1.8096983432769775,
+      "learning_rate": 1.9580597254205713e-05,
+      "loss": 1.0322,
+      "step": 38085
+    },
+    {
+      "epoch": 11.4,
+      "grad_norm": 5.962859153747559,
+      "learning_rate": 1.9574861897831175e-05,
+      "loss": 0.9794,
+      "step": 38090
+    },
+    {
+      "epoch": 11.4,
+      "grad_norm": 1.6979426145553589,
+      "learning_rate": 1.9569126841081392e-05,
+      "loss": 1.1257,
+      "step": 38095
+    },
+    {
+      "epoch": 11.4,
+      "grad_norm": 3.6008379459381104,
+      "learning_rate": 1.9563392084273074e-05,
+      "loss": 0.9345,
+      "step": 38100
+    },
+    {
+      "epoch": 11.4,
+      "grad_norm": 1.6322747468948364,
+      "learning_rate": 1.955765762772297e-05,
+      "loss": 0.9787,
+      "step": 38105
+    },
+    {
+      "epoch": 11.4,
+      "grad_norm": 3.6460819244384766,
+      "learning_rate": 1.955192347174778e-05,
+      "loss": 1.0202,
+      "step": 38110
+    },
+    {
+      "epoch": 11.4,
+      "grad_norm": 6.339641094207764,
+      "learning_rate": 1.9546189616664183e-05,
+      "loss": 1.0028,
+      "step": 38115
+    },
+    {
+      "epoch": 11.41,
+      "grad_norm": 5.438929557800293,
+      "learning_rate": 1.954045606278888e-05,
+      "loss": 1.1827,
+      "step": 38120
+    },
+    {
+      "epoch": 11.41,
+      "grad_norm": 1.7382398843765259,
+      "learning_rate": 1.9534722810438496e-05,
+      "loss": 1.0428,
+      "step": 38125
+    },
+    {
+      "epoch": 11.41,
+      "grad_norm": 3.361985206604004,
+      "learning_rate": 1.9528989859929704e-05,
+      "loss": 1.1343,
+      "step": 38130
+    },
+    {
+      "epoch": 11.41,
+      "grad_norm": 1.6024760007858276,
+      "learning_rate": 1.952325721157911e-05,
+      "loss": 1.0125,
+      "step": 38135
+    },
+    {
+      "epoch": 11.41,
+      "grad_norm": 5.525008201599121,
+      "learning_rate": 1.9517524865703322e-05,
+      "loss": 1.1382,
+      "step": 38140
+    },
+    {
+      "epoch": 11.41,
+      "grad_norm": 2.9428493976593018,
+      "learning_rate": 1.9511792822618947e-05,
+      "loss": 1.017,
+      "step": 38145
+    },
+    {
+      "epoch": 11.41,
+      "grad_norm": 4.371325492858887,
+      "learning_rate": 1.9506061082642538e-05,
+      "loss": 1.1891,
+      "step": 38150
+    },
+    {
+      "epoch": 11.42,
+      "grad_norm": 2.9674203395843506,
+      "learning_rate": 1.9500329646090677e-05,
+      "loss": 1.0249,
+      "step": 38155
+    },
+    {
+      "epoch": 11.42,
+      "grad_norm": 2.463282346725464,
+      "learning_rate": 1.949459851327988e-05,
+      "loss": 1.0911,
+      "step": 38160
+    },
+    {
+      "epoch": 11.42,
+      "grad_norm": 3.0271716117858887,
+      "learning_rate": 1.94888676845267e-05,
+      "loss": 1.041,
+      "step": 38165
+    },
+    {
+      "epoch": 11.42,
+      "grad_norm": 2.7793784141540527,
+      "learning_rate": 1.9483137160147626e-05,
+      "loss": 1.0276,
+      "step": 38170
+    },
+    {
+      "epoch": 11.42,
+      "grad_norm": 1.6293712854385376,
+      "learning_rate": 1.947740694045914e-05,
+      "loss": 0.9052,
+      "step": 38175
+    },
+    {
+      "epoch": 11.42,
+      "grad_norm": 2.220062017440796,
+      "learning_rate": 1.9471677025777743e-05,
+      "loss": 1.1091,
+      "step": 38180
+    },
+    {
+      "epoch": 11.42,
+      "grad_norm": 2.7983016967773438,
+      "learning_rate": 1.9465947416419867e-05,
+      "loss": 1.1761,
+      "step": 38185
+    },
+    {
+      "epoch": 11.43,
+      "grad_norm": 2.7694313526153564,
+      "learning_rate": 1.946021811270197e-05,
+      "loss": 0.9134,
+      "step": 38190
+    },
+    {
+      "epoch": 11.43,
+      "grad_norm": 4.779193878173828,
+      "learning_rate": 1.9454489114940458e-05,
+      "loss": 1.0319,
+      "step": 38195
+    },
+    {
+      "epoch": 11.43,
+      "grad_norm": 2.7941768169403076,
+      "learning_rate": 1.944876042345176e-05,
+      "loss": 1.1433,
+      "step": 38200
+    },
+    {
+      "epoch": 11.43,
+      "grad_norm": 2.305264949798584,
+      "learning_rate": 1.944303203855225e-05,
+      "loss": 1.2263,
+      "step": 38205
+    },
+    {
+      "epoch": 11.43,
+      "grad_norm": 8.440693855285645,
+      "learning_rate": 1.94373039605583e-05,
+      "loss": 1.0192,
+      "step": 38210
+    },
+    {
+      "epoch": 11.43,
+      "grad_norm": 3.348942995071411,
+      "learning_rate": 1.9431576189786276e-05,
+      "loss": 1.0054,
+      "step": 38215
+    },
+    {
+      "epoch": 11.43,
+      "grad_norm": 3.2682738304138184,
+      "learning_rate": 1.9425848726552503e-05,
+      "loss": 1.0656,
+      "step": 38220
+    },
+    {
+      "epoch": 11.44,
+      "grad_norm": 1.6970369815826416,
+      "learning_rate": 1.9420121571173315e-05,
+      "loss": 0.943,
+      "step": 38225
+    },
+    {
+      "epoch": 11.44,
+      "grad_norm": 3.24284029006958,
+      "learning_rate": 1.941439472396502e-05,
+      "loss": 1.1863,
+      "step": 38230
+    },
+    {
+      "epoch": 11.44,
+      "grad_norm": 4.399949550628662,
+      "learning_rate": 1.9408668185243885e-05,
+      "loss": 1.0284,
+      "step": 38235
+    },
+    {
+      "epoch": 11.44,
+      "grad_norm": 1.228689193725586,
+      "learning_rate": 1.9402941955326207e-05,
+      "loss": 1.1199,
+      "step": 38240
+    },
+    {
+      "epoch": 11.44,
+      "grad_norm": 2.4281301498413086,
+      "learning_rate": 1.9397216034528216e-05,
+      "loss": 1.0223,
+      "step": 38245
+    },
+    {
+      "epoch": 11.44,
+      "grad_norm": 6.723038673400879,
+      "learning_rate": 1.939149042316617e-05,
+      "loss": 1.0873,
+      "step": 38250
+    },
+    {
+      "epoch": 11.45,
+      "grad_norm": 2.267920970916748,
+      "learning_rate": 1.9385765121556266e-05,
+      "loss": 1.1129,
+      "step": 38255
+    },
+    {
+      "epoch": 11.45,
+      "grad_norm": 1.6286221742630005,
+      "learning_rate": 1.9380040130014733e-05,
+      "loss": 0.951,
+      "step": 38260
+    },
+    {
+      "epoch": 11.45,
+      "grad_norm": 5.079461574554443,
+      "learning_rate": 1.9374315448857737e-05,
+      "loss": 1.0332,
+      "step": 38265
+    },
+    {
+      "epoch": 11.45,
+      "grad_norm": 1.5790833234786987,
+      "learning_rate": 1.9368591078401442e-05,
+      "loss": 0.983,
+      "step": 38270
+    },
+    {
+      "epoch": 11.45,
+      "grad_norm": 4.2222394943237305,
+      "learning_rate": 1.9362867018962022e-05,
+      "loss": 1.0405,
+      "step": 38275
+    },
+    {
+      "epoch": 11.45,
+      "grad_norm": 1.2576323747634888,
+      "learning_rate": 1.9357143270855586e-05,
+      "loss": 1.0631,
+      "step": 38280
+    },
+    {
+      "epoch": 11.45,
+      "grad_norm": 3.4669594764709473,
+      "learning_rate": 1.935141983439827e-05,
+      "loss": 1.0043,
+      "step": 38285
+    },
+    {
+      "epoch": 11.46,
+      "grad_norm": 3.4095098972320557,
+      "learning_rate": 1.9345696709906152e-05,
+      "loss": 1.146,
+      "step": 38290
+    },
+    {
+      "epoch": 11.46,
+      "grad_norm": 2.086794137954712,
+      "learning_rate": 1.9339973897695344e-05,
+      "loss": 1.0618,
+      "step": 38295
+    },
+    {
+      "epoch": 11.46,
+      "grad_norm": 2.766907215118408,
+      "learning_rate": 1.933425139808189e-05,
+      "loss": 1.1128,
+      "step": 38300
+    },
+    {
+      "epoch": 11.46,
+      "grad_norm": 5.271542072296143,
+      "learning_rate": 1.9328529211381835e-05,
+      "loss": 1.0,
+      "step": 38305
+    },
+    {
+      "epoch": 11.46,
+      "grad_norm": 2.415320634841919,
+      "learning_rate": 1.932280733791122e-05,
+      "loss": 1.0116,
+      "step": 38310
+    },
+    {
+      "epoch": 11.46,
+      "grad_norm": 1.7332686185836792,
+      "learning_rate": 1.9317085777986054e-05,
+      "loss": 1.2115,
+      "step": 38315
+    },
+    {
+      "epoch": 11.46,
+      "grad_norm": 4.1479058265686035,
+      "learning_rate": 1.9311364531922333e-05,
+      "loss": 1.1115,
+      "step": 38320
+    },
+    {
+      "epoch": 11.47,
+      "grad_norm": 1.9687248468399048,
+      "learning_rate": 1.9305643600036037e-05,
+      "loss": 1.0981,
+      "step": 38325
+    },
+    {
+      "epoch": 11.47,
+      "grad_norm": 2.3544981479644775,
+      "learning_rate": 1.9299922982643127e-05,
+      "loss": 1.0191,
+      "step": 38330
+    },
+    {
+      "epoch": 11.47,
+      "grad_norm": 3.2068850994110107,
+      "learning_rate": 1.929420268005955e-05,
+      "loss": 0.9429,
+      "step": 38335
+    },
+    {
+      "epoch": 11.47,
+      "grad_norm": 2.9129559993743896,
+      "learning_rate": 1.9288482692601217e-05,
+      "loss": 1.004,
+      "step": 38340
+    },
+    {
+      "epoch": 11.47,
+      "grad_norm": 1.2691713571548462,
+      "learning_rate": 1.928276302058406e-05,
+      "loss": 1.013,
+      "step": 38345
+    },
+    {
+      "epoch": 11.47,
+      "grad_norm": 2.6996827125549316,
+      "learning_rate": 1.9277043664323945e-05,
+      "loss": 1.0327,
+      "step": 38350
+    },
+    {
+      "epoch": 11.48,
+      "grad_norm": 3.7437167167663574,
+      "learning_rate": 1.9271324624136775e-05,
+      "loss": 0.9692,
+      "step": 38355
+    },
+    {
+      "epoch": 11.48,
+      "grad_norm": 1.4719038009643555,
+      "learning_rate": 1.926560590033839e-05,
+      "loss": 1.1914,
+      "step": 38360
+    },
+    {
+      "epoch": 11.48,
+      "grad_norm": 2.6418561935424805,
+      "learning_rate": 1.9259887493244615e-05,
+      "loss": 1.0561,
+      "step": 38365
+    },
+    {
+      "epoch": 11.48,
+      "grad_norm": 1.7341514825820923,
+      "learning_rate": 1.9254169403171303e-05,
+      "loss": 1.1022,
+      "step": 38370
+    },
+    {
+      "epoch": 11.48,
+      "grad_norm": 2.533081531524658,
+      "learning_rate": 1.9248451630434232e-05,
+      "loss": 1.0766,
+      "step": 38375
+    },
+    {
+      "epoch": 11.48,
+      "grad_norm": 3.991834878921509,
+      "learning_rate": 1.9242734175349208e-05,
+      "loss": 1.2385,
+      "step": 38380
+    },
+    {
+      "epoch": 11.48,
+      "grad_norm": 1.5860527753829956,
+      "learning_rate": 1.923701703823198e-05,
+      "loss": 1.1093,
+      "step": 38385
+    },
+    {
+      "epoch": 11.49,
+      "grad_norm": 3.969327688217163,
+      "learning_rate": 1.923130021939832e-05,
+      "loss": 0.9571,
+      "step": 38390
+    },
+    {
+      "epoch": 11.49,
+      "grad_norm": 3.739284038543701,
+      "learning_rate": 1.922558371916395e-05,
+      "loss": 1.0152,
+      "step": 38395
+    },
+    {
+      "epoch": 11.49,
+      "grad_norm": 10.35611343383789,
+      "learning_rate": 1.9219867537844588e-05,
+      "loss": 1.0696,
+      "step": 38400
+    },
+    {
+      "epoch": 11.49,
+      "grad_norm": 2.626101016998291,
+      "learning_rate": 1.9214151675755935e-05,
+      "loss": 1.0215,
+      "step": 38405
+    },
+    {
+      "epoch": 11.49,
+      "grad_norm": 3.1883928775787354,
+      "learning_rate": 1.920843613321367e-05,
+      "loss": 1.1055,
+      "step": 38410
+    },
+    {
+      "epoch": 11.49,
+      "grad_norm": 3.719695568084717,
+      "learning_rate": 1.920272091053346e-05,
+      "loss": 0.9234,
+      "step": 38415
+    },
+    {
+      "epoch": 11.49,
+      "grad_norm": 1.9183346033096313,
+      "learning_rate": 1.9197006008030942e-05,
+      "loss": 1.0771,
+      "step": 38420
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 10.849863052368164,
+      "learning_rate": 1.9191291426021756e-05,
+      "loss": 1.1378,
+      "step": 38425
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 2.5107712745666504,
+      "learning_rate": 1.9185577164821507e-05,
+      "loss": 0.9376,
+      "step": 38430
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 3.5658061504364014,
+      "learning_rate": 1.917986322474578e-05,
+      "loss": 0.8868,
+      "step": 38435
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 4.931859493255615,
+      "learning_rate": 1.917414960611017e-05,
+      "loss": 0.9333,
+      "step": 38440
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 3.9199514389038086,
+      "learning_rate": 1.916843630923021e-05,
+      "loss": 1.1173,
+      "step": 38445
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 1.2305980920791626,
+      "learning_rate": 1.916272333442146e-05,
+      "loss": 1.0331,
+      "step": 38450
+    },
+    {
+      "epoch": 11.51,
+      "grad_norm": 1.6451231241226196,
+      "learning_rate": 1.915701068199942e-05,
+      "loss": 0.9607,
+      "step": 38455
+    },
+    {
+      "epoch": 11.51,
+      "grad_norm": 1.1699979305267334,
+      "learning_rate": 1.915129835227962e-05,
+      "loss": 0.914,
+      "step": 38460
+    },
+    {
+      "epoch": 11.51,
+      "grad_norm": 2.1757733821868896,
+      "learning_rate": 1.9145586345577533e-05,
+      "loss": 1.1439,
+      "step": 38465
+    },
+    {
+      "epoch": 11.51,
+      "grad_norm": 1.6203327178955078,
+      "learning_rate": 1.9139874662208615e-05,
+      "loss": 1.2016,
+      "step": 38470
+    },
+    {
+      "epoch": 11.51,
+      "grad_norm": 3.481635093688965,
+      "learning_rate": 1.913416330248834e-05,
+      "loss": 0.7547,
+      "step": 38475
+    },
+    {
+      "epoch": 11.51,
+      "grad_norm": 2.200890064239502,
+      "learning_rate": 1.9128452266732115e-05,
+      "loss": 1.0624,
+      "step": 38480
+    },
+    {
+      "epoch": 11.51,
+      "grad_norm": 1.81036376953125,
+      "learning_rate": 1.9122741555255384e-05,
+      "loss": 0.9228,
+      "step": 38485
+    },
+    {
+      "epoch": 11.52,
+      "grad_norm": 4.309178829193115,
+      "learning_rate": 1.911703116837352e-05,
+      "loss": 1.0101,
+      "step": 38490
+    },
+    {
+      "epoch": 11.52,
+      "grad_norm": 1.9071286916732788,
+      "learning_rate": 1.911132110640191e-05,
+      "loss": 1.0724,
+      "step": 38495
+    },
+    {
+      "epoch": 11.52,
+      "grad_norm": 2.7584476470947266,
+      "learning_rate": 1.9105611369655914e-05,
+      "loss": 1.0121,
+      "step": 38500
+    },
+    {
+      "epoch": 11.52,
+      "grad_norm": 3.3853213787078857,
+      "learning_rate": 1.9099901958450868e-05,
+      "loss": 0.9683,
+      "step": 38505
+    },
+    {
+      "epoch": 11.52,
+      "grad_norm": 2.5267248153686523,
+      "learning_rate": 1.9094192873102117e-05,
+      "loss": 1.0502,
+      "step": 38510
+    },
+    {
+      "epoch": 11.52,
+      "grad_norm": 3.773902416229248,
+      "learning_rate": 1.9088484113924944e-05,
+      "loss": 1.029,
+      "step": 38515
+    },
+    {
+      "epoch": 11.52,
+      "grad_norm": 2.5307250022888184,
+      "learning_rate": 1.908277568123466e-05,
+      "loss": 0.9967,
+      "step": 38520
+    },
+    {
+      "epoch": 11.53,
+      "grad_norm": 4.490280628204346,
+      "learning_rate": 1.907706757534652e-05,
+      "loss": 1.0186,
+      "step": 38525
+    },
+    {
+      "epoch": 11.53,
+      "grad_norm": 1.3024266958236694,
+      "learning_rate": 1.9071359796575774e-05,
+      "loss": 1.1027,
+      "step": 38530
+    },
+    {
+      "epoch": 11.53,
+      "grad_norm": 5.981440544128418,
+      "learning_rate": 1.906565234523767e-05,
+      "loss": 1.1221,
+      "step": 38535
+    },
+    {
+      "epoch": 11.53,
+      "grad_norm": 2.8567864894866943,
+      "learning_rate": 1.9059945221647414e-05,
+      "loss": 1.1096,
+      "step": 38540
+    },
+    {
+      "epoch": 11.53,
+      "grad_norm": 2.297405481338501,
+      "learning_rate": 1.9054238426120218e-05,
+      "loss": 1.0762,
+      "step": 38545
+    },
+    {
+      "epoch": 11.53,
+      "grad_norm": 1.9504579305648804,
+      "learning_rate": 1.9048531958971245e-05,
+      "loss": 1.1548,
+      "step": 38550
+    },
+    {
+      "epoch": 11.54,
+      "grad_norm": 4.491879463195801,
+      "learning_rate": 1.9042825820515675e-05,
+      "loss": 0.8179,
+      "step": 38555
+    },
+    {
+      "epoch": 11.54,
+      "grad_norm": 3.1393933296203613,
+      "learning_rate": 1.903712001106864e-05,
+      "loss": 0.8162,
+      "step": 38560
+    },
+    {
+      "epoch": 11.54,
+      "grad_norm": 1.3722354173660278,
+      "learning_rate": 1.9031414530945263e-05,
+      "loss": 1.096,
+      "step": 38565
+    },
+    {
+      "epoch": 11.54,
+      "grad_norm": 3.8262624740600586,
+      "learning_rate": 1.9025709380460662e-05,
+      "loss": 0.9238,
+      "step": 38570
+    },
+    {
+      "epoch": 11.54,
+      "grad_norm": 3.3715012073516846,
+      "learning_rate": 1.9020004559929915e-05,
+      "loss": 1.0101,
+      "step": 38575
+    },
+    {
+      "epoch": 11.54,
+      "grad_norm": 4.317600250244141,
+      "learning_rate": 1.9014300069668113e-05,
+      "loss": 0.9422,
+      "step": 38580
+    },
+    {
+      "epoch": 11.54,
+      "grad_norm": 3.4299378395080566,
+      "learning_rate": 1.9008595909990286e-05,
+      "loss": 1.0151,
+      "step": 38585
+    },
+    {
+      "epoch": 11.55,
+      "grad_norm": 5.137393474578857,
+      "learning_rate": 1.9002892081211478e-05,
+      "loss": 1.1151,
+      "step": 38590
+    },
+    {
+      "epoch": 11.55,
+      "grad_norm": 2.469620704650879,
+      "learning_rate": 1.8997188583646713e-05,
+      "loss": 1.0459,
+      "step": 38595
+    },
+    {
+      "epoch": 11.55,
+      "grad_norm": 1.945117473602295,
+      "learning_rate": 1.899148541761098e-05,
+      "loss": 1.037,
+      "step": 38600
+    },
+    {
+      "epoch": 11.55,
+      "grad_norm": 2.0627458095550537,
+      "learning_rate": 1.8985782583419266e-05,
+      "loss": 1.0576,
+      "step": 38605
+    },
+    {
+      "epoch": 11.55,
+      "grad_norm": 4.658207416534424,
+      "learning_rate": 1.8980080081386517e-05,
+      "loss": 0.9801,
+      "step": 38610
+    },
+    {
+      "epoch": 11.55,
+      "grad_norm": 3.8801674842834473,
+      "learning_rate": 1.89743779118277e-05,
+      "loss": 0.9549,
+      "step": 38615
+    },
+    {
+      "epoch": 11.55,
+      "grad_norm": 2.336965322494507,
+      "learning_rate": 1.896867607505773e-05,
+      "loss": 1.236,
+      "step": 38620
+    },
+    {
+      "epoch": 11.56,
+      "grad_norm": 2.502366304397583,
+      "learning_rate": 1.89629745713915e-05,
+      "loss": 0.8492,
+      "step": 38625
+    },
+    {
+      "epoch": 11.56,
+      "grad_norm": 3.108949661254883,
+      "learning_rate": 1.895727340114392e-05,
+      "loss": 0.9725,
+      "step": 38630
+    },
+    {
+      "epoch": 11.56,
+      "grad_norm": 4.816553115844727,
+      "learning_rate": 1.8951572564629835e-05,
+      "loss": 1.0565,
+      "step": 38635
+    },
+    {
+      "epoch": 11.56,
+      "grad_norm": 2.322685956954956,
+      "learning_rate": 1.8945872062164122e-05,
+      "loss": 0.9627,
+      "step": 38640
+    },
+    {
+      "epoch": 11.56,
+      "grad_norm": 2.1832380294799805,
+      "learning_rate": 1.894017189406159e-05,
+      "loss": 1.0276,
+      "step": 38645
+    },
+    {
+      "epoch": 11.56,
+      "grad_norm": 2.5433905124664307,
+      "learning_rate": 1.8934472060637083e-05,
+      "loss": 1.0334,
+      "step": 38650
+    },
+    {
+      "epoch": 11.57,
+      "grad_norm": 1.171791672706604,
+      "learning_rate": 1.8928772562205373e-05,
+      "loss": 1.0299,
+      "step": 38655
+    },
+    {
+      "epoch": 11.57,
+      "grad_norm": 12.511678695678711,
+      "learning_rate": 1.8923073399081238e-05,
+      "loss": 0.8337,
+      "step": 38660
+    },
+    {
+      "epoch": 11.57,
+      "grad_norm": 2.81550669670105,
+      "learning_rate": 1.8917374571579444e-05,
+      "loss": 1.1925,
+      "step": 38665
+    },
+    {
+      "epoch": 11.57,
+      "grad_norm": 1.520538568496704,
+      "learning_rate": 1.891167608001473e-05,
+      "loss": 0.9966,
+      "step": 38670
+    },
+    {
+      "epoch": 11.57,
+      "grad_norm": 6.8623456954956055,
+      "learning_rate": 1.890597792470182e-05,
+      "loss": 1.1197,
+      "step": 38675
+    },
+    {
+      "epoch": 11.57,
+      "grad_norm": 2.9777519702911377,
+      "learning_rate": 1.890028010595541e-05,
+      "loss": 1.0691,
+      "step": 38680
+    },
+    {
+      "epoch": 11.57,
+      "grad_norm": 5.088373184204102,
+      "learning_rate": 1.8894582624090194e-05,
+      "loss": 0.8929,
+      "step": 38685
+    },
+    {
+      "epoch": 11.58,
+      "grad_norm": 3.7132937908172607,
+      "learning_rate": 1.8888885479420838e-05,
+      "loss": 1.0176,
+      "step": 38690
+    },
+    {
+      "epoch": 11.58,
+      "grad_norm": 3.316521167755127,
+      "learning_rate": 1.8883188672261972e-05,
+      "loss": 1.0659,
+      "step": 38695
+    },
+    {
+      "epoch": 11.58,
+      "grad_norm": 2.1407995223999023,
+      "learning_rate": 1.8877492202928252e-05,
+      "loss": 1.0775,
+      "step": 38700
+    },
+    {
+      "epoch": 11.58,
+      "grad_norm": 2.758033037185669,
+      "learning_rate": 1.887179607173426e-05,
+      "loss": 1.0208,
+      "step": 38705
+    },
+    {
+      "epoch": 11.58,
+      "grad_norm": 3.8762590885162354,
+      "learning_rate": 1.8866100278994613e-05,
+      "loss": 0.9107,
+      "step": 38710
+    },
+    {
+      "epoch": 11.58,
+      "grad_norm": 2.9407236576080322,
+      "learning_rate": 1.886040482502387e-05,
+      "loss": 1.0697,
+      "step": 38715
+    },
+    {
+      "epoch": 11.58,
+      "grad_norm": 2.600522994995117,
+      "learning_rate": 1.885470971013658e-05,
+      "loss": 1.0183,
+      "step": 38720
+    },
+    {
+      "epoch": 11.59,
+      "grad_norm": 3.5335729122161865,
+      "learning_rate": 1.884901493464729e-05,
+      "loss": 0.9398,
+      "step": 38725
+    },
+    {
+      "epoch": 11.59,
+      "grad_norm": 1.7185178995132446,
+      "learning_rate": 1.8843320498870504e-05,
+      "loss": 0.8392,
+      "step": 38730
+    },
+    {
+      "epoch": 11.59,
+      "grad_norm": 2.37362003326416,
+      "learning_rate": 1.883762640312074e-05,
+      "loss": 1.0308,
+      "step": 38735
+    },
+    {
+      "epoch": 11.59,
+      "grad_norm": 5.628195285797119,
+      "learning_rate": 1.883193264771245e-05,
+      "loss": 0.9699,
+      "step": 38740
+    },
+    {
+      "epoch": 11.59,
+      "grad_norm": 2.0216867923736572,
+      "learning_rate": 1.882623923296012e-05,
+      "loss": 1.0402,
+      "step": 38745
+    },
+    {
+      "epoch": 11.59,
+      "grad_norm": 4.347649097442627,
+      "learning_rate": 1.8820546159178175e-05,
+      "loss": 0.9695,
+      "step": 38750
+    },
+    {
+      "epoch": 11.6,
+      "grad_norm": 5.039327621459961,
+      "learning_rate": 1.8814853426681046e-05,
+      "loss": 0.9951,
+      "step": 38755
+    },
+    {
+      "epoch": 11.6,
+      "grad_norm": 3.6531269550323486,
+      "learning_rate": 1.880916103578313e-05,
+      "loss": 1.0957,
+      "step": 38760
+    },
+    {
+      "epoch": 11.6,
+      "grad_norm": 2.0845232009887695,
+      "learning_rate": 1.8803468986798814e-05,
+      "loss": 1.05,
+      "step": 38765
+    },
+    {
+      "epoch": 11.6,
+      "grad_norm": 4.7361979484558105,
+      "learning_rate": 1.8797777280042467e-05,
+      "loss": 0.9722,
+      "step": 38770
+    },
+    {
+      "epoch": 11.6,
+      "grad_norm": 4.363923072814941,
+      "learning_rate": 1.879208591582843e-05,
+      "loss": 0.9497,
+      "step": 38775
+    },
+    {
+      "epoch": 11.6,
+      "grad_norm": 3.8049824237823486,
+      "learning_rate": 1.8786394894471038e-05,
+      "loss": 1.1666,
+      "step": 38780
+    },
+    {
+      "epoch": 11.6,
+      "grad_norm": 4.981505870819092,
+      "learning_rate": 1.8780704216284604e-05,
+      "loss": 1.0055,
+      "step": 38785
+    },
+    {
+      "epoch": 11.61,
+      "grad_norm": 2.457186222076416,
+      "learning_rate": 1.87750138815834e-05,
+      "loss": 1.1097,
+      "step": 38790
+    },
+    {
+      "epoch": 11.61,
+      "grad_norm": 2.6897389888763428,
+      "learning_rate": 1.8769323890681717e-05,
+      "loss": 1.2053,
+      "step": 38795
+    },
+    {
+      "epoch": 11.61,
+      "grad_norm": 2.4741759300231934,
+      "learning_rate": 1.876363424389379e-05,
+      "loss": 1.0683,
+      "step": 38800
+    },
+    {
+      "epoch": 11.61,
+      "grad_norm": 3.079468011856079,
+      "learning_rate": 1.875794494153387e-05,
+      "loss": 0.8789,
+      "step": 38805
+    },
+    {
+      "epoch": 11.61,
+      "grad_norm": 4.948914051055908,
+      "learning_rate": 1.875225598391616e-05,
+      "loss": 0.9235,
+      "step": 38810
+    },
+    {
+      "epoch": 11.61,
+      "grad_norm": 9.49067211151123,
+      "learning_rate": 1.874656737135486e-05,
+      "loss": 0.7728,
+      "step": 38815
+    },
+    {
+      "epoch": 11.61,
+      "grad_norm": 4.153932094573975,
+      "learning_rate": 1.874087910416415e-05,
+      "loss": 0.8923,
+      "step": 38820
+    },
+    {
+      "epoch": 11.62,
+      "grad_norm": 5.080899715423584,
+      "learning_rate": 1.8735191182658164e-05,
+      "loss": 1.1195,
+      "step": 38825
+    },
+    {
+      "epoch": 11.62,
+      "grad_norm": 14.455235481262207,
+      "learning_rate": 1.872950360715107e-05,
+      "loss": 1.0142,
+      "step": 38830
+    },
+    {
+      "epoch": 11.62,
+      "grad_norm": 2.518566131591797,
+      "learning_rate": 1.8723816377956966e-05,
+      "loss": 1.0349,
+      "step": 38835
+    },
+    {
+      "epoch": 11.62,
+      "grad_norm": 1.6086268424987793,
+      "learning_rate": 1.871812949538997e-05,
+      "loss": 1.0412,
+      "step": 38840
+    },
+    {
+      "epoch": 11.62,
+      "grad_norm": 3.0783514976501465,
+      "learning_rate": 1.8712442959764144e-05,
+      "loss": 1.0691,
+      "step": 38845
+    },
+    {
+      "epoch": 11.62,
+      "grad_norm": 2.9750258922576904,
+      "learning_rate": 1.870675677139356e-05,
+      "loss": 0.8981,
+      "step": 38850
+    },
+    {
+      "epoch": 11.62,
+      "grad_norm": 1.0562318563461304,
+      "learning_rate": 1.8701070930592257e-05,
+      "loss": 1.0965,
+      "step": 38855
+    },
+    {
+      "epoch": 11.63,
+      "grad_norm": 3.368023157119751,
+      "learning_rate": 1.8695385437674263e-05,
+      "loss": 1.0257,
+      "step": 38860
+    },
+    {
+      "epoch": 11.63,
+      "grad_norm": 2.06657338142395,
+      "learning_rate": 1.868970029295357e-05,
+      "loss": 1.0595,
+      "step": 38865
+    },
+    {
+      "epoch": 11.63,
+      "grad_norm": 3.3948874473571777,
+      "learning_rate": 1.8684015496744173e-05,
+      "loss": 1.0824,
+      "step": 38870
+    },
+    {
+      "epoch": 11.63,
+      "grad_norm": 3.1580018997192383,
+      "learning_rate": 1.8678331049360044e-05,
+      "loss": 0.994,
+      "step": 38875
+    },
+    {
+      "epoch": 11.63,
+      "grad_norm": 2.4036195278167725,
+      "learning_rate": 1.8672646951115116e-05,
+      "loss": 1.1658,
+      "step": 38880
+    },
+    {
+      "epoch": 11.63,
+      "grad_norm": 1.6055760383605957,
+      "learning_rate": 1.866696320232331e-05,
+      "loss": 1.0514,
+      "step": 38885
+    },
+    {
+      "epoch": 11.64,
+      "grad_norm": 2.9293153285980225,
+      "learning_rate": 1.866127980329856e-05,
+      "loss": 1.0034,
+      "step": 38890
+    },
+    {
+      "epoch": 11.64,
+      "grad_norm": 6.062731742858887,
+      "learning_rate": 1.865559675435472e-05,
+      "loss": 0.9402,
+      "step": 38895
+    },
+    {
+      "epoch": 11.64,
+      "grad_norm": 2.678724527359009,
+      "learning_rate": 1.8649914055805696e-05,
+      "loss": 0.8018,
+      "step": 38900
+    },
+    {
+      "epoch": 11.64,
+      "grad_norm": 1.7814666032791138,
+      "learning_rate": 1.8644231707965297e-05,
+      "loss": 1.0356,
+      "step": 38905
+    },
+    {
+      "epoch": 11.64,
+      "grad_norm": 2.179852247238159,
+      "learning_rate": 1.863854971114739e-05,
+      "loss": 1.0481,
+      "step": 38910
+    },
+    {
+      "epoch": 11.64,
+      "grad_norm": 3.8820948600769043,
+      "learning_rate": 1.863286806566577e-05,
+      "loss": 1.0233,
+      "step": 38915
+    },
+    {
+      "epoch": 11.64,
+      "grad_norm": 4.843132019042969,
+      "learning_rate": 1.862718677183422e-05,
+      "loss": 1.1038,
+      "step": 38920
+    },
+    {
+      "epoch": 11.65,
+      "grad_norm": 2.708319902420044,
+      "learning_rate": 1.862150582996653e-05,
+      "loss": 1.1368,
+      "step": 38925
+    },
+    {
+      "epoch": 11.65,
+      "grad_norm": 2.447589159011841,
+      "learning_rate": 1.861582524037643e-05,
+      "loss": 1.096,
+      "step": 38930
+    },
+    {
+      "epoch": 11.65,
+      "grad_norm": 2.1640899181365967,
+      "learning_rate": 1.8610145003377676e-05,
+      "loss": 0.9314,
+      "step": 38935
+    },
+    {
+      "epoch": 11.65,
+      "grad_norm": 6.040550708770752,
+      "learning_rate": 1.860446511928397e-05,
+      "loss": 1.1117,
+      "step": 38940
+    },
+    {
+      "epoch": 11.65,
+      "grad_norm": 18.708223342895508,
+      "learning_rate": 1.8598785588409005e-05,
+      "loss": 1.0053,
+      "step": 38945
+    },
+    {
+      "epoch": 11.65,
+      "grad_norm": 2.9335033893585205,
+      "learning_rate": 1.859310641106646e-05,
+      "loss": 1.2107,
+      "step": 38950
+    },
+    {
+      "epoch": 11.65,
+      "grad_norm": 3.3120362758636475,
+      "learning_rate": 1.8587427587569982e-05,
+      "loss": 1.193,
+      "step": 38955
+    },
+    {
+      "epoch": 11.66,
+      "grad_norm": 1.8769471645355225,
+      "learning_rate": 1.8581749118233225e-05,
+      "loss": 1.1355,
+      "step": 38960
+    },
+    {
+      "epoch": 11.66,
+      "grad_norm": 3.2708353996276855,
+      "learning_rate": 1.857607100336978e-05,
+      "loss": 0.9209,
+      "step": 38965
+    },
+    {
+      "epoch": 11.66,
+      "grad_norm": 2.209517478942871,
+      "learning_rate": 1.857039324329327e-05,
+      "loss": 1.1848,
+      "step": 38970
+    },
+    {
+      "epoch": 11.66,
+      "grad_norm": 2.2054500579833984,
+      "learning_rate": 1.8564715838317252e-05,
+      "loss": 1.0466,
+      "step": 38975
+    },
+    {
+      "epoch": 11.66,
+      "grad_norm": 1.8774044513702393,
+      "learning_rate": 1.8559038788755285e-05,
+      "loss": 0.9097,
+      "step": 38980
+    },
+    {
+      "epoch": 11.66,
+      "grad_norm": 2.7075273990631104,
+      "learning_rate": 1.8553362094920918e-05,
+      "loss": 1.1306,
+      "step": 38985
+    },
+    {
+      "epoch": 11.67,
+      "grad_norm": 3.6890761852264404,
+      "learning_rate": 1.854768575712765e-05,
+      "loss": 1.0088,
+      "step": 38990
+    },
+    {
+      "epoch": 11.67,
+      "grad_norm": 4.63300085067749,
+      "learning_rate": 1.8542009775689003e-05,
+      "loss": 1.1951,
+      "step": 38995
+    },
+    {
+      "epoch": 11.67,
+      "grad_norm": 2.1743202209472656,
+      "learning_rate": 1.8536334150918432e-05,
+      "loss": 1.0896,
+      "step": 39000
+    },
+    {
+      "epoch": 11.67,
+      "grad_norm": 2.1397712230682373,
+      "learning_rate": 1.853065888312942e-05,
+      "loss": 0.9551,
+      "step": 39005
+    },
+    {
+      "epoch": 11.67,
+      "grad_norm": 3.1289427280426025,
+      "learning_rate": 1.852498397263539e-05,
+      "loss": 1.0159,
+      "step": 39010
+    },
+    {
+      "epoch": 11.67,
+      "grad_norm": 2.917625904083252,
+      "learning_rate": 1.8519309419749757e-05,
+      "loss": 1.0579,
+      "step": 39015
+    },
+    {
+      "epoch": 11.67,
+      "grad_norm": 1.9072588682174683,
+      "learning_rate": 1.851363522478594e-05,
+      "loss": 1.0325,
+      "step": 39020
+    },
+    {
+      "epoch": 11.68,
+      "grad_norm": 1.4245704412460327,
+      "learning_rate": 1.8507961388057292e-05,
+      "loss": 1.1429,
+      "step": 39025
+    },
+    {
+      "epoch": 11.68,
+      "grad_norm": 5.188929080963135,
+      "learning_rate": 1.8502287909877197e-05,
+      "loss": 1.1144,
+      "step": 39030
+    },
+    {
+      "epoch": 11.68,
+      "grad_norm": 6.824649333953857,
+      "learning_rate": 1.8496614790558987e-05,
+      "loss": 0.9599,
+      "step": 39035
+    },
+    {
+      "epoch": 11.68,
+      "grad_norm": 5.712022304534912,
+      "learning_rate": 1.8490942030415982e-05,
+      "loss": 1.105,
+      "step": 39040
+    },
+    {
+      "epoch": 11.68,
+      "grad_norm": 5.067729949951172,
+      "learning_rate": 1.8485269629761487e-05,
+      "loss": 1.0556,
+      "step": 39045
+    },
+    {
+      "epoch": 11.68,
+      "grad_norm": 1.9118393659591675,
+      "learning_rate": 1.847959758890877e-05,
+      "loss": 1.129,
+      "step": 39050
+    },
+    {
+      "epoch": 11.68,
+      "grad_norm": 2.5474183559417725,
+      "learning_rate": 1.847392590817111e-05,
+      "loss": 0.9264,
+      "step": 39055
+    },
+    {
+      "epoch": 11.69,
+      "grad_norm": 3.004509925842285,
+      "learning_rate": 1.8468254587861728e-05,
+      "loss": 0.9561,
+      "step": 39060
+    },
+    {
+      "epoch": 11.69,
+      "grad_norm": 2.7860445976257324,
+      "learning_rate": 1.8462583628293867e-05,
+      "loss": 1.0989,
+      "step": 39065
+    },
+    {
+      "epoch": 11.69,
+      "grad_norm": 3.1194570064544678,
+      "learning_rate": 1.8456913029780714e-05,
+      "loss": 0.9351,
+      "step": 39070
+    },
+    {
+      "epoch": 11.69,
+      "grad_norm": 2.584099054336548,
+      "learning_rate": 1.8451242792635447e-05,
+      "loss": 1.0934,
+      "step": 39075
+    },
+    {
+      "epoch": 11.69,
+      "grad_norm": 4.399397850036621,
+      "learning_rate": 1.8445572917171238e-05,
+      "loss": 1.0019,
+      "step": 39080
+    },
+    {
+      "epoch": 11.69,
+      "grad_norm": 1.7044438123703003,
+      "learning_rate": 1.843990340370122e-05,
+      "loss": 1.1474,
+      "step": 39085
+    },
+    {
+      "epoch": 11.7,
+      "grad_norm": 1.9364503622055054,
+      "learning_rate": 1.8434234252538523e-05,
+      "loss": 1.0992,
+      "step": 39090
+    },
+    {
+      "epoch": 11.7,
+      "grad_norm": 2.9224693775177,
+      "learning_rate": 1.842856546399624e-05,
+      "loss": 0.8731,
+      "step": 39095
+    },
+    {
+      "epoch": 11.7,
+      "grad_norm": 2.569241523742676,
+      "learning_rate": 1.842289703838746e-05,
+      "loss": 1.0233,
+      "step": 39100
+    },
+    {
+      "epoch": 11.7,
+      "grad_norm": 6.116538047790527,
+      "learning_rate": 1.841722897602524e-05,
+      "loss": 0.9919,
+      "step": 39105
+    },
+    {
+      "epoch": 11.7,
+      "grad_norm": 5.045174598693848,
+      "learning_rate": 1.8411561277222616e-05,
+      "loss": 0.9874,
+      "step": 39110
+    },
+    {
+      "epoch": 11.7,
+      "grad_norm": 4.61328125,
+      "learning_rate": 1.8405893942292617e-05,
+      "loss": 0.913,
+      "step": 39115
+    },
+    {
+      "epoch": 11.7,
+      "grad_norm": 3.24153470993042,
+      "learning_rate": 1.840022697154824e-05,
+      "loss": 0.908,
+      "step": 39120
+    },
+    {
+      "epoch": 11.71,
+      "grad_norm": 1.127431869506836,
+      "learning_rate": 1.839456036530247e-05,
+      "loss": 1.0368,
+      "step": 39125
+    },
+    {
+      "epoch": 11.71,
+      "grad_norm": 2.7838566303253174,
+      "learning_rate": 1.838889412386826e-05,
+      "loss": 0.9142,
+      "step": 39130
+    },
+    {
+      "epoch": 11.71,
+      "grad_norm": 2.489769458770752,
+      "learning_rate": 1.838322824755856e-05,
+      "loss": 1.0572,
+      "step": 39135
+    },
+    {
+      "epoch": 11.71,
+      "grad_norm": 2.236867904663086,
+      "learning_rate": 1.837756273668629e-05,
+      "loss": 1.0315,
+      "step": 39140
+    },
+    {
+      "epoch": 11.71,
+      "grad_norm": 1.7803845405578613,
+      "learning_rate": 1.8371897591564335e-05,
+      "loss": 0.9635,
+      "step": 39145
+    },
+    {
+      "epoch": 11.71,
+      "grad_norm": 19.32113265991211,
+      "learning_rate": 1.8366232812505602e-05,
+      "loss": 1.2016,
+      "step": 39150
+    },
+    {
+      "epoch": 11.71,
+      "grad_norm": 3.2473385334014893,
+      "learning_rate": 1.836056839982292e-05,
+      "loss": 0.95,
+      "step": 39155
+    },
+    {
+      "epoch": 11.72,
+      "grad_norm": 3.1559784412384033,
+      "learning_rate": 1.8354904353829156e-05,
+      "loss": 0.9813,
+      "step": 39160
+    },
+    {
+      "epoch": 11.72,
+      "grad_norm": 2.750032663345337,
+      "learning_rate": 1.8349240674837105e-05,
+      "loss": 0.8542,
+      "step": 39165
+    },
+    {
+      "epoch": 11.72,
+      "grad_norm": 5.297883033752441,
+      "learning_rate": 1.834357736315959e-05,
+      "loss": 0.9693,
+      "step": 39170
+    },
+    {
+      "epoch": 11.72,
+      "grad_norm": 3.035358190536499,
+      "learning_rate": 1.8337914419109383e-05,
+      "loss": 1.0519,
+      "step": 39175
+    },
+    {
+      "epoch": 11.72,
+      "grad_norm": 5.685046195983887,
+      "learning_rate": 1.833225184299922e-05,
+      "loss": 0.9673,
+      "step": 39180
+    },
+    {
+      "epoch": 11.72,
+      "grad_norm": 4.552941799163818,
+      "learning_rate": 1.8326589635141874e-05,
+      "loss": 1.1577,
+      "step": 39185
+    },
+    {
+      "epoch": 11.73,
+      "grad_norm": 11.873933792114258,
+      "learning_rate": 1.832092779585003e-05,
+      "loss": 1.0136,
+      "step": 39190
+    },
+    {
+      "epoch": 11.73,
+      "grad_norm": 1.8182165622711182,
+      "learning_rate": 1.8315266325436412e-05,
+      "loss": 1.2264,
+      "step": 39195
+    },
+    {
+      "epoch": 11.73,
+      "grad_norm": 4.9064621925354,
+      "learning_rate": 1.8309605224213682e-05,
+      "loss": 1.0442,
+      "step": 39200
+    },
+    {
+      "epoch": 11.73,
+      "grad_norm": 4.886104106903076,
+      "learning_rate": 1.83039444924945e-05,
+      "loss": 0.9959,
+      "step": 39205
+    },
+    {
+      "epoch": 11.73,
+      "grad_norm": 2.1193058490753174,
+      "learning_rate": 1.8298284130591508e-05,
+      "loss": 0.9811,
+      "step": 39210
+    },
+    {
+      "epoch": 11.73,
+      "grad_norm": 2.320329427719116,
+      "learning_rate": 1.8292624138817317e-05,
+      "loss": 1.1301,
+      "step": 39215
+    },
+    {
+      "epoch": 11.73,
+      "grad_norm": 8.177576065063477,
+      "learning_rate": 1.8286964517484518e-05,
+      "loss": 1.0834,
+      "step": 39220
+    },
+    {
+      "epoch": 11.74,
+      "grad_norm": 3.5780060291290283,
+      "learning_rate": 1.8281305266905697e-05,
+      "loss": 0.9501,
+      "step": 39225
+    },
+    {
+      "epoch": 11.74,
+      "grad_norm": 4.063993453979492,
+      "learning_rate": 1.8275646387393396e-05,
+      "loss": 1.252,
+      "step": 39230
+    },
+    {
+      "epoch": 11.74,
+      "grad_norm": 1.9857643842697144,
+      "learning_rate": 1.8269987879260163e-05,
+      "loss": 1.1019,
+      "step": 39235
+    },
+    {
+      "epoch": 11.74,
+      "grad_norm": 1.7458826303482056,
+      "learning_rate": 1.8264329742818496e-05,
+      "loss": 1.0769,
+      "step": 39240
+    },
+    {
+      "epoch": 11.74,
+      "grad_norm": 1.2692326307296753,
+      "learning_rate": 1.8258671978380902e-05,
+      "loss": 1.1112,
+      "step": 39245
+    },
+    {
+      "epoch": 11.74,
+      "grad_norm": 1.531681776046753,
+      "learning_rate": 1.8253014586259842e-05,
+      "loss": 1.0832,
+      "step": 39250
+    },
+    {
+      "epoch": 11.74,
+      "grad_norm": 3.519178628921509,
+      "learning_rate": 1.824735756676778e-05,
+      "loss": 1.0214,
+      "step": 39255
+    },
+    {
+      "epoch": 11.75,
+      "grad_norm": 1.5379389524459839,
+      "learning_rate": 1.824170092021713e-05,
+      "loss": 0.942,
+      "step": 39260
+    },
+    {
+      "epoch": 11.75,
+      "grad_norm": 1.4689420461654663,
+      "learning_rate": 1.823604464692033e-05,
+      "loss": 1.0427,
+      "step": 39265
+    },
+    {
+      "epoch": 11.75,
+      "grad_norm": 1.6258187294006348,
+      "learning_rate": 1.8230388747189746e-05,
+      "loss": 1.0556,
+      "step": 39270
+    },
+    {
+      "epoch": 11.75,
+      "grad_norm": 1.4137862920761108,
+      "learning_rate": 1.8224733221337748e-05,
+      "loss": 1.2594,
+      "step": 39275
+    },
+    {
+      "epoch": 11.75,
+      "grad_norm": 3.4396555423736572,
+      "learning_rate": 1.82190780696767e-05,
+      "loss": 1.1517,
+      "step": 39280
+    },
+    {
+      "epoch": 11.75,
+      "grad_norm": 2.3807373046875,
+      "learning_rate": 1.8213423292518912e-05,
+      "loss": 0.9228,
+      "step": 39285
+    },
+    {
+      "epoch": 11.76,
+      "grad_norm": 3.9610815048217773,
+      "learning_rate": 1.8207768890176714e-05,
+      "loss": 0.975,
+      "step": 39290
+    },
+    {
+      "epoch": 11.76,
+      "grad_norm": 2.3501439094543457,
+      "learning_rate": 1.8202114862962376e-05,
+      "loss": 1.0386,
+      "step": 39295
+    },
+    {
+      "epoch": 11.76,
+      "grad_norm": 2.7657525539398193,
+      "learning_rate": 1.8196461211188164e-05,
+      "loss": 1.0143,
+      "step": 39300
+    },
+    {
+      "epoch": 11.76,
+      "grad_norm": 1.5906193256378174,
+      "learning_rate": 1.8190807935166327e-05,
+      "loss": 1.1036,
+      "step": 39305
+    },
+    {
+      "epoch": 11.76,
+      "grad_norm": 26.443702697753906,
+      "learning_rate": 1.8185155035209096e-05,
+      "loss": 0.9669,
+      "step": 39310
+    },
+    {
+      "epoch": 11.76,
+      "grad_norm": 4.151071548461914,
+      "learning_rate": 1.817950251162866e-05,
+      "loss": 1.0132,
+      "step": 39315
+    },
+    {
+      "epoch": 11.76,
+      "grad_norm": 2.691179037094116,
+      "learning_rate": 1.817385036473721e-05,
+      "loss": 1.0296,
+      "step": 39320
+    },
+    {
+      "epoch": 11.77,
+      "grad_norm": 3.2042014598846436,
+      "learning_rate": 1.816819859484692e-05,
+      "loss": 1.0784,
+      "step": 39325
+    },
+    {
+      "epoch": 11.77,
+      "grad_norm": 2.0353968143463135,
+      "learning_rate": 1.816254720226992e-05,
+      "loss": 0.9752,
+      "step": 39330
+    },
+    {
+      "epoch": 11.77,
+      "grad_norm": 1.9836816787719727,
+      "learning_rate": 1.8156896187318322e-05,
+      "loss": 1.1108,
+      "step": 39335
+    },
+    {
+      "epoch": 11.77,
+      "grad_norm": 1.6825382709503174,
+      "learning_rate": 1.815124555030424e-05,
+      "loss": 1.0884,
+      "step": 39340
+    },
+    {
+      "epoch": 11.77,
+      "grad_norm": 3.492959976196289,
+      "learning_rate": 1.8145595291539742e-05,
+      "loss": 1.1597,
+      "step": 39345
+    },
+    {
+      "epoch": 11.77,
+      "grad_norm": 3.515138864517212,
+      "learning_rate": 1.81399454113369e-05,
+      "loss": 1.0825,
+      "step": 39350
+    },
+    {
+      "epoch": 11.77,
+      "grad_norm": 2.3991005420684814,
+      "learning_rate": 1.8134295910007733e-05,
+      "loss": 1.1757,
+      "step": 39355
+    },
+    {
+      "epoch": 11.78,
+      "grad_norm": 4.532209396362305,
+      "learning_rate": 1.8128646787864274e-05,
+      "loss": 0.9928,
+      "step": 39360
+    },
+    {
+      "epoch": 11.78,
+      "grad_norm": 1.8209837675094604,
+      "learning_rate": 1.8122998045218515e-05,
+      "loss": 0.9525,
+      "step": 39365
+    },
+    {
+      "epoch": 11.78,
+      "grad_norm": 1.6788138151168823,
+      "learning_rate": 1.811734968238241e-05,
+      "loss": 1.1068,
+      "step": 39370
+    },
+    {
+      "epoch": 11.78,
+      "grad_norm": 1.9423623085021973,
+      "learning_rate": 1.8111701699667945e-05,
+      "loss": 0.9662,
+      "step": 39375
+    },
+    {
+      "epoch": 11.78,
+      "grad_norm": 1.7735339403152466,
+      "learning_rate": 1.8106054097387017e-05,
+      "loss": 1.1168,
+      "step": 39380
+    },
+    {
+      "epoch": 11.78,
+      "grad_norm": 4.587143898010254,
+      "learning_rate": 1.810040687585157e-05,
+      "loss": 1.0909,
+      "step": 39385
+    },
+    {
+      "epoch": 11.79,
+      "grad_norm": 2.695783853530884,
+      "learning_rate": 1.809476003537347e-05,
+      "loss": 1.0518,
+      "step": 39390
+    },
+    {
+      "epoch": 11.79,
+      "grad_norm": 1.4317811727523804,
+      "learning_rate": 1.80891135762646e-05,
+      "loss": 1.0841,
+      "step": 39395
+    },
+    {
+      "epoch": 11.79,
+      "grad_norm": 2.6233603954315186,
+      "learning_rate": 1.8083467498836805e-05,
+      "loss": 0.8933,
+      "step": 39400
+    },
+    {
+      "epoch": 11.79,
+      "grad_norm": 1.8944311141967773,
+      "learning_rate": 1.8077821803401903e-05,
+      "loss": 1.0322,
+      "step": 39405
+    },
+    {
+      "epoch": 11.79,
+      "grad_norm": 4.844879627227783,
+      "learning_rate": 1.807217649027172e-05,
+      "loss": 1.0349,
+      "step": 39410
+    },
+    {
+      "epoch": 11.79,
+      "grad_norm": 3.5124640464782715,
+      "learning_rate": 1.8066531559758015e-05,
+      "loss": 0.9921,
+      "step": 39415
+    },
+    {
+      "epoch": 11.79,
+      "grad_norm": 4.222341060638428,
+      "learning_rate": 1.8060887012172578e-05,
+      "loss": 1.0685,
+      "step": 39420
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 2.961749792098999,
+      "learning_rate": 1.8055242847827137e-05,
+      "loss": 1.0769,
+      "step": 39425
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 6.683197498321533,
+      "learning_rate": 1.8049599067033406e-05,
+      "loss": 1.1211,
+      "step": 39430
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 2.078709840774536,
+      "learning_rate": 1.80439556701031e-05,
+      "loss": 1.2193,
+      "step": 39435
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 3.6731183528900146,
+      "learning_rate": 1.803831265734789e-05,
+      "loss": 1.0561,
+      "step": 39440
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 2.6240220069885254,
+      "learning_rate": 1.803267002907944e-05,
+      "loss": 0.9361,
+      "step": 39445
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 5.170044898986816,
+      "learning_rate": 1.802702778560937e-05,
+      "loss": 1.033,
+      "step": 39450
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 2.432077407836914,
+      "learning_rate": 1.802138592724932e-05,
+      "loss": 1.1533,
+      "step": 39455
+    },
+    {
+      "epoch": 11.81,
+      "grad_norm": 1.541792869567871,
+      "learning_rate": 1.8015744454310873e-05,
+      "loss": 1.0976,
+      "step": 39460
+    },
+    {
+      "epoch": 11.81,
+      "grad_norm": 2.662036180496216,
+      "learning_rate": 1.8010103367105588e-05,
+      "loss": 0.9965,
+      "step": 39465
+    },
+    {
+      "epoch": 11.81,
+      "grad_norm": 3.3465163707733154,
+      "learning_rate": 1.800446266594504e-05,
+      "loss": 1.0623,
+      "step": 39470
+    },
+    {
+      "epoch": 11.81,
+      "grad_norm": 2.513610601425171,
+      "learning_rate": 1.799882235114074e-05,
+      "loss": 0.9073,
+      "step": 39475
+    },
+    {
+      "epoch": 11.81,
+      "grad_norm": 5.069911956787109,
+      "learning_rate": 1.79931824230042e-05,
+      "loss": 1.2199,
+      "step": 39480
+    },
+    {
+      "epoch": 11.81,
+      "grad_norm": 2.737182378768921,
+      "learning_rate": 1.7987542881846918e-05,
+      "loss": 1.194,
+      "step": 39485
+    },
+    {
+      "epoch": 11.81,
+      "grad_norm": 4.385614395141602,
+      "learning_rate": 1.7981903727980355e-05,
+      "loss": 1.0531,
+      "step": 39490
+    },
+    {
+      "epoch": 11.82,
+      "grad_norm": 1.9035724401474,
+      "learning_rate": 1.7976264961715954e-05,
+      "loss": 1.0845,
+      "step": 39495
+    },
+    {
+      "epoch": 11.82,
+      "grad_norm": 2.0744688510894775,
+      "learning_rate": 1.7970626583365135e-05,
+      "loss": 0.9908,
+      "step": 39500
+    },
+    {
+      "epoch": 11.82,
+      "grad_norm": 1.704471468925476,
+      "learning_rate": 1.796498859323931e-05,
+      "loss": 1.0739,
+      "step": 39505
+    },
+    {
+      "epoch": 11.82,
+      "grad_norm": 4.32279634475708,
+      "learning_rate": 1.7959350991649843e-05,
+      "loss": 0.9117,
+      "step": 39510
+    },
+    {
+      "epoch": 11.82,
+      "grad_norm": 3.1391208171844482,
+      "learning_rate": 1.795371377890811e-05,
+      "loss": 0.9438,
+      "step": 39515
+    },
+    {
+      "epoch": 11.82,
+      "grad_norm": 6.020325183868408,
+      "learning_rate": 1.7948076955325445e-05,
+      "loss": 0.9606,
+      "step": 39520
+    },
+    {
+      "epoch": 11.83,
+      "grad_norm": 2.498509645462036,
+      "learning_rate": 1.7942440521213144e-05,
+      "loss": 1.0956,
+      "step": 39525
+    },
+    {
+      "epoch": 11.83,
+      "grad_norm": 2.654123544692993,
+      "learning_rate": 1.793680447688253e-05,
+      "loss": 0.9559,
+      "step": 39530
+    },
+    {
+      "epoch": 11.83,
+      "grad_norm": 3.5224146842956543,
+      "learning_rate": 1.7931168822644857e-05,
+      "loss": 1.0071,
+      "step": 39535
+    },
+    {
+      "epoch": 11.83,
+      "grad_norm": 2.877861499786377,
+      "learning_rate": 1.7925533558811386e-05,
+      "loss": 1.0302,
+      "step": 39540
+    },
+    {
+      "epoch": 11.83,
+      "grad_norm": 2.9893412590026855,
+      "learning_rate": 1.7919898685693336e-05,
+      "loss": 0.9003,
+      "step": 39545
+    },
+    {
+      "epoch": 11.83,
+      "grad_norm": 3.915611505508423,
+      "learning_rate": 1.7914264203601933e-05,
+      "loss": 1.0966,
+      "step": 39550
+    },
+    {
+      "epoch": 11.83,
+      "grad_norm": 4.270025730133057,
+      "learning_rate": 1.7908630112848345e-05,
+      "loss": 1.0488,
+      "step": 39555
+    },
+    {
+      "epoch": 11.84,
+      "grad_norm": 1.905087947845459,
+      "learning_rate": 1.7902996413743743e-05,
+      "loss": 1.0429,
+      "step": 39560
+    },
+    {
+      "epoch": 11.84,
+      "grad_norm": 1.7518643140792847,
+      "learning_rate": 1.789736310659928e-05,
+      "loss": 0.9564,
+      "step": 39565
+    },
+    {
+      "epoch": 11.84,
+      "grad_norm": 2.0850372314453125,
+      "learning_rate": 1.7891730191726062e-05,
+      "loss": 1.0715,
+      "step": 39570
+    },
+    {
+      "epoch": 11.84,
+      "grad_norm": 4.21657133102417,
+      "learning_rate": 1.7886097669435203e-05,
+      "loss": 1.169,
+      "step": 39575
+    },
+    {
+      "epoch": 11.84,
+      "grad_norm": 1.5663069486618042,
+      "learning_rate": 1.7880465540037772e-05,
+      "loss": 0.9148,
+      "step": 39580
+    },
+    {
+      "epoch": 11.84,
+      "grad_norm": 1.8661555051803589,
+      "learning_rate": 1.787483380384483e-05,
+      "loss": 0.9692,
+      "step": 39585
+    },
+    {
+      "epoch": 11.84,
+      "grad_norm": 2.8097341060638428,
+      "learning_rate": 1.7869202461167414e-05,
+      "loss": 1.1088,
+      "step": 39590
+    },
+    {
+      "epoch": 11.85,
+      "grad_norm": 10.885028839111328,
+      "learning_rate": 1.7863571512316525e-05,
+      "loss": 1.0241,
+      "step": 39595
+    },
+    {
+      "epoch": 11.85,
+      "grad_norm": 3.2606093883514404,
+      "learning_rate": 1.7857940957603175e-05,
+      "loss": 1.1048,
+      "step": 39600
+    },
+    {
+      "epoch": 11.85,
+      "grad_norm": 2.2225308418273926,
+      "learning_rate": 1.785231079733831e-05,
+      "loss": 1.0134,
+      "step": 39605
+    },
+    {
+      "epoch": 11.85,
+      "grad_norm": 3.4681246280670166,
+      "learning_rate": 1.7846681031832902e-05,
+      "loss": 1.014,
+      "step": 39610
+    },
+    {
+      "epoch": 11.85,
+      "grad_norm": 2.1326732635498047,
+      "learning_rate": 1.784105166139785e-05,
+      "loss": 1.0633,
+      "step": 39615
+    },
+    {
+      "epoch": 11.85,
+      "grad_norm": 3.486208200454712,
+      "learning_rate": 1.7835422686344088e-05,
+      "loss": 1.0031,
+      "step": 39620
+    },
+    {
+      "epoch": 11.86,
+      "grad_norm": 2.6152260303497314,
+      "learning_rate": 1.7829794106982487e-05,
+      "loss": 1.1336,
+      "step": 39625
+    },
+    {
+      "epoch": 11.86,
+      "grad_norm": 3.0266525745391846,
+      "learning_rate": 1.782416592362389e-05,
+      "loss": 1.2247,
+      "step": 39630
+    },
+    {
+      "epoch": 11.86,
+      "grad_norm": 1.3934903144836426,
+      "learning_rate": 1.781853813657916e-05,
+      "loss": 1.1949,
+      "step": 39635
+    },
+    {
+      "epoch": 11.86,
+      "grad_norm": 1.248276948928833,
+      "learning_rate": 1.7812910746159096e-05,
+      "loss": 1.0491,
+      "step": 39640
+    },
+    {
+      "epoch": 11.86,
+      "grad_norm": 3.2080233097076416,
+      "learning_rate": 1.780728375267451e-05,
+      "loss": 1.0669,
+      "step": 39645
+    },
+    {
+      "epoch": 11.86,
+      "grad_norm": 6.742861747741699,
+      "learning_rate": 1.7801657156436162e-05,
+      "loss": 1.0856,
+      "step": 39650
+    },
+    {
+      "epoch": 11.86,
+      "grad_norm": 2.66377592086792,
+      "learning_rate": 1.7796030957754806e-05,
+      "loss": 1.2409,
+      "step": 39655
+    },
+    {
+      "epoch": 11.87,
+      "grad_norm": 2.5184435844421387,
+      "learning_rate": 1.779040515694117e-05,
+      "loss": 1.0872,
+      "step": 39660
+    },
+    {
+      "epoch": 11.87,
+      "grad_norm": 2.7967371940612793,
+      "learning_rate": 1.7784779754305962e-05,
+      "loss": 1.0393,
+      "step": 39665
+    },
+    {
+      "epoch": 11.87,
+      "grad_norm": 2.9046332836151123,
+      "learning_rate": 1.7779154750159874e-05,
+      "loss": 1.0531,
+      "step": 39670
+    },
+    {
+      "epoch": 11.87,
+      "grad_norm": 1.2060468196868896,
+      "learning_rate": 1.777353014481356e-05,
+      "loss": 1.0541,
+      "step": 39675
+    },
+    {
+      "epoch": 11.87,
+      "grad_norm": 5.9001359939575195,
+      "learning_rate": 1.7767905938577666e-05,
+      "loss": 1.0902,
+      "step": 39680
+    },
+    {
+      "epoch": 11.87,
+      "grad_norm": 3.407264471054077,
+      "learning_rate": 1.7762282131762814e-05,
+      "loss": 1.0468,
+      "step": 39685
+    },
+    {
+      "epoch": 11.87,
+      "grad_norm": 2.291767120361328,
+      "learning_rate": 1.7756658724679588e-05,
+      "loss": 0.9659,
+      "step": 39690
+    },
+    {
+      "epoch": 11.88,
+      "grad_norm": 1.3889952898025513,
+      "learning_rate": 1.775103571763858e-05,
+      "loss": 1.0298,
+      "step": 39695
+    },
+    {
+      "epoch": 11.88,
+      "grad_norm": 2.6598963737487793,
+      "learning_rate": 1.7745413110950325e-05,
+      "loss": 1.1888,
+      "step": 39700
+    },
+    {
+      "epoch": 11.88,
+      "grad_norm": 5.673483371734619,
+      "learning_rate": 1.7739790904925374e-05,
+      "loss": 1.0024,
+      "step": 39705
+    },
+    {
+      "epoch": 11.88,
+      "grad_norm": 4.94851016998291,
+      "learning_rate": 1.7734169099874216e-05,
+      "loss": 1.051,
+      "step": 39710
+    },
+    {
+      "epoch": 11.88,
+      "grad_norm": 10.22668743133545,
+      "learning_rate": 1.772854769610735e-05,
+      "loss": 1.0193,
+      "step": 39715
+    },
+    {
+      "epoch": 11.88,
+      "grad_norm": 3.5006158351898193,
+      "learning_rate": 1.7722926693935242e-05,
+      "loss": 1.0393,
+      "step": 39720
+    },
+    {
+      "epoch": 11.89,
+      "grad_norm": 2.18442964553833,
+      "learning_rate": 1.771730609366832e-05,
+      "loss": 0.9004,
+      "step": 39725
+    },
+    {
+      "epoch": 11.89,
+      "grad_norm": 2.1316640377044678,
+      "learning_rate": 1.771168589561702e-05,
+      "loss": 1.0415,
+      "step": 39730
+    },
+    {
+      "epoch": 11.89,
+      "grad_norm": 7.090075492858887,
+      "learning_rate": 1.770606610009172e-05,
+      "loss": 1.0966,
+      "step": 39735
+    },
+    {
+      "epoch": 11.89,
+      "grad_norm": 4.611660480499268,
+      "learning_rate": 1.770044670740282e-05,
+      "loss": 0.9932,
+      "step": 39740
+    },
+    {
+      "epoch": 11.89,
+      "grad_norm": 2.1602795124053955,
+      "learning_rate": 1.769482771786065e-05,
+      "loss": 1.2849,
+      "step": 39745
+    },
+    {
+      "epoch": 11.89,
+      "grad_norm": 5.2523274421691895,
+      "learning_rate": 1.7689209131775563e-05,
+      "loss": 1.0831,
+      "step": 39750
+    },
+    {
+      "epoch": 11.89,
+      "grad_norm": 1.25748610496521,
+      "learning_rate": 1.7683590949457852e-05,
+      "loss": 1.0985,
+      "step": 39755
+    },
+    {
+      "epoch": 11.9,
+      "grad_norm": 1.3775397539138794,
+      "learning_rate": 1.7677973171217805e-05,
+      "loss": 1.0786,
+      "step": 39760
+    },
+    {
+      "epoch": 11.9,
+      "grad_norm": 1.7527154684066772,
+      "learning_rate": 1.767235579736569e-05,
+      "loss": 1.0301,
+      "step": 39765
+    },
+    {
+      "epoch": 11.9,
+      "grad_norm": 3.6977381706237793,
+      "learning_rate": 1.7666738828211742e-05,
+      "loss": 1.2242,
+      "step": 39770
+    },
+    {
+      "epoch": 11.9,
+      "grad_norm": 2.5999135971069336,
+      "learning_rate": 1.76611222640662e-05,
+      "loss": 1.0193,
+      "step": 39775
+    },
+    {
+      "epoch": 11.9,
+      "grad_norm": 5.089832782745361,
+      "learning_rate": 1.7655506105239243e-05,
+      "loss": 1.0719,
+      "step": 39780
+    },
+    {
+      "epoch": 11.9,
+      "grad_norm": 3.162248373031616,
+      "learning_rate": 1.764989035204104e-05,
+      "loss": 0.87,
+      "step": 39785
+    },
+    {
+      "epoch": 11.9,
+      "grad_norm": 3.2377114295959473,
+      "learning_rate": 1.7644275004781763e-05,
+      "loss": 1.0549,
+      "step": 39790
+    },
+    {
+      "epoch": 11.91,
+      "grad_norm": 1.3557695150375366,
+      "learning_rate": 1.7638660063771523e-05,
+      "loss": 1.0341,
+      "step": 39795
+    },
+    {
+      "epoch": 11.91,
+      "grad_norm": 2.794524669647217,
+      "learning_rate": 1.7633045529320445e-05,
+      "loss": 1.0422,
+      "step": 39800
+    },
+    {
+      "epoch": 11.91,
+      "grad_norm": 5.456986427307129,
+      "learning_rate": 1.7627431401738597e-05,
+      "loss": 0.9706,
+      "step": 39805
+    },
+    {
+      "epoch": 11.91,
+      "grad_norm": 2.9997246265411377,
+      "learning_rate": 1.7621817681336055e-05,
+      "loss": 1.0415,
+      "step": 39810
+    },
+    {
+      "epoch": 11.91,
+      "grad_norm": 3.4345109462738037,
+      "learning_rate": 1.761620436842286e-05,
+      "loss": 0.884,
+      "step": 39815
+    },
+    {
+      "epoch": 11.91,
+      "grad_norm": 3.6423003673553467,
+      "learning_rate": 1.7610591463309007e-05,
+      "loss": 1.1453,
+      "step": 39820
+    },
+    {
+      "epoch": 11.92,
+      "grad_norm": 3.724964141845703,
+      "learning_rate": 1.760497896630452e-05,
+      "loss": 1.1898,
+      "step": 39825
+    },
+    {
+      "epoch": 11.92,
+      "grad_norm": 3.579566478729248,
+      "learning_rate": 1.759936687771934e-05,
+      "loss": 1.0795,
+      "step": 39830
+    },
+    {
+      "epoch": 11.92,
+      "grad_norm": 6.6637492179870605,
+      "learning_rate": 1.7593755197863455e-05,
+      "loss": 1.0689,
+      "step": 39835
+    },
+    {
+      "epoch": 11.92,
+      "grad_norm": 3.4842472076416016,
+      "learning_rate": 1.758814392704676e-05,
+      "loss": 1.0492,
+      "step": 39840
+    },
+    {
+      "epoch": 11.92,
+      "grad_norm": 4.1526079177856445,
+      "learning_rate": 1.758253306557917e-05,
+      "loss": 0.9039,
+      "step": 39845
+    },
+    {
+      "epoch": 11.92,
+      "grad_norm": 2.869919538497925,
+      "learning_rate": 1.757692261377058e-05,
+      "loss": 1.0529,
+      "step": 39850
+    },
+    {
+      "epoch": 11.92,
+      "grad_norm": 1.4743778705596924,
+      "learning_rate": 1.7571312571930827e-05,
+      "loss": 0.9677,
+      "step": 39855
+    },
+    {
+      "epoch": 11.93,
+      "grad_norm": 5.439683437347412,
+      "learning_rate": 1.756570294036977e-05,
+      "loss": 1.0239,
+      "step": 39860
+    },
+    {
+      "epoch": 11.93,
+      "grad_norm": 1.5706872940063477,
+      "learning_rate": 1.7560093719397204e-05,
+      "loss": 1.0556,
+      "step": 39865
+    },
+    {
+      "epoch": 11.93,
+      "grad_norm": 4.678708553314209,
+      "learning_rate": 1.7554484909322933e-05,
+      "loss": 1.1842,
+      "step": 39870
+    },
+    {
+      "epoch": 11.93,
+      "grad_norm": 3.4976863861083984,
+      "learning_rate": 1.7548876510456723e-05,
+      "loss": 0.873,
+      "step": 39875
+    },
+    {
+      "epoch": 11.93,
+      "grad_norm": 3.4963254928588867,
+      "learning_rate": 1.7543268523108308e-05,
+      "loss": 0.9826,
+      "step": 39880
+    },
+    {
+      "epoch": 11.93,
+      "grad_norm": 4.180238246917725,
+      "learning_rate": 1.7537660947587434e-05,
+      "loss": 0.9779,
+      "step": 39885
+    },
+    {
+      "epoch": 11.93,
+      "grad_norm": 0.9406687617301941,
+      "learning_rate": 1.753205378420378e-05,
+      "loss": 1.0133,
+      "step": 39890
+    },
+    {
+      "epoch": 11.94,
+      "grad_norm": 2.034942865371704,
+      "learning_rate": 1.7526447033267037e-05,
+      "loss": 1.203,
+      "step": 39895
+    },
+    {
+      "epoch": 11.94,
+      "grad_norm": 2.6159539222717285,
+      "learning_rate": 1.7520840695086847e-05,
+      "loss": 1.1254,
+      "step": 39900
+    },
+    {
+      "epoch": 11.94,
+      "grad_norm": 4.698697566986084,
+      "learning_rate": 1.7515234769972865e-05,
+      "loss": 0.823,
+      "step": 39905
+    },
+    {
+      "epoch": 11.94,
+      "grad_norm": 3.892012119293213,
+      "learning_rate": 1.750962925823469e-05,
+      "loss": 0.9369,
+      "step": 39910
+    },
+    {
+      "epoch": 11.94,
+      "grad_norm": 1.37578284740448,
+      "learning_rate": 1.750402416018189e-05,
+      "loss": 1.1761,
+      "step": 39915
+    },
+    {
+      "epoch": 11.94,
+      "grad_norm": 3.057701349258423,
+      "learning_rate": 1.749841947612405e-05,
+      "loss": 0.9876,
+      "step": 39920
+    },
+    {
+      "epoch": 11.95,
+      "grad_norm": 4.663394451141357,
+      "learning_rate": 1.7492815206370705e-05,
+      "loss": 0.8152,
+      "step": 39925
+    },
+    {
+      "epoch": 11.95,
+      "grad_norm": 2.727992296218872,
+      "learning_rate": 1.748721135123137e-05,
+      "loss": 1.1597,
+      "step": 39930
+    },
+    {
+      "epoch": 11.95,
+      "grad_norm": 1.3559057712554932,
+      "learning_rate": 1.7481607911015547e-05,
+      "loss": 1.0632,
+      "step": 39935
+    },
+    {
+      "epoch": 11.95,
+      "grad_norm": 2.3270246982574463,
+      "learning_rate": 1.7476004886032703e-05,
+      "loss": 0.8817,
+      "step": 39940
+    },
+    {
+      "epoch": 11.95,
+      "grad_norm": 3.695523977279663,
+      "learning_rate": 1.7470402276592294e-05,
+      "loss": 0.9562,
+      "step": 39945
+    },
+    {
+      "epoch": 11.95,
+      "grad_norm": 2.7991878986358643,
+      "learning_rate": 1.746480008300373e-05,
+      "loss": 1.0675,
+      "step": 39950
+    },
+    {
+      "epoch": 11.95,
+      "grad_norm": 10.5833740234375,
+      "learning_rate": 1.7459198305576434e-05,
+      "loss": 1.0423,
+      "step": 39955
+    },
+    {
+      "epoch": 11.96,
+      "grad_norm": 1.6265363693237305,
+      "learning_rate": 1.745359694461977e-05,
+      "loss": 0.9417,
+      "step": 39960
+    },
+    {
+      "epoch": 11.96,
+      "grad_norm": 2.9523098468780518,
+      "learning_rate": 1.744799600044311e-05,
+      "loss": 1.2033,
+      "step": 39965
+    },
+    {
+      "epoch": 11.96,
+      "grad_norm": 3.3546805381774902,
+      "learning_rate": 1.744239547335577e-05,
+      "loss": 1.0307,
+      "step": 39970
+    },
+    {
+      "epoch": 11.96,
+      "grad_norm": 4.862351894378662,
+      "learning_rate": 1.7436795363667086e-05,
+      "loss": 0.9044,
+      "step": 39975
+    },
+    {
+      "epoch": 11.96,
+      "grad_norm": 3.834397792816162,
+      "learning_rate": 1.743119567168633e-05,
+      "loss": 0.9981,
+      "step": 39980
+    },
+    {
+      "epoch": 11.96,
+      "grad_norm": 5.2290472984313965,
+      "learning_rate": 1.742559639772276e-05,
+      "loss": 1.0151,
+      "step": 39985
+    },
+    {
+      "epoch": 11.96,
+      "grad_norm": 1.2269209623336792,
+      "learning_rate": 1.741999754208564e-05,
+      "loss": 0.915,
+      "step": 39990
+    },
+    {
+      "epoch": 11.97,
+      "grad_norm": 9.868966102600098,
+      "learning_rate": 1.7414399105084166e-05,
+      "loss": 0.9631,
+      "step": 39995
+    },
+    {
+      "epoch": 11.97,
+      "grad_norm": 4.046831130981445,
+      "learning_rate": 1.7408801087027554e-05,
+      "loss": 1.0331,
+      "step": 40000
+    },
+    {
+      "epoch": 11.97,
+      "grad_norm": 7.04620361328125,
+      "learning_rate": 1.740320348822496e-05,
+      "loss": 1.0635,
+      "step": 40005
+    },
+    {
+      "epoch": 11.97,
+      "grad_norm": 4.793135166168213,
+      "learning_rate": 1.739760630898554e-05,
+      "loss": 1.0377,
+      "step": 40010
+    },
+    {
+      "epoch": 11.97,
+      "grad_norm": 1.5962762832641602,
+      "learning_rate": 1.7392009549618426e-05,
+      "loss": 1.3054,
+      "step": 40015
+    },
+    {
+      "epoch": 11.97,
+      "grad_norm": 3.573862314224243,
+      "learning_rate": 1.7386413210432717e-05,
+      "loss": 1.1205,
+      "step": 40020
+    },
+    {
+      "epoch": 11.98,
+      "grad_norm": 2.5612030029296875,
+      "learning_rate": 1.7380817291737488e-05,
+      "loss": 0.9382,
+      "step": 40025
+    },
+    {
+      "epoch": 11.98,
+      "grad_norm": 4.043525218963623,
+      "learning_rate": 1.7375221793841806e-05,
+      "loss": 1.083,
+      "step": 40030
+    },
+    {
+      "epoch": 11.98,
+      "grad_norm": 3.7928805351257324,
+      "learning_rate": 1.7369626717054693e-05,
+      "loss": 1.1117,
+      "step": 40035
+    },
+    {
+      "epoch": 11.98,
+      "grad_norm": 2.3192451000213623,
+      "learning_rate": 1.7364032061685174e-05,
+      "loss": 1.0247,
+      "step": 40040
+    },
+    {
+      "epoch": 11.98,
+      "grad_norm": 3.9967002868652344,
+      "learning_rate": 1.7358437828042215e-05,
+      "loss": 0.9724,
+      "step": 40045
+    },
+    {
+      "epoch": 11.98,
+      "grad_norm": 4.295874118804932,
+      "learning_rate": 1.7352844016434802e-05,
+      "loss": 1.17,
+      "step": 40050
+    },
+    {
+      "epoch": 11.98,
+      "grad_norm": 2.709665536880493,
+      "learning_rate": 1.7347250627171857e-05,
+      "loss": 1.0567,
+      "step": 40055
+    },
+    {
+      "epoch": 11.99,
+      "grad_norm": 1.6827901601791382,
+      "learning_rate": 1.7341657660562313e-05,
+      "loss": 1.0409,
+      "step": 40060
+    },
+    {
+      "epoch": 11.99,
+      "grad_norm": 1.5551859140396118,
+      "learning_rate": 1.7336065116915045e-05,
+      "loss": 0.9164,
+      "step": 40065
+    },
+    {
+      "epoch": 11.99,
+      "grad_norm": 4.49882698059082,
+      "learning_rate": 1.7330472996538948e-05,
+      "loss": 0.9778,
+      "step": 40070
+    },
+    {
+      "epoch": 11.99,
+      "grad_norm": 8.409674644470215,
+      "learning_rate": 1.7324881299742855e-05,
+      "loss": 1.1369,
+      "step": 40075
+    },
+    {
+      "epoch": 11.99,
+      "grad_norm": 3.090298891067505,
+      "learning_rate": 1.7319290026835577e-05,
+      "loss": 1.0897,
+      "step": 40080
+    },
+    {
+      "epoch": 11.99,
+      "grad_norm": 4.631008625030518,
+      "learning_rate": 1.731369917812594e-05,
+      "loss": 1.0815,
+      "step": 40085
+    },
+    {
+      "epoch": 11.99,
+      "grad_norm": 2.0959417819976807,
+      "learning_rate": 1.7308108753922698e-05,
+      "loss": 0.944,
+      "step": 40090
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 2.209991931915283,
+      "learning_rate": 1.730251875453462e-05,
+      "loss": 1.1075,
+      "step": 40095
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 1.7672827243804932,
+      "learning_rate": 1.7296929180270424e-05,
+      "loss": 1.0148,
+      "step": 40100
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 2.572972536087036,
+      "learning_rate": 1.7291340031438828e-05,
+      "loss": 0.9899,
+      "step": 40105
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 3.605733871459961,
+      "learning_rate": 1.728575130834851e-05,
+      "loss": 1.1367,
+      "step": 40110
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.9864242672920227,
+      "learning_rate": 1.7280163011308127e-05,
+      "loss": 1.0373,
+      "step": 40115
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 2.625638484954834,
+      "learning_rate": 1.7274575140626318e-05,
+      "loss": 0.9886,
+      "step": 40120
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 1.6979814767837524,
+      "learning_rate": 1.726898769661169e-05,
+      "loss": 0.9172,
+      "step": 40125
+    },
+    {
+      "epoch": 12.01,
+      "grad_norm": 3.1777591705322266,
+      "learning_rate": 1.7263400679572838e-05,
+      "loss": 1.1741,
+      "step": 40130
+    },
+    {
+      "epoch": 12.01,
+      "grad_norm": 1.9709137678146362,
+      "learning_rate": 1.725781408981833e-05,
+      "loss": 0.9846,
+      "step": 40135
+    },
+    {
+      "epoch": 12.01,
+      "grad_norm": 2.2092223167419434,
+      "learning_rate": 1.7252227927656692e-05,
+      "loss": 1.1,
+      "step": 40140
+    },
+    {
+      "epoch": 12.01,
+      "grad_norm": 3.6297123432159424,
+      "learning_rate": 1.7246642193396463e-05,
+      "loss": 1.1311,
+      "step": 40145
+    },
+    {
+      "epoch": 12.01,
+      "grad_norm": 2.822312831878662,
+      "learning_rate": 1.7241056887346115e-05,
+      "loss": 1.0686,
+      "step": 40150
+    },
+    {
+      "epoch": 12.01,
+      "grad_norm": 2.791294574737549,
+      "learning_rate": 1.7235472009814142e-05,
+      "loss": 1.0233,
+      "step": 40155
+    },
+    {
+      "epoch": 12.02,
+      "grad_norm": 2.6758475303649902,
+      "learning_rate": 1.7229887561108967e-05,
+      "loss": 0.8289,
+      "step": 40160
+    },
+    {
+      "epoch": 12.02,
+      "grad_norm": 1.5019391775131226,
+      "learning_rate": 1.7224303541539034e-05,
+      "loss": 1.0052,
+      "step": 40165
+    },
+    {
+      "epoch": 12.02,
+      "grad_norm": 2.995958089828491,
+      "learning_rate": 1.7218719951412736e-05,
+      "loss": 1.0393,
+      "step": 40170
+    },
+    {
+      "epoch": 12.02,
+      "grad_norm": 2.246119976043701,
+      "learning_rate": 1.7213136791038436e-05,
+      "loss": 1.0853,
+      "step": 40175
+    },
+    {
+      "epoch": 12.02,
+      "grad_norm": 5.3217315673828125,
+      "learning_rate": 1.7207554060724505e-05,
+      "loss": 1.0277,
+      "step": 40180
+    },
+    {
+      "epoch": 12.02,
+      "grad_norm": 1.9030908346176147,
+      "learning_rate": 1.720197176077926e-05,
+      "loss": 1.0401,
+      "step": 40185
+    },
+    {
+      "epoch": 12.02,
+      "grad_norm": 3.2368953227996826,
+      "learning_rate": 1.7196389891511017e-05,
+      "loss": 0.9992,
+      "step": 40190
+    },
+    {
+      "epoch": 12.03,
+      "grad_norm": 2.3075368404388428,
+      "learning_rate": 1.7190808453228035e-05,
+      "loss": 1.1156,
+      "step": 40195
+    },
+    {
+      "epoch": 12.03,
+      "grad_norm": 3.3374037742614746,
+      "learning_rate": 1.7185227446238597e-05,
+      "loss": 1.2066,
+      "step": 40200
+    },
+    {
+      "epoch": 12.03,
+      "grad_norm": 2.5891878604888916,
+      "learning_rate": 1.7179646870850917e-05,
+      "loss": 0.8088,
+      "step": 40205
+    },
+    {
+      "epoch": 12.03,
+      "grad_norm": 2.94810152053833,
+      "learning_rate": 1.7174066727373212e-05,
+      "loss": 0.9585,
+      "step": 40210
+    },
+    {
+      "epoch": 12.03,
+      "grad_norm": 1.5498286485671997,
+      "learning_rate": 1.7168487016113675e-05,
+      "loss": 1.1713,
+      "step": 40215
+    },
+    {
+      "epoch": 12.03,
+      "grad_norm": 2.1500444412231445,
+      "learning_rate": 1.7162907737380447e-05,
+      "loss": 1.0594,
+      "step": 40220
+    },
+    {
+      "epoch": 12.03,
+      "grad_norm": 2.072866439819336,
+      "learning_rate": 1.7157328891481688e-05,
+      "loss": 1.0718,
+      "step": 40225
+    },
+    {
+      "epoch": 12.04,
+      "grad_norm": 2.1537182331085205,
+      "learning_rate": 1.7151750478725506e-05,
+      "loss": 0.953,
+      "step": 40230
+    },
+    {
+      "epoch": 12.04,
+      "grad_norm": 3.077263116836548,
+      "learning_rate": 1.7146172499419976e-05,
+      "loss": 1.0168,
+      "step": 40235
+    },
+    {
+      "epoch": 12.04,
+      "grad_norm": 3.462986469268799,
+      "learning_rate": 1.7140594953873184e-05,
+      "loss": 0.8017,
+      "step": 40240
+    },
+    {
+      "epoch": 12.04,
+      "grad_norm": 2.716017723083496,
+      "learning_rate": 1.7135017842393156e-05,
+      "loss": 1.1061,
+      "step": 40245
+    },
+    {
+      "epoch": 12.04,
+      "grad_norm": 2.1912074089050293,
+      "learning_rate": 1.7129441165287923e-05,
+      "loss": 0.9462,
+      "step": 40250
+    },
+    {
+      "epoch": 12.04,
+      "grad_norm": 1.6843229532241821,
+      "learning_rate": 1.7123864922865468e-05,
+      "loss": 0.9333,
+      "step": 40255
+    },
+    {
+      "epoch": 12.05,
+      "grad_norm": 1.5522648096084595,
+      "learning_rate": 1.7118289115433774e-05,
+      "loss": 0.9716,
+      "step": 40260
+    },
+    {
+      "epoch": 12.05,
+      "grad_norm": 3.79538631439209,
+      "learning_rate": 1.7112713743300778e-05,
+      "loss": 0.9679,
+      "step": 40265
+    },
+    {
+      "epoch": 12.05,
+      "grad_norm": 2.5258116722106934,
+      "learning_rate": 1.7107138806774398e-05,
+      "loss": 1.0447,
+      "step": 40270
+    },
+    {
+      "epoch": 12.05,
+      "grad_norm": 1.7278646230697632,
+      "learning_rate": 1.7101564306162546e-05,
+      "loss": 1.0542,
+      "step": 40275
+    },
+    {
+      "epoch": 12.05,
+      "grad_norm": 3.5281360149383545,
+      "learning_rate": 1.7095990241773076e-05,
+      "loss": 0.9029,
+      "step": 40280
+    },
+    {
+      "epoch": 12.05,
+      "grad_norm": 2.533038854598999,
+      "learning_rate": 1.7090416613913863e-05,
+      "loss": 1.1102,
+      "step": 40285
+    },
+    {
+      "epoch": 12.05,
+      "grad_norm": 2.665400505065918,
+      "learning_rate": 1.7084843422892705e-05,
+      "loss": 0.9122,
+      "step": 40290
+    },
+    {
+      "epoch": 12.06,
+      "grad_norm": 3.1313700675964355,
+      "learning_rate": 1.7079270669017422e-05,
+      "loss": 1.0415,
+      "step": 40295
+    },
+    {
+      "epoch": 12.06,
+      "grad_norm": 1.9384442567825317,
+      "learning_rate": 1.707369835259579e-05,
+      "loss": 0.8205,
+      "step": 40300
+    },
+    {
+      "epoch": 12.06,
+      "grad_norm": 3.2962498664855957,
+      "learning_rate": 1.7068126473935552e-05,
+      "loss": 1.1216,
+      "step": 40305
+    },
+    {
+      "epoch": 12.06,
+      "grad_norm": 4.275677680969238,
+      "learning_rate": 1.7062555033344457e-05,
+      "loss": 0.9877,
+      "step": 40310
+    },
+    {
+      "epoch": 12.06,
+      "grad_norm": 2.7179315090179443,
+      "learning_rate": 1.705698403113018e-05,
+      "loss": 0.8647,
+      "step": 40315
+    },
+    {
+      "epoch": 12.06,
+      "grad_norm": 2.34228515625,
+      "learning_rate": 1.7051413467600435e-05,
+      "loss": 0.9374,
+      "step": 40320
+    },
+    {
+      "epoch": 12.06,
+      "grad_norm": 3.802971124649048,
+      "learning_rate": 1.704584334306285e-05,
+      "loss": 0.9489,
+      "step": 40325
+    },
+    {
+      "epoch": 12.07,
+      "grad_norm": 3.335583448410034,
+      "learning_rate": 1.704027365782508e-05,
+      "loss": 1.0729,
+      "step": 40330
+    },
+    {
+      "epoch": 12.07,
+      "grad_norm": 2.546466827392578,
+      "learning_rate": 1.7034704412194722e-05,
+      "loss": 0.991,
+      "step": 40335
+    },
+    {
+      "epoch": 12.07,
+      "grad_norm": 3.204293727874756,
+      "learning_rate": 1.7029135606479346e-05,
+      "loss": 1.137,
+      "step": 40340
+    },
+    {
+      "epoch": 12.07,
+      "grad_norm": 4.195864677429199,
+      "learning_rate": 1.702356724098654e-05,
+      "loss": 0.9883,
+      "step": 40345
+    },
+    {
+      "epoch": 12.07,
+      "grad_norm": 2.6011598110198975,
+      "learning_rate": 1.7017999316023814e-05,
+      "loss": 1.0274,
+      "step": 40350
+    },
+    {
+      "epoch": 12.07,
+      "grad_norm": 1.9285248517990112,
+      "learning_rate": 1.7012431831898696e-05,
+      "loss": 0.9519,
+      "step": 40355
+    },
+    {
+      "epoch": 12.08,
+      "grad_norm": 1.2027796506881714,
+      "learning_rate": 1.700686478891867e-05,
+      "loss": 1.2077,
+      "step": 40360
+    },
+    {
+      "epoch": 12.08,
+      "grad_norm": 13.604466438293457,
+      "learning_rate": 1.700129818739118e-05,
+      "loss": 0.802,
+      "step": 40365
+    },
+    {
+      "epoch": 12.08,
+      "grad_norm": 2.9307985305786133,
+      "learning_rate": 1.6995732027623677e-05,
+      "loss": 1.0621,
+      "step": 40370
+    },
+    {
+      "epoch": 12.08,
+      "grad_norm": 3.506657838821411,
+      "learning_rate": 1.6990166309923584e-05,
+      "loss": 1.0037,
+      "step": 40375
+    },
+    {
+      "epoch": 12.08,
+      "grad_norm": 3.7252209186553955,
+      "learning_rate": 1.6984601034598273e-05,
+      "loss": 1.0522,
+      "step": 40380
+    },
+    {
+      "epoch": 12.08,
+      "grad_norm": 3.6333189010620117,
+      "learning_rate": 1.6979036201955118e-05,
+      "loss": 1.1053,
+      "step": 40385
+    },
+    {
+      "epoch": 12.08,
+      "grad_norm": 1.9753819704055786,
+      "learning_rate": 1.6973471812301456e-05,
+      "loss": 0.9674,
+      "step": 40390
+    },
+    {
+      "epoch": 12.09,
+      "grad_norm": 1.905187964439392,
+      "learning_rate": 1.6967907865944608e-05,
+      "loss": 1.0107,
+      "step": 40395
+    },
+    {
+      "epoch": 12.09,
+      "grad_norm": 4.743467807769775,
+      "learning_rate": 1.6962344363191846e-05,
+      "loss": 0.8817,
+      "step": 40400
+    },
+    {
+      "epoch": 12.09,
+      "grad_norm": 1.3228265047073364,
+      "learning_rate": 1.6956781304350466e-05,
+      "loss": 0.9523,
+      "step": 40405
+    },
+    {
+      "epoch": 12.09,
+      "grad_norm": 2.1767237186431885,
+      "learning_rate": 1.695121868972768e-05,
+      "loss": 1.177,
+      "step": 40410
+    },
+    {
+      "epoch": 12.09,
+      "grad_norm": 2.85136342048645,
+      "learning_rate": 1.694565651963073e-05,
+      "loss": 1.2009,
+      "step": 40415
+    },
+    {
+      "epoch": 12.09,
+      "grad_norm": 1.4177536964416504,
+      "learning_rate": 1.694009479436679e-05,
+      "loss": 1.0331,
+      "step": 40420
+    },
+    {
+      "epoch": 12.09,
+      "grad_norm": 3.8098344802856445,
+      "learning_rate": 1.6934533514243046e-05,
+      "loss": 1.0316,
+      "step": 40425
+    },
+    {
+      "epoch": 12.1,
+      "grad_norm": 2.5896317958831787,
+      "learning_rate": 1.6928972679566633e-05,
+      "loss": 1.1835,
+      "step": 40430
+    },
+    {
+      "epoch": 12.1,
+      "grad_norm": 1.0999228954315186,
+      "learning_rate": 1.692341229064466e-05,
+      "loss": 1.1264,
+      "step": 40435
+    },
+    {
+      "epoch": 12.1,
+      "grad_norm": 3.237612009048462,
+      "learning_rate": 1.691785234778424e-05,
+      "loss": 0.9797,
+      "step": 40440
+    },
+    {
+      "epoch": 12.1,
+      "grad_norm": 2.606142520904541,
+      "learning_rate": 1.691229285129242e-05,
+      "loss": 0.9886,
+      "step": 40445
+    },
+    {
+      "epoch": 12.1,
+      "grad_norm": 5.350770950317383,
+      "learning_rate": 1.6906733801476275e-05,
+      "loss": 0.9763,
+      "step": 40450
+    },
+    {
+      "epoch": 12.1,
+      "grad_norm": 1.0530993938446045,
+      "learning_rate": 1.69011751986428e-05,
+      "loss": 0.9137,
+      "step": 40455
+    },
+    {
+      "epoch": 12.11,
+      "grad_norm": 5.592156887054443,
+      "learning_rate": 1.6895617043099006e-05,
+      "loss": 1.0207,
+      "step": 40460
+    },
+    {
+      "epoch": 12.11,
+      "grad_norm": 2.4852864742279053,
+      "learning_rate": 1.6890059335151856e-05,
+      "loss": 1.008,
+      "step": 40465
+    },
+    {
+      "epoch": 12.11,
+      "grad_norm": 2.4040777683258057,
+      "learning_rate": 1.6884502075108298e-05,
+      "loss": 1.0699,
+      "step": 40470
+    },
+    {
+      "epoch": 12.11,
+      "grad_norm": 2.1579723358154297,
+      "learning_rate": 1.687894526327526e-05,
+      "loss": 0.9733,
+      "step": 40475
+    },
+    {
+      "epoch": 12.11,
+      "grad_norm": 3.1468398571014404,
+      "learning_rate": 1.6873388899959625e-05,
+      "loss": 1.0083,
+      "step": 40480
+    },
+    {
+      "epoch": 12.11,
+      "grad_norm": 2.4095633029937744,
+      "learning_rate": 1.686783298546828e-05,
+      "loss": 1.0388,
+      "step": 40485
+    },
+    {
+      "epoch": 12.11,
+      "grad_norm": 3.516322612762451,
+      "learning_rate": 1.686227752010807e-05,
+      "loss": 1.0025,
+      "step": 40490
+    },
+    {
+      "epoch": 12.12,
+      "grad_norm": 4.202288627624512,
+      "learning_rate": 1.6856722504185802e-05,
+      "loss": 1.0996,
+      "step": 40495
+    },
+    {
+      "epoch": 12.12,
+      "grad_norm": 3.198098659515381,
+      "learning_rate": 1.68511679380083e-05,
+      "loss": 0.8971,
+      "step": 40500
+    },
+    {
+      "epoch": 12.12,
+      "grad_norm": 6.3249993324279785,
+      "learning_rate": 1.684561382188231e-05,
+      "loss": 0.9876,
+      "step": 40505
+    },
+    {
+      "epoch": 12.12,
+      "grad_norm": 1.2453722953796387,
+      "learning_rate": 1.68400601561146e-05,
+      "loss": 0.9331,
+      "step": 40510
+    },
+    {
+      "epoch": 12.12,
+      "grad_norm": 3.7218070030212402,
+      "learning_rate": 1.683450694101188e-05,
+      "loss": 0.9366,
+      "step": 40515
+    },
+    {
+      "epoch": 12.12,
+      "grad_norm": 2.971968412399292,
+      "learning_rate": 1.682895417688086e-05,
+      "loss": 1.0184,
+      "step": 40520
+    },
+    {
+      "epoch": 12.12,
+      "grad_norm": 2.1106045246124268,
+      "learning_rate": 1.682340186402821e-05,
+      "loss": 0.9095,
+      "step": 40525
+    },
+    {
+      "epoch": 12.13,
+      "grad_norm": 1.3278517723083496,
+      "learning_rate": 1.6817850002760565e-05,
+      "loss": 1.1384,
+      "step": 40530
+    },
+    {
+      "epoch": 12.13,
+      "grad_norm": 1.8929804563522339,
+      "learning_rate": 1.6812298593384574e-05,
+      "loss": 1.029,
+      "step": 40535
+    },
+    {
+      "epoch": 12.13,
+      "grad_norm": 1.8360497951507568,
+      "learning_rate": 1.6806747636206804e-05,
+      "loss": 0.9871,
+      "step": 40540
+    },
+    {
+      "epoch": 12.13,
+      "grad_norm": 1.8714280128479004,
+      "learning_rate": 1.680119713153386e-05,
+      "loss": 1.0098,
+      "step": 40545
+    },
+    {
+      "epoch": 12.13,
+      "grad_norm": 2.2284882068634033,
+      "learning_rate": 1.6795647079672262e-05,
+      "loss": 0.9422,
+      "step": 40550
+    },
+    {
+      "epoch": 12.13,
+      "grad_norm": 2.5915369987487793,
+      "learning_rate": 1.6790097480928562e-05,
+      "loss": 0.9804,
+      "step": 40555
+    },
+    {
+      "epoch": 12.14,
+      "grad_norm": 1.8112045526504517,
+      "learning_rate": 1.678454833560924e-05,
+      "loss": 0.8877,
+      "step": 40560
+    },
+    {
+      "epoch": 12.14,
+      "grad_norm": 2.2570388317108154,
+      "learning_rate": 1.677899964402077e-05,
+      "loss": 0.8855,
+      "step": 40565
+    },
+    {
+      "epoch": 12.14,
+      "grad_norm": 2.4070281982421875,
+      "learning_rate": 1.6773451406469607e-05,
+      "loss": 1.2104,
+      "step": 40570
+    },
+    {
+      "epoch": 12.14,
+      "grad_norm": 1.3471062183380127,
+      "learning_rate": 1.6767903623262168e-05,
+      "loss": 1.1283,
+      "step": 40575
+    },
+    {
+      "epoch": 12.14,
+      "grad_norm": 1.2536911964416504,
+      "learning_rate": 1.6762356294704863e-05,
+      "loss": 1.0132,
+      "step": 40580
+    },
+    {
+      "epoch": 12.14,
+      "grad_norm": 4.913021564483643,
+      "learning_rate": 1.675680942110406e-05,
+      "loss": 1.1661,
+      "step": 40585
+    },
+    {
+      "epoch": 12.14,
+      "grad_norm": 3.420980215072632,
+      "learning_rate": 1.675126300276609e-05,
+      "loss": 0.7853,
+      "step": 40590
+    },
+    {
+      "epoch": 12.15,
+      "grad_norm": 4.242785930633545,
+      "learning_rate": 1.6745717039997303e-05,
+      "loss": 0.8756,
+      "step": 40595
+    },
+    {
+      "epoch": 12.15,
+      "grad_norm": 4.858880519866943,
+      "learning_rate": 1.6740171533103974e-05,
+      "loss": 0.894,
+      "step": 40600
+    },
+    {
+      "epoch": 12.15,
+      "grad_norm": 1.5555437803268433,
+      "learning_rate": 1.6734626482392397e-05,
+      "loss": 0.9653,
+      "step": 40605
+    },
+    {
+      "epoch": 12.15,
+      "grad_norm": 2.0157172679901123,
+      "learning_rate": 1.6729081888168794e-05,
+      "loss": 1.0733,
+      "step": 40610
+    },
+    {
+      "epoch": 12.15,
+      "grad_norm": 5.87971830368042,
+      "learning_rate": 1.6723537750739414e-05,
+      "loss": 1.0843,
+      "step": 40615
+    },
+    {
+      "epoch": 12.15,
+      "grad_norm": 3.524643659591675,
+      "learning_rate": 1.6717994070410442e-05,
+      "loss": 1.0036,
+      "step": 40620
+    },
+    {
+      "epoch": 12.15,
+      "grad_norm": 3.6206631660461426,
+      "learning_rate": 1.6712450847488037e-05,
+      "loss": 1.038,
+      "step": 40625
+    },
+    {
+      "epoch": 12.16,
+      "grad_norm": 3.202336072921753,
+      "learning_rate": 1.6706908082278368e-05,
+      "loss": 1.021,
+      "step": 40630
+    },
+    {
+      "epoch": 12.16,
+      "grad_norm": 1.4924439191818237,
+      "learning_rate": 1.6701365775087534e-05,
+      "loss": 1.123,
+      "step": 40635
+    },
+    {
+      "epoch": 12.16,
+      "grad_norm": 2.8223538398742676,
+      "learning_rate": 1.669582392622165e-05,
+      "loss": 1.1115,
+      "step": 40640
+    },
+    {
+      "epoch": 12.16,
+      "grad_norm": 4.758267879486084,
+      "learning_rate": 1.6690282535986775e-05,
+      "loss": 1.0646,
+      "step": 40645
+    },
+    {
+      "epoch": 12.16,
+      "grad_norm": 3.9418559074401855,
+      "learning_rate": 1.6684741604688962e-05,
+      "loss": 1.0262,
+      "step": 40650
+    },
+    {
+      "epoch": 12.16,
+      "grad_norm": 2.5239322185516357,
+      "learning_rate": 1.6679201132634227e-05,
+      "loss": 0.9649,
+      "step": 40655
+    },
+    {
+      "epoch": 12.17,
+      "grad_norm": 2.7712440490722656,
+      "learning_rate": 1.667366112012856e-05,
+      "loss": 0.9846,
+      "step": 40660
+    },
+    {
+      "epoch": 12.17,
+      "grad_norm": 1.613471269607544,
+      "learning_rate": 1.666812156747794e-05,
+      "loss": 0.9812,
+      "step": 40665
+    },
+    {
+      "epoch": 12.17,
+      "grad_norm": 3.2598836421966553,
+      "learning_rate": 1.6662582474988297e-05,
+      "loss": 1.1001,
+      "step": 40670
+    },
+    {
+      "epoch": 12.17,
+      "grad_norm": 4.2924346923828125,
+      "learning_rate": 1.665704384296557e-05,
+      "loss": 1.0433,
+      "step": 40675
+    },
+    {
+      "epoch": 12.17,
+      "grad_norm": 3.4734883308410645,
+      "learning_rate": 1.6651505671715628e-05,
+      "loss": 1.083,
+      "step": 40680
+    },
+    {
+      "epoch": 12.17,
+      "grad_norm": 3.842501401901245,
+      "learning_rate": 1.6645967961544357e-05,
+      "loss": 0.8934,
+      "step": 40685
+    },
+    {
+      "epoch": 12.17,
+      "grad_norm": 3.126307725906372,
+      "learning_rate": 1.6640430712757594e-05,
+      "loss": 0.8845,
+      "step": 40690
+    },
+    {
+      "epoch": 12.18,
+      "grad_norm": 3.3595333099365234,
+      "learning_rate": 1.6634893925661142e-05,
+      "loss": 1.0184,
+      "step": 40695
+    },
+    {
+      "epoch": 12.18,
+      "grad_norm": 2.7751407623291016,
+      "learning_rate": 1.662935760056082e-05,
+      "loss": 0.8099,
+      "step": 40700
+    },
+    {
+      "epoch": 12.18,
+      "grad_norm": 4.374361038208008,
+      "learning_rate": 1.6623821737762362e-05,
+      "loss": 0.89,
+      "step": 40705
+    },
+    {
+      "epoch": 12.18,
+      "grad_norm": 3.8457541465759277,
+      "learning_rate": 1.6618286337571532e-05,
+      "loss": 1.0311,
+      "step": 40710
+    },
+    {
+      "epoch": 12.18,
+      "grad_norm": 4.070710182189941,
+      "learning_rate": 1.661275140029404e-05,
+      "loss": 1.0774,
+      "step": 40715
+    },
+    {
+      "epoch": 12.18,
+      "grad_norm": 1.868833303451538,
+      "learning_rate": 1.6607216926235552e-05,
+      "loss": 1.1075,
+      "step": 40720
+    },
+    {
+      "epoch": 12.18,
+      "grad_norm": 2.962254524230957,
+      "learning_rate": 1.6601682915701767e-05,
+      "loss": 0.7483,
+      "step": 40725
+    },
+    {
+      "epoch": 12.19,
+      "grad_norm": 1.097212553024292,
+      "learning_rate": 1.659614936899829e-05,
+      "loss": 0.9267,
+      "step": 40730
+    },
+    {
+      "epoch": 12.19,
+      "grad_norm": 1.6510802507400513,
+      "learning_rate": 1.6590616286430754e-05,
+      "loss": 1.0181,
+      "step": 40735
+    },
+    {
+      "epoch": 12.19,
+      "grad_norm": 2.0439059734344482,
+      "learning_rate": 1.658508366830474e-05,
+      "loss": 1.1101,
+      "step": 40740
+    },
+    {
+      "epoch": 12.19,
+      "grad_norm": 3.3008670806884766,
+      "learning_rate": 1.6579551514925812e-05,
+      "loss": 1.0038,
+      "step": 40745
+    },
+    {
+      "epoch": 12.19,
+      "grad_norm": 1.4634921550750732,
+      "learning_rate": 1.65740198265995e-05,
+      "loss": 1.0533,
+      "step": 40750
+    },
+    {
+      "epoch": 12.19,
+      "grad_norm": 2.8518662452697754,
+      "learning_rate": 1.656848860363131e-05,
+      "loss": 1.0689,
+      "step": 40755
+    },
+    {
+      "epoch": 12.19,
+      "grad_norm": 2.9406023025512695,
+      "learning_rate": 1.6562957846326738e-05,
+      "loss": 1.1724,
+      "step": 40760
+    },
+    {
+      "epoch": 12.2,
+      "grad_norm": 4.297204971313477,
+      "learning_rate": 1.6557427554991222e-05,
+      "loss": 1.0288,
+      "step": 40765
+    },
+    {
+      "epoch": 12.2,
+      "grad_norm": 1.472741723060608,
+      "learning_rate": 1.655189772993022e-05,
+      "loss": 1.1269,
+      "step": 40770
+    },
+    {
+      "epoch": 12.2,
+      "grad_norm": 2.547549247741699,
+      "learning_rate": 1.6546368371449115e-05,
+      "loss": 1.1671,
+      "step": 40775
+    },
+    {
+      "epoch": 12.2,
+      "grad_norm": 2.158825159072876,
+      "learning_rate": 1.6540839479853305e-05,
+      "loss": 1.0053,
+      "step": 40780
+    },
+    {
+      "epoch": 12.2,
+      "grad_norm": 2.1777453422546387,
+      "learning_rate": 1.653531105544814e-05,
+      "loss": 1.0576,
+      "step": 40785
+    },
+    {
+      "epoch": 12.2,
+      "grad_norm": 3.722043991088867,
+      "learning_rate": 1.6529783098538937e-05,
+      "loss": 1.0552,
+      "step": 40790
+    },
+    {
+      "epoch": 12.21,
+      "grad_norm": 9.3829984664917,
+      "learning_rate": 1.6524255609431018e-05,
+      "loss": 0.8906,
+      "step": 40795
+    },
+    {
+      "epoch": 12.21,
+      "grad_norm": 1.8850293159484863,
+      "learning_rate": 1.6518728588429643e-05,
+      "loss": 0.9639,
+      "step": 40800
+    },
+    {
+      "epoch": 12.21,
+      "grad_norm": 4.943428993225098,
+      "learning_rate": 1.651320203584008e-05,
+      "loss": 0.8564,
+      "step": 40805
+    },
+    {
+      "epoch": 12.21,
+      "grad_norm": 1.9969452619552612,
+      "learning_rate": 1.650767595196754e-05,
+      "loss": 0.9259,
+      "step": 40810
+    },
+    {
+      "epoch": 12.21,
+      "grad_norm": 3.900078773498535,
+      "learning_rate": 1.650215033711724e-05,
+      "loss": 1.0503,
+      "step": 40815
+    },
+    {
+      "epoch": 12.21,
+      "grad_norm": 6.483584880828857,
+      "learning_rate": 1.6496625191594335e-05,
+      "loss": 1.1011,
+      "step": 40820
+    },
+    {
+      "epoch": 12.21,
+      "grad_norm": 3.2015042304992676,
+      "learning_rate": 1.6491100515703984e-05,
+      "loss": 0.9268,
+      "step": 40825
+    },
+    {
+      "epoch": 12.22,
+      "grad_norm": 1.9813512563705444,
+      "learning_rate": 1.6485576309751304e-05,
+      "loss": 0.7646,
+      "step": 40830
+    },
+    {
+      "epoch": 12.22,
+      "grad_norm": 2.6845803260803223,
+      "learning_rate": 1.6480052574041398e-05,
+      "loss": 1.1781,
+      "step": 40835
+    },
+    {
+      "epoch": 12.22,
+      "grad_norm": 5.457139492034912,
+      "learning_rate": 1.647452930887933e-05,
+      "loss": 1.0875,
+      "step": 40840
+    },
+    {
+      "epoch": 12.22,
+      "grad_norm": 2.0619680881500244,
+      "learning_rate": 1.6469006514570158e-05,
+      "loss": 1.136,
+      "step": 40845
+    },
+    {
+      "epoch": 12.22,
+      "grad_norm": 3.572672128677368,
+      "learning_rate": 1.646348419141887e-05,
+      "loss": 0.893,
+      "step": 40850
+    },
+    {
+      "epoch": 12.22,
+      "grad_norm": 3.076113700866699,
+      "learning_rate": 1.6457962339730492e-05,
+      "loss": 0.948,
+      "step": 40855
+    },
+    {
+      "epoch": 12.22,
+      "grad_norm": 2.0957534313201904,
+      "learning_rate": 1.6452440959809965e-05,
+      "loss": 0.9835,
+      "step": 40860
+    },
+    {
+      "epoch": 12.23,
+      "grad_norm": 5.54816198348999,
+      "learning_rate": 1.6446920051962247e-05,
+      "loss": 0.9175,
+      "step": 40865
+    },
+    {
+      "epoch": 12.23,
+      "grad_norm": 2.4465866088867188,
+      "learning_rate": 1.6441399616492238e-05,
+      "loss": 1.0556,
+      "step": 40870
+    },
+    {
+      "epoch": 12.23,
+      "grad_norm": 2.606534957885742,
+      "learning_rate": 1.6435879653704835e-05,
+      "loss": 1.0207,
+      "step": 40875
+    },
+    {
+      "epoch": 12.23,
+      "grad_norm": 2.173264980316162,
+      "learning_rate": 1.6430360163904902e-05,
+      "loss": 1.1046,
+      "step": 40880
+    },
+    {
+      "epoch": 12.23,
+      "grad_norm": 2.238013744354248,
+      "learning_rate": 1.6424841147397256e-05,
+      "loss": 1.0992,
+      "step": 40885
+    },
+    {
+      "epoch": 12.23,
+      "grad_norm": 1.3038547039031982,
+      "learning_rate": 1.6419322604486737e-05,
+      "loss": 0.9191,
+      "step": 40890
+    },
+    {
+      "epoch": 12.24,
+      "grad_norm": 1.2264683246612549,
+      "learning_rate": 1.6413804535478095e-05,
+      "loss": 1.0838,
+      "step": 40895
+    },
+    {
+      "epoch": 12.24,
+      "grad_norm": 4.56827974319458,
+      "learning_rate": 1.6408286940676114e-05,
+      "loss": 0.986,
+      "step": 40900
+    },
+    {
+      "epoch": 12.24,
+      "grad_norm": 1.0923278331756592,
+      "learning_rate": 1.6402769820385504e-05,
+      "loss": 1.0315,
+      "step": 40905
+    },
+    {
+      "epoch": 12.24,
+      "grad_norm": 1.2232002019882202,
+      "learning_rate": 1.6397253174910997e-05,
+      "loss": 1.0501,
+      "step": 40910
+    },
+    {
+      "epoch": 12.24,
+      "grad_norm": 1.501046895980835,
+      "learning_rate": 1.6391737004557246e-05,
+      "loss": 0.9132,
+      "step": 40915
+    },
+    {
+      "epoch": 12.24,
+      "grad_norm": 0.9044210910797119,
+      "learning_rate": 1.638622130962891e-05,
+      "loss": 1.017,
+      "step": 40920
+    },
+    {
+      "epoch": 12.24,
+      "grad_norm": 2.844947576522827,
+      "learning_rate": 1.638070609043062e-05,
+      "loss": 0.9969,
+      "step": 40925
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 4.111929893493652,
+      "learning_rate": 1.637519134726697e-05,
+      "loss": 1.0534,
+      "step": 40930
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 1.8732011318206787,
+      "learning_rate": 1.636967708044254e-05,
+      "loss": 0.888,
+      "step": 40935
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 2.614795446395874,
+      "learning_rate": 1.636416329026188e-05,
+      "loss": 0.9893,
+      "step": 40940
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 2.9741506576538086,
+      "learning_rate": 1.6358649977029493e-05,
+      "loss": 1.0112,
+      "step": 40945
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 4.53338098526001,
+      "learning_rate": 1.63531371410499e-05,
+      "loss": 0.9482,
+      "step": 40950
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 1.6804264783859253,
+      "learning_rate": 1.634762478262754e-05,
+      "loss": 0.9047,
+      "step": 40955
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 5.280777931213379,
+      "learning_rate": 1.634211290206688e-05,
+      "loss": 0.8929,
+      "step": 40960
+    },
+    {
+      "epoch": 12.26,
+      "grad_norm": 2.2232463359832764,
+      "learning_rate": 1.6336601499672316e-05,
+      "loss": 1.0002,
+      "step": 40965
+    },
+    {
+      "epoch": 12.26,
+      "grad_norm": 3.6429049968719482,
+      "learning_rate": 1.633109057574826e-05,
+      "loss": 0.9099,
+      "step": 40970
+    },
+    {
+      "epoch": 12.26,
+      "grad_norm": 4.826415538787842,
+      "learning_rate": 1.6325580130599054e-05,
+      "loss": 0.9408,
+      "step": 40975
+    },
+    {
+      "epoch": 12.26,
+      "grad_norm": 1.3283343315124512,
+      "learning_rate": 1.6320070164529033e-05,
+      "loss": 0.9633,
+      "step": 40980
+    },
+    {
+      "epoch": 12.26,
+      "grad_norm": 4.262852191925049,
+      "learning_rate": 1.6314560677842526e-05,
+      "loss": 1.0573,
+      "step": 40985
+    },
+    {
+      "epoch": 12.26,
+      "grad_norm": 1.9915564060211182,
+      "learning_rate": 1.6309051670843794e-05,
+      "loss": 0.9936,
+      "step": 40990
+    },
+    {
+      "epoch": 12.27,
+      "grad_norm": 2.673431634902954,
+      "learning_rate": 1.6303543143837113e-05,
+      "loss": 0.9722,
+      "step": 40995
+    },
+    {
+      "epoch": 12.27,
+      "grad_norm": 2.525904417037964,
+      "learning_rate": 1.6298035097126698e-05,
+      "loss": 1.1012,
+      "step": 41000
+    },
+    {
+      "epoch": 12.27,
+      "grad_norm": 2.020827054977417,
+      "learning_rate": 1.629252753101677e-05,
+      "loss": 1.0946,
+      "step": 41005
+    },
+    {
+      "epoch": 12.27,
+      "grad_norm": 2.497849941253662,
+      "learning_rate": 1.6287020445811485e-05,
+      "loss": 1.1731,
+      "step": 41010
+    },
+    {
+      "epoch": 12.27,
+      "grad_norm": 2.420525312423706,
+      "learning_rate": 1.6281513841815006e-05,
+      "loss": 0.9534,
+      "step": 41015
+    },
+    {
+      "epoch": 12.27,
+      "grad_norm": 1.6529130935668945,
+      "learning_rate": 1.627600771933146e-05,
+      "loss": 1.0972,
+      "step": 41020
+    },
+    {
+      "epoch": 12.27,
+      "grad_norm": 2.2307538986206055,
+      "learning_rate": 1.6270502078664927e-05,
+      "loss": 0.9837,
+      "step": 41025
+    },
+    {
+      "epoch": 12.28,
+      "grad_norm": 4.9482269287109375,
+      "learning_rate": 1.6264996920119507e-05,
+      "loss": 0.7897,
+      "step": 41030
+    },
+    {
+      "epoch": 12.28,
+      "grad_norm": 1.9426223039627075,
+      "learning_rate": 1.6259492243999215e-05,
+      "loss": 1.088,
+      "step": 41035
+    },
+    {
+      "epoch": 12.28,
+      "grad_norm": 2.4098875522613525,
+      "learning_rate": 1.6253988050608092e-05,
+      "loss": 1.014,
+      "step": 41040
+    },
+    {
+      "epoch": 12.28,
+      "grad_norm": 3.1415224075317383,
+      "learning_rate": 1.6248484340250114e-05,
+      "loss": 1.0839,
+      "step": 41045
+    },
+    {
+      "epoch": 12.28,
+      "grad_norm": 3.377718687057495,
+      "learning_rate": 1.6242981113229245e-05,
+      "loss": 0.8829,
+      "step": 41050
+    },
+    {
+      "epoch": 12.28,
+      "grad_norm": 2.4279189109802246,
+      "learning_rate": 1.6237478369849433e-05,
+      "loss": 1.1367,
+      "step": 41055
+    },
+    {
+      "epoch": 12.28,
+      "grad_norm": 5.114649295806885,
+      "learning_rate": 1.6231976110414574e-05,
+      "loss": 1.0176,
+      "step": 41060
+    },
+    {
+      "epoch": 12.29,
+      "grad_norm": 3.2667107582092285,
+      "learning_rate": 1.622647433522857e-05,
+      "loss": 1.0747,
+      "step": 41065
+    },
+    {
+      "epoch": 12.29,
+      "grad_norm": 3.784885883331299,
+      "learning_rate": 1.6220973044595267e-05,
+      "loss": 1.0822,
+      "step": 41070
+    },
+    {
+      "epoch": 12.29,
+      "grad_norm": 2.1055243015289307,
+      "learning_rate": 1.621547223881849e-05,
+      "loss": 1.1663,
+      "step": 41075
+    },
+    {
+      "epoch": 12.29,
+      "grad_norm": 1.5692771673202515,
+      "learning_rate": 1.620997191820206e-05,
+      "loss": 0.9983,
+      "step": 41080
+    },
+    {
+      "epoch": 12.29,
+      "grad_norm": 2.18072772026062,
+      "learning_rate": 1.620447208304973e-05,
+      "loss": 0.805,
+      "step": 41085
+    },
+    {
+      "epoch": 12.29,
+      "grad_norm": 6.3626933097839355,
+      "learning_rate": 1.6198972733665284e-05,
+      "loss": 0.8722,
+      "step": 41090
+    },
+    {
+      "epoch": 12.3,
+      "grad_norm": 2.6465694904327393,
+      "learning_rate": 1.6193473870352408e-05,
+      "loss": 1.124,
+      "step": 41095
+    },
+    {
+      "epoch": 12.3,
+      "grad_norm": 1.850497841835022,
+      "learning_rate": 1.6187975493414825e-05,
+      "loss": 0.909,
+      "step": 41100
+    },
+    {
+      "epoch": 12.3,
+      "grad_norm": 3.717559576034546,
+      "learning_rate": 1.61824776031562e-05,
+      "loss": 0.9344,
+      "step": 41105
+    },
+    {
+      "epoch": 12.3,
+      "grad_norm": 1.314609408378601,
+      "learning_rate": 1.6176980199880158e-05,
+      "loss": 0.9273,
+      "step": 41110
+    },
+    {
+      "epoch": 12.3,
+      "grad_norm": 1.5568089485168457,
+      "learning_rate": 1.6171483283890342e-05,
+      "loss": 1.0734,
+      "step": 41115
+    },
+    {
+      "epoch": 12.3,
+      "grad_norm": 4.105639457702637,
+      "learning_rate": 1.6165986855490316e-05,
+      "loss": 0.9675,
+      "step": 41120
+    },
+    {
+      "epoch": 12.3,
+      "grad_norm": 4.530340194702148,
+      "learning_rate": 1.6160490914983667e-05,
+      "loss": 1.1482,
+      "step": 41125
+    },
+    {
+      "epoch": 12.31,
+      "grad_norm": Infinity,
+      "learning_rate": 1.6156094514065535e-05,
+      "loss": 1.1886,
+      "step": 41130
+    },
+    {
+      "epoch": 12.31,
+      "grad_norm": 6.879144191741943,
+      "learning_rate": 1.615059945253183e-05,
+      "loss": 1.0541,
+      "step": 41135
+    },
+    {
+      "epoch": 12.31,
+      "grad_norm": 5.159574508666992,
+      "learning_rate": 1.6145104879741307e-05,
+      "loss": 1.0869,
+      "step": 41140
+    },
+    {
+      "epoch": 12.31,
+      "grad_norm": 3.31752872467041,
+      "learning_rate": 1.6139610795997448e-05,
+      "loss": 0.8379,
+      "step": 41145
+    },
+    {
+      "epoch": 12.31,
+      "grad_norm": 1.4643511772155762,
+      "learning_rate": 1.6134117201603662e-05,
+      "loss": 1.0015,
+      "step": 41150
+    },
+    {
+      "epoch": 12.31,
+      "grad_norm": 1.95048189163208,
+      "learning_rate": 1.612862409686338e-05,
+      "loss": 1.1556,
+      "step": 41155
+    },
+    {
+      "epoch": 12.31,
+      "grad_norm": 3.2634193897247314,
+      "learning_rate": 1.6123131482079962e-05,
+      "loss": 1.2138,
+      "step": 41160
+    },
+    {
+      "epoch": 12.32,
+      "grad_norm": 2.3244948387145996,
+      "learning_rate": 1.6117639357556767e-05,
+      "loss": 1.0905,
+      "step": 41165
+    },
+    {
+      "epoch": 12.32,
+      "grad_norm": 5.133475303649902,
+      "learning_rate": 1.6112147723597116e-05,
+      "loss": 1.0629,
+      "step": 41170
+    },
+    {
+      "epoch": 12.32,
+      "grad_norm": 2.374720811843872,
+      "learning_rate": 1.610665658050431e-05,
+      "loss": 1.0795,
+      "step": 41175
+    },
+    {
+      "epoch": 12.32,
+      "grad_norm": 2.306103229522705,
+      "learning_rate": 1.6101165928581612e-05,
+      "loss": 1.022,
+      "step": 41180
+    },
+    {
+      "epoch": 12.32,
+      "grad_norm": 2.1258509159088135,
+      "learning_rate": 1.6095675768132273e-05,
+      "loss": 0.8965,
+      "step": 41185
+    },
+    {
+      "epoch": 12.32,
+      "grad_norm": 4.365717887878418,
+      "learning_rate": 1.6090186099459505e-05,
+      "loss": 1.1589,
+      "step": 41190
+    },
+    {
+      "epoch": 12.33,
+      "grad_norm": 3.533360481262207,
+      "learning_rate": 1.6084696922866504e-05,
+      "loss": 0.9771,
+      "step": 41195
+    },
+    {
+      "epoch": 12.33,
+      "grad_norm": 1.57905113697052,
+      "learning_rate": 1.6079208238656414e-05,
+      "loss": 1.1706,
+      "step": 41200
+    },
+    {
+      "epoch": 12.33,
+      "grad_norm": 3.772719144821167,
+      "learning_rate": 1.607372004713239e-05,
+      "loss": 0.9388,
+      "step": 41205
+    },
+    {
+      "epoch": 12.33,
+      "grad_norm": 2.236943006515503,
+      "learning_rate": 1.606823234859752e-05,
+      "loss": 1.0367,
+      "step": 41210
+    },
+    {
+      "epoch": 12.33,
+      "grad_norm": 1.4967836141586304,
+      "learning_rate": 1.6062745143354903e-05,
+      "loss": 0.9601,
+      "step": 41215
+    },
+    {
+      "epoch": 12.33,
+      "grad_norm": 1.7550181150436401,
+      "learning_rate": 1.6057258431707585e-05,
+      "loss": 1.1271,
+      "step": 41220
+    },
+    {
+      "epoch": 12.33,
+      "grad_norm": 1.6419475078582764,
+      "learning_rate": 1.6051772213958575e-05,
+      "loss": 1.1171,
+      "step": 41225
+    },
+    {
+      "epoch": 12.34,
+      "grad_norm": 3.1746506690979004,
+      "learning_rate": 1.6046286490410895e-05,
+      "loss": 0.9319,
+      "step": 41230
+    },
+    {
+      "epoch": 12.34,
+      "grad_norm": 4.4260993003845215,
+      "learning_rate": 1.6040801261367493e-05,
+      "loss": 0.9774,
+      "step": 41235
+    },
+    {
+      "epoch": 12.34,
+      "grad_norm": 1.5141162872314453,
+      "learning_rate": 1.603531652713134e-05,
+      "loss": 1.0768,
+      "step": 41240
+    },
+    {
+      "epoch": 12.34,
+      "grad_norm": 1.775514841079712,
+      "learning_rate": 1.602983228800532e-05,
+      "loss": 0.874,
+      "step": 41245
+    },
+    {
+      "epoch": 12.34,
+      "grad_norm": 5.238076210021973,
+      "learning_rate": 1.6024348544292357e-05,
+      "loss": 0.9991,
+      "step": 41250
+    },
+    {
+      "epoch": 12.34,
+      "grad_norm": 2.956698417663574,
+      "learning_rate": 1.6018865296295284e-05,
+      "loss": 1.137,
+      "step": 41255
+    },
+    {
+      "epoch": 12.34,
+      "grad_norm": 3.052151918411255,
+      "learning_rate": 1.6013382544316947e-05,
+      "loss": 1.0055,
+      "step": 41260
+    },
+    {
+      "epoch": 12.35,
+      "grad_norm": 7.823723316192627,
+      "learning_rate": 1.6007900288660148e-05,
+      "loss": 0.9589,
+      "step": 41265
+    },
+    {
+      "epoch": 12.35,
+      "grad_norm": 2.2807137966156006,
+      "learning_rate": 1.6002418529627673e-05,
+      "loss": 0.9707,
+      "step": 41270
+    },
+    {
+      "epoch": 12.35,
+      "grad_norm": 5.3980207443237305,
+      "learning_rate": 1.5996937267522265e-05,
+      "loss": 1.0007,
+      "step": 41275
+    },
+    {
+      "epoch": 12.35,
+      "grad_norm": 4.970637321472168,
+      "learning_rate": 1.5991456502646658e-05,
+      "loss": 0.8831,
+      "step": 41280
+    },
+    {
+      "epoch": 12.35,
+      "grad_norm": 2.0372869968414307,
+      "learning_rate": 1.5985976235303547e-05,
+      "loss": 1.0731,
+      "step": 41285
+    },
+    {
+      "epoch": 12.35,
+      "grad_norm": 1.5453438758850098,
+      "learning_rate": 1.59804964657956e-05,
+      "loss": 1.1497,
+      "step": 41290
+    },
+    {
+      "epoch": 12.35,
+      "grad_norm": 3.3459970951080322,
+      "learning_rate": 1.5975017194425448e-05,
+      "loss": 0.9553,
+      "step": 41295
+    },
+    {
+      "epoch": 12.36,
+      "grad_norm": 1.7910807132720947,
+      "learning_rate": 1.5969538421495728e-05,
+      "loss": 0.9187,
+      "step": 41300
+    },
+    {
+      "epoch": 12.36,
+      "grad_norm": 2.4004948139190674,
+      "learning_rate": 1.5964060147309e-05,
+      "loss": 0.8181,
+      "step": 41305
+    },
+    {
+      "epoch": 12.36,
+      "grad_norm": 3.5690810680389404,
+      "learning_rate": 1.5958582372167853e-05,
+      "loss": 1.1377,
+      "step": 41310
+    },
+    {
+      "epoch": 12.36,
+      "grad_norm": 7.011974334716797,
+      "learning_rate": 1.5953105096374797e-05,
+      "loss": 0.9167,
+      "step": 41315
+    },
+    {
+      "epoch": 12.36,
+      "grad_norm": 2.1069743633270264,
+      "learning_rate": 1.594762832023234e-05,
+      "loss": 1.0594,
+      "step": 41320
+    },
+    {
+      "epoch": 12.36,
+      "grad_norm": 1.5728492736816406,
+      "learning_rate": 1.594215204404297e-05,
+      "loss": 1.1337,
+      "step": 41325
+    },
+    {
+      "epoch": 12.37,
+      "grad_norm": 3.168928384780884,
+      "learning_rate": 1.5936676268109113e-05,
+      "loss": 0.933,
+      "step": 41330
+    },
+    {
+      "epoch": 12.37,
+      "grad_norm": 1.6599355936050415,
+      "learning_rate": 1.5931200992733217e-05,
+      "loss": 1.1395,
+      "step": 41335
+    },
+    {
+      "epoch": 12.37,
+      "grad_norm": 3.265467643737793,
+      "learning_rate": 1.5925726218217653e-05,
+      "loss": 1.0227,
+      "step": 41340
+    },
+    {
+      "epoch": 12.37,
+      "grad_norm": 3.9105184078216553,
+      "learning_rate": 1.5920251944864812e-05,
+      "loss": 1.0851,
+      "step": 41345
+    },
+    {
+      "epoch": 12.37,
+      "grad_norm": 7.27264928817749,
+      "learning_rate": 1.5914778172977008e-05,
+      "loss": 1.0477,
+      "step": 41350
+    },
+    {
+      "epoch": 12.37,
+      "grad_norm": 3.6267623901367188,
+      "learning_rate": 1.5909304902856563e-05,
+      "loss": 1.0303,
+      "step": 41355
+    },
+    {
+      "epoch": 12.37,
+      "grad_norm": 3.7175920009613037,
+      "learning_rate": 1.590383213480576e-05,
+      "loss": 1.0979,
+      "step": 41360
+    },
+    {
+      "epoch": 12.38,
+      "grad_norm": 1.885209321975708,
+      "learning_rate": 1.589835986912685e-05,
+      "loss": 1.1829,
+      "step": 41365
+    },
+    {
+      "epoch": 12.38,
+      "grad_norm": 4.020153045654297,
+      "learning_rate": 1.5892888106122073e-05,
+      "loss": 1.1045,
+      "step": 41370
+    },
+    {
+      "epoch": 12.38,
+      "grad_norm": 1.572926640510559,
+      "learning_rate": 1.5887416846093605e-05,
+      "loss": 0.992,
+      "step": 41375
+    },
+    {
+      "epoch": 12.38,
+      "grad_norm": 3.5849032402038574,
+      "learning_rate": 1.5881946089343646e-05,
+      "loss": 1.1051,
+      "step": 41380
+    },
+    {
+      "epoch": 12.38,
+      "grad_norm": 2.3222670555114746,
+      "learning_rate": 1.5876475836174328e-05,
+      "loss": 0.9915,
+      "step": 41385
+    },
+    {
+      "epoch": 12.38,
+      "grad_norm": 2.578094482421875,
+      "learning_rate": 1.5871006086887756e-05,
+      "loss": 0.9607,
+      "step": 41390
+    },
+    {
+      "epoch": 12.38,
+      "grad_norm": 4.564905166625977,
+      "learning_rate": 1.586553684178604e-05,
+      "loss": 1.0403,
+      "step": 41395
+    },
+    {
+      "epoch": 12.39,
+      "grad_norm": 4.259903907775879,
+      "learning_rate": 1.5860068101171214e-05,
+      "loss": 0.9894,
+      "step": 41400
+    },
+    {
+      "epoch": 12.39,
+      "grad_norm": 3.7057387828826904,
+      "learning_rate": 1.5854599865345342e-05,
+      "loss": 1.1087,
+      "step": 41405
+    },
+    {
+      "epoch": 12.39,
+      "grad_norm": 2.632147789001465,
+      "learning_rate": 1.58491321346104e-05,
+      "loss": 1.029,
+      "step": 41410
+    },
+    {
+      "epoch": 12.39,
+      "grad_norm": 1.1085238456726074,
+      "learning_rate": 1.5843664909268392e-05,
+      "loss": 1.0203,
+      "step": 41415
+    },
+    {
+      "epoch": 12.39,
+      "grad_norm": 3.5292742252349854,
+      "learning_rate": 1.583819818962125e-05,
+      "loss": 0.9154,
+      "step": 41420
+    },
+    {
+      "epoch": 12.39,
+      "grad_norm": 4.0045552253723145,
+      "learning_rate": 1.5832731975970887e-05,
+      "loss": 1.0444,
+      "step": 41425
+    },
+    {
+      "epoch": 12.4,
+      "grad_norm": 1.9779725074768066,
+      "learning_rate": 1.5827266268619224e-05,
+      "loss": 1.0704,
+      "step": 41430
+    },
+    {
+      "epoch": 12.4,
+      "grad_norm": 4.693383693695068,
+      "learning_rate": 1.5821801067868096e-05,
+      "loss": 0.9821,
+      "step": 41435
+    },
+    {
+      "epoch": 12.4,
+      "grad_norm": 1.2160636186599731,
+      "learning_rate": 1.5816336374019363e-05,
+      "loss": 0.8789,
+      "step": 41440
+    },
+    {
+      "epoch": 12.4,
+      "grad_norm": 1.0221775770187378,
+      "learning_rate": 1.581087218737483e-05,
+      "loss": 1.0346,
+      "step": 41445
+    },
+    {
+      "epoch": 12.4,
+      "grad_norm": 3.203967809677124,
+      "learning_rate": 1.5805408508236263e-05,
+      "loss": 0.8631,
+      "step": 41450
+    },
+    {
+      "epoch": 12.4,
+      "grad_norm": 10.920624732971191,
+      "learning_rate": 1.5799945336905438e-05,
+      "loss": 0.9483,
+      "step": 41455
+    },
+    {
+      "epoch": 12.4,
+      "grad_norm": 1.881516456604004,
+      "learning_rate": 1.5794482673684056e-05,
+      "loss": 1.0587,
+      "step": 41460
+    },
+    {
+      "epoch": 12.41,
+      "grad_norm": 3.7939016819000244,
+      "learning_rate": 1.5789020518873842e-05,
+      "loss": 1.1071,
+      "step": 41465
+    },
+    {
+      "epoch": 12.41,
+      "grad_norm": 2.4270174503326416,
+      "learning_rate": 1.5783558872776438e-05,
+      "loss": 1.1607,
+      "step": 41470
+    },
+    {
+      "epoch": 12.41,
+      "grad_norm": 3.0483105182647705,
+      "learning_rate": 1.5778097735693508e-05,
+      "loss": 0.9586,
+      "step": 41475
+    },
+    {
+      "epoch": 12.41,
+      "grad_norm": 3.0462000370025635,
+      "learning_rate": 1.5772637107926658e-05,
+      "loss": 1.072,
+      "step": 41480
+    },
+    {
+      "epoch": 12.41,
+      "grad_norm": 1.6828171014785767,
+      "learning_rate": 1.5767176989777455e-05,
+      "loss": 1.138,
+      "step": 41485
+    },
+    {
+      "epoch": 12.41,
+      "grad_norm": 3.1212778091430664,
+      "learning_rate": 1.576171738154748e-05,
+      "loss": 0.9104,
+      "step": 41490
+    },
+    {
+      "epoch": 12.41,
+      "grad_norm": 3.6774206161499023,
+      "learning_rate": 1.5756258283538243e-05,
+      "loss": 0.9859,
+      "step": 41495
+    },
+    {
+      "epoch": 12.42,
+      "grad_norm": 5.639192581176758,
+      "learning_rate": 1.575079969605126e-05,
+      "loss": 1.0641,
+      "step": 41500
+    },
+    {
+      "epoch": 12.42,
+      "grad_norm": 2.7606070041656494,
+      "learning_rate": 1.5745341619387986e-05,
+      "loss": 0.9769,
+      "step": 41505
+    },
+    {
+      "epoch": 12.42,
+      "grad_norm": 2.2884132862091064,
+      "learning_rate": 1.573988405384989e-05,
+      "loss": 1.0192,
+      "step": 41510
+    },
+    {
+      "epoch": 12.42,
+      "grad_norm": 2.4269514083862305,
+      "learning_rate": 1.573442699973837e-05,
+      "loss": 1.0783,
+      "step": 41515
+    },
+    {
+      "epoch": 12.42,
+      "grad_norm": 4.481752872467041,
+      "learning_rate": 1.5728970457354802e-05,
+      "loss": 1.0585,
+      "step": 41520
+    },
+    {
+      "epoch": 12.42,
+      "grad_norm": 2.6316633224487305,
+      "learning_rate": 1.572351442700057e-05,
+      "loss": 1.1228,
+      "step": 41525
+    },
+    {
+      "epoch": 12.43,
+      "grad_norm": 3.698345184326172,
+      "learning_rate": 1.5718058908976988e-05,
+      "loss": 1.0415,
+      "step": 41530
+    },
+    {
+      "epoch": 12.43,
+      "grad_norm": 3.938213348388672,
+      "learning_rate": 1.5712603903585367e-05,
+      "loss": 0.9013,
+      "step": 41535
+    },
+    {
+      "epoch": 12.43,
+      "grad_norm": 2.3214263916015625,
+      "learning_rate": 1.5707149411126975e-05,
+      "loss": 0.9264,
+      "step": 41540
+    },
+    {
+      "epoch": 12.43,
+      "grad_norm": 2.7372500896453857,
+      "learning_rate": 1.5701695431903068e-05,
+      "loss": 0.8975,
+      "step": 41545
+    },
+    {
+      "epoch": 12.43,
+      "grad_norm": 1.4826568365097046,
+      "learning_rate": 1.569624196621486e-05,
+      "loss": 1.1417,
+      "step": 41550
+    },
+    {
+      "epoch": 12.43,
+      "grad_norm": 2.281169891357422,
+      "learning_rate": 1.5690789014363526e-05,
+      "loss": 0.9454,
+      "step": 41555
+    },
+    {
+      "epoch": 12.43,
+      "grad_norm": 1.4119058847427368,
+      "learning_rate": 1.568533657665025e-05,
+      "loss": 1.0374,
+      "step": 41560
+    },
+    {
+      "epoch": 12.44,
+      "grad_norm": 14.846397399902344,
+      "learning_rate": 1.5679884653376138e-05,
+      "loss": 0.9577,
+      "step": 41565
+    },
+    {
+      "epoch": 12.44,
+      "grad_norm": 5.682839393615723,
+      "learning_rate": 1.5674433244842324e-05,
+      "loss": 0.8656,
+      "step": 41570
+    },
+    {
+      "epoch": 12.44,
+      "grad_norm": 6.977272033691406,
+      "learning_rate": 1.566898235134987e-05,
+      "loss": 1.0464,
+      "step": 41575
+    },
+    {
+      "epoch": 12.44,
+      "grad_norm": 3.747130870819092,
+      "learning_rate": 1.5663531973199807e-05,
+      "loss": 0.8694,
+      "step": 41580
+    },
+    {
+      "epoch": 12.44,
+      "grad_norm": 3.2453455924987793,
+      "learning_rate": 1.565808211069318e-05,
+      "loss": 1.191,
+      "step": 41585
+    },
+    {
+      "epoch": 12.44,
+      "grad_norm": 1.451064944267273,
+      "learning_rate": 1.565263276413096e-05,
+      "loss": 1.0963,
+      "step": 41590
+    },
+    {
+      "epoch": 12.44,
+      "grad_norm": 1.9707701206207275,
+      "learning_rate": 1.5647183933814124e-05,
+      "loss": 0.9704,
+      "step": 41595
+    },
+    {
+      "epoch": 12.45,
+      "grad_norm": 2.4262611865997314,
+      "learning_rate": 1.5641735620043586e-05,
+      "loss": 0.9327,
+      "step": 41600
+    },
+    {
+      "epoch": 12.45,
+      "grad_norm": 2.2412772178649902,
+      "learning_rate": 1.5636287823120278e-05,
+      "loss": 1.0346,
+      "step": 41605
+    },
+    {
+      "epoch": 12.45,
+      "grad_norm": 3.1475157737731934,
+      "learning_rate": 1.5630840543345048e-05,
+      "loss": 0.9991,
+      "step": 41610
+    },
+    {
+      "epoch": 12.45,
+      "grad_norm": 5.283227920532227,
+      "learning_rate": 1.562539378101876e-05,
+      "loss": 0.9999,
+      "step": 41615
+    },
+    {
+      "epoch": 12.45,
+      "grad_norm": 3.1468896865844727,
+      "learning_rate": 1.561994753644223e-05,
+      "loss": 1.0675,
+      "step": 41620
+    },
+    {
+      "epoch": 12.45,
+      "grad_norm": 6.32867431640625,
+      "learning_rate": 1.5614501809916245e-05,
+      "loss": 1.0708,
+      "step": 41625
+    },
+    {
+      "epoch": 12.46,
+      "grad_norm": 3.040778875350952,
+      "learning_rate": 1.5609056601741573e-05,
+      "loss": 1.26,
+      "step": 41630
+    },
+    {
+      "epoch": 12.46,
+      "grad_norm": 3.2033233642578125,
+      "learning_rate": 1.560361191221894e-05,
+      "loss": 1.1999,
+      "step": 41635
+    },
+    {
+      "epoch": 12.46,
+      "grad_norm": 1.406636118888855,
+      "learning_rate": 1.5598167741649054e-05,
+      "loss": 0.8328,
+      "step": 41640
+    },
+    {
+      "epoch": 12.46,
+      "grad_norm": 2.998223066329956,
+      "learning_rate": 1.55927240903326e-05,
+      "loss": 1.0981,
+      "step": 41645
+    },
+    {
+      "epoch": 12.46,
+      "grad_norm": 4.719427108764648,
+      "learning_rate": 1.5587280958570206e-05,
+      "loss": 0.8436,
+      "step": 41650
+    },
+    {
+      "epoch": 12.46,
+      "grad_norm": 5.524830341339111,
+      "learning_rate": 1.5581838346662506e-05,
+      "loss": 0.8711,
+      "step": 41655
+    },
+    {
+      "epoch": 12.46,
+      "grad_norm": 2.657658576965332,
+      "learning_rate": 1.5576396254910074e-05,
+      "loss": 0.9805,
+      "step": 41660
+    },
+    {
+      "epoch": 12.47,
+      "grad_norm": 3.9892799854278564,
+      "learning_rate": 1.5570954683613496e-05,
+      "loss": 1.0028,
+      "step": 41665
+    },
+    {
+      "epoch": 12.47,
+      "grad_norm": 3.137242555618286,
+      "learning_rate": 1.556551363307329e-05,
+      "loss": 1.0821,
+      "step": 41670
+    },
+    {
+      "epoch": 12.47,
+      "grad_norm": 3.95418119430542,
+      "learning_rate": 1.5560073103589947e-05,
+      "loss": 1.0106,
+      "step": 41675
+    },
+    {
+      "epoch": 12.47,
+      "grad_norm": 1.6314005851745605,
+      "learning_rate": 1.5554633095463966e-05,
+      "loss": 1.0254,
+      "step": 41680
+    },
+    {
+      "epoch": 12.47,
+      "grad_norm": 1.4221240282058716,
+      "learning_rate": 1.5549193608995772e-05,
+      "loss": 1.0632,
+      "step": 41685
+    },
+    {
+      "epoch": 12.47,
+      "grad_norm": 3.80000901222229,
+      "learning_rate": 1.5543754644485797e-05,
+      "loss": 1.1136,
+      "step": 41690
+    },
+    {
+      "epoch": 12.47,
+      "grad_norm": 1.8849966526031494,
+      "learning_rate": 1.5538316202234416e-05,
+      "loss": 1.1025,
+      "step": 41695
+    },
+    {
+      "epoch": 12.48,
+      "grad_norm": 1.476726770401001,
+      "learning_rate": 1.5532878282542007e-05,
+      "loss": 1.0641,
+      "step": 41700
+    },
+    {
+      "epoch": 12.48,
+      "grad_norm": 2.6277527809143066,
+      "learning_rate": 1.5527440885708884e-05,
+      "loss": 1.0265,
+      "step": 41705
+    },
+    {
+      "epoch": 12.48,
+      "grad_norm": 1.5605323314666748,
+      "learning_rate": 1.5522004012035358e-05,
+      "loss": 0.9976,
+      "step": 41710
+    },
+    {
+      "epoch": 12.48,
+      "grad_norm": 2.3000760078430176,
+      "learning_rate": 1.551656766182169e-05,
+      "loss": 0.9773,
+      "step": 41715
+    },
+    {
+      "epoch": 12.48,
+      "grad_norm": 1.9093542098999023,
+      "learning_rate": 1.551113183536814e-05,
+      "loss": 1.0161,
+      "step": 41720
+    },
+    {
+      "epoch": 12.48,
+      "grad_norm": 3.195117712020874,
+      "learning_rate": 1.5505696532974918e-05,
+      "loss": 1.1003,
+      "step": 41725
+    },
+    {
+      "epoch": 12.49,
+      "grad_norm": 1.47775399684906,
+      "learning_rate": 1.55002617549422e-05,
+      "loss": 1.0035,
+      "step": 41730
+    },
+    {
+      "epoch": 12.49,
+      "grad_norm": 4.95327615737915,
+      "learning_rate": 1.549482750157016e-05,
+      "loss": 1.1078,
+      "step": 41735
+    },
+    {
+      "epoch": 12.49,
+      "grad_norm": 1.3294042348861694,
+      "learning_rate": 1.548939377315891e-05,
+      "loss": 1.0844,
+      "step": 41740
+    },
+    {
+      "epoch": 12.49,
+      "grad_norm": 1.624887228012085,
+      "learning_rate": 1.5483960570008555e-05,
+      "loss": 1.076,
+      "step": 41745
+    },
+    {
+      "epoch": 12.49,
+      "grad_norm": 1.299483060836792,
+      "learning_rate": 1.5478527892419176e-05,
+      "loss": 1.0632,
+      "step": 41750
+    },
+    {
+      "epoch": 12.49,
+      "grad_norm": 1.710900068283081,
+      "learning_rate": 1.5473095740690792e-05,
+      "loss": 1.0242,
+      "step": 41755
+    },
+    {
+      "epoch": 12.49,
+      "grad_norm": 3.3313851356506348,
+      "learning_rate": 1.5467664115123435e-05,
+      "loss": 0.9548,
+      "step": 41760
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 2.4123222827911377,
+      "learning_rate": 1.5462233016017074e-05,
+      "loss": 1.0284,
+      "step": 41765
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 2.309690237045288,
+      "learning_rate": 1.545680244367168e-05,
+      "loss": 1.0553,
+      "step": 41770
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 1.3467509746551514,
+      "learning_rate": 1.545137239838717e-05,
+      "loss": 1.1693,
+      "step": 41775
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 1.2531408071517944,
+      "learning_rate": 1.5445942880463422e-05,
+      "loss": 1.0461,
+      "step": 41780
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 3.7026236057281494,
+      "learning_rate": 1.5440513890200333e-05,
+      "loss": 1.0725,
+      "step": 41785
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 1.7098792791366577,
+      "learning_rate": 1.543508542789771e-05,
+      "loss": 1.1343,
+      "step": 41790
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 3.3441429138183594,
+      "learning_rate": 1.542965749385539e-05,
+      "loss": 1.07,
+      "step": 41795
+    },
+    {
+      "epoch": 12.51,
+      "grad_norm": 4.822731018066406,
+      "learning_rate": 1.5424230088373132e-05,
+      "loss": 0.8686,
+      "step": 41800
+    },
+    {
+      "epoch": 12.51,
+      "grad_norm": 3.751056671142578,
+      "learning_rate": 1.541880321175069e-05,
+      "loss": 1.0426,
+      "step": 41805
+    },
+    {
+      "epoch": 12.51,
+      "grad_norm": 1.4619433879852295,
+      "learning_rate": 1.5413376864287793e-05,
+      "loss": 0.9761,
+      "step": 41810
+    },
+    {
+      "epoch": 12.51,
+      "grad_norm": 4.5311737060546875,
+      "learning_rate": 1.5407951046284118e-05,
+      "loss": 0.9352,
+      "step": 41815
+    },
+    {
+      "epoch": 12.51,
+      "grad_norm": 1.865403175354004,
+      "learning_rate": 1.5402525758039348e-05,
+      "loss": 0.9368,
+      "step": 41820
+    },
+    {
+      "epoch": 12.51,
+      "grad_norm": 1.6462074518203735,
+      "learning_rate": 1.53971009998531e-05,
+      "loss": 0.9772,
+      "step": 41825
+    },
+    {
+      "epoch": 12.52,
+      "grad_norm": 1.4321706295013428,
+      "learning_rate": 1.5391676772024983e-05,
+      "loss": 0.9295,
+      "step": 41830
+    },
+    {
+      "epoch": 12.52,
+      "grad_norm": 3.5952343940734863,
+      "learning_rate": 1.5386253074854572e-05,
+      "loss": 1.0067,
+      "step": 41835
+    },
+    {
+      "epoch": 12.52,
+      "grad_norm": 3.0504446029663086,
+      "learning_rate": 1.5380829908641407e-05,
+      "loss": 0.9688,
+      "step": 41840
+    },
+    {
+      "epoch": 12.52,
+      "grad_norm": 2.3888778686523438,
+      "learning_rate": 1.537540727368501e-05,
+      "loss": 1.1385,
+      "step": 41845
+    },
+    {
+      "epoch": 12.52,
+      "grad_norm": 3.4064996242523193,
+      "learning_rate": 1.5369985170284864e-05,
+      "loss": 1.0932,
+      "step": 41850
+    },
+    {
+      "epoch": 12.52,
+      "grad_norm": 2.349851131439209,
+      "learning_rate": 1.536456359874043e-05,
+      "loss": 1.0279,
+      "step": 41855
+    },
+    {
+      "epoch": 12.52,
+      "grad_norm": 2.4067983627319336,
+      "learning_rate": 1.5359142559351124e-05,
+      "loss": 0.8575,
+      "step": 41860
+    },
+    {
+      "epoch": 12.53,
+      "grad_norm": 7.212216854095459,
+      "learning_rate": 1.5353722052416362e-05,
+      "loss": 0.9648,
+      "step": 41865
+    },
+    {
+      "epoch": 12.53,
+      "grad_norm": 2.0969083309173584,
+      "learning_rate": 1.534830207823551e-05,
+      "loss": 1.1525,
+      "step": 41870
+    },
+    {
+      "epoch": 12.53,
+      "grad_norm": 1.6566853523254395,
+      "learning_rate": 1.5342882637107885e-05,
+      "loss": 1.058,
+      "step": 41875
+    },
+    {
+      "epoch": 12.53,
+      "grad_norm": 4.8544921875,
+      "learning_rate": 1.533746372933283e-05,
+      "loss": 0.9298,
+      "step": 41880
+    },
+    {
+      "epoch": 12.53,
+      "grad_norm": 3.7266921997070312,
+      "learning_rate": 1.5332045355209597e-05,
+      "loss": 0.9846,
+      "step": 41885
+    },
+    {
+      "epoch": 12.53,
+      "grad_norm": 3.181262254714966,
+      "learning_rate": 1.5326627515037452e-05,
+      "loss": 0.8741,
+      "step": 41890
+    },
+    {
+      "epoch": 12.53,
+      "grad_norm": 2.2438769340515137,
+      "learning_rate": 1.532121020911562e-05,
+      "loss": 0.966,
+      "step": 41895
+    },
+    {
+      "epoch": 12.54,
+      "grad_norm": 4.2861409187316895,
+      "learning_rate": 1.5315793437743284e-05,
+      "loss": 1.0051,
+      "step": 41900
+    },
+    {
+      "epoch": 12.54,
+      "grad_norm": 6.4033026695251465,
+      "learning_rate": 1.5310377201219618e-05,
+      "loss": 0.8874,
+      "step": 41905
+    },
+    {
+      "epoch": 12.54,
+      "grad_norm": 3.8851475715637207,
+      "learning_rate": 1.5304961499843734e-05,
+      "loss": 0.9808,
+      "step": 41910
+    },
+    {
+      "epoch": 12.54,
+      "grad_norm": 2.0215234756469727,
+      "learning_rate": 1.529954633391476e-05,
+      "loss": 1.1129,
+      "step": 41915
+    },
+    {
+      "epoch": 12.54,
+      "grad_norm": 5.1186699867248535,
+      "learning_rate": 1.529413170373175e-05,
+      "loss": 0.908,
+      "step": 41920
+    },
+    {
+      "epoch": 12.54,
+      "grad_norm": 2.9314262866973877,
+      "learning_rate": 1.5288717609593764e-05,
+      "loss": 1.0398,
+      "step": 41925
+    },
+    {
+      "epoch": 12.54,
+      "grad_norm": 4.335308074951172,
+      "learning_rate": 1.5283304051799813e-05,
+      "loss": 1.0499,
+      "step": 41930
+    },
+    {
+      "epoch": 12.55,
+      "grad_norm": 2.7027440071105957,
+      "learning_rate": 1.5277891030648868e-05,
+      "loss": 1.075,
+      "step": 41935
+    },
+    {
+      "epoch": 12.55,
+      "grad_norm": 1.977004051208496,
+      "learning_rate": 1.5272478546439907e-05,
+      "loss": 1.1353,
+      "step": 41940
+    },
+    {
+      "epoch": 12.55,
+      "grad_norm": 4.210170745849609,
+      "learning_rate": 1.5267066599471836e-05,
+      "loss": 1.1134,
+      "step": 41945
+    },
+    {
+      "epoch": 12.55,
+      "grad_norm": 3.0245361328125,
+      "learning_rate": 1.5261655190043568e-05,
+      "loss": 0.8765,
+      "step": 41950
+    },
+    {
+      "epoch": 12.55,
+      "grad_norm": 2.2794244289398193,
+      "learning_rate": 1.525624431845395e-05,
+      "loss": 0.9708,
+      "step": 41955
+    },
+    {
+      "epoch": 12.55,
+      "grad_norm": 4.885735034942627,
+      "learning_rate": 1.5250833985001845e-05,
+      "loss": 1.0793,
+      "step": 41960
+    },
+    {
+      "epoch": 12.56,
+      "grad_norm": 4.991076946258545,
+      "learning_rate": 1.5245424189986035e-05,
+      "loss": 1.1405,
+      "step": 41965
+    },
+    {
+      "epoch": 12.56,
+      "grad_norm": 2.991661787033081,
+      "learning_rate": 1.524001493370531e-05,
+      "loss": 0.8436,
+      "step": 41970
+    },
+    {
+      "epoch": 12.56,
+      "grad_norm": 4.58644962310791,
+      "learning_rate": 1.5234606216458414e-05,
+      "loss": 0.955,
+      "step": 41975
+    },
+    {
+      "epoch": 12.56,
+      "grad_norm": 4.122666358947754,
+      "learning_rate": 1.5229198038544068e-05,
+      "loss": 0.9454,
+      "step": 41980
+    },
+    {
+      "epoch": 12.56,
+      "grad_norm": 2.8381059169769287,
+      "learning_rate": 1.5223790400260956e-05,
+      "loss": 1.2621,
+      "step": 41985
+    },
+    {
+      "epoch": 12.56,
+      "grad_norm": 2.797647476196289,
+      "learning_rate": 1.521838330190774e-05,
+      "loss": 1.1286,
+      "step": 41990
+    },
+    {
+      "epoch": 12.56,
+      "grad_norm": 2.7422313690185547,
+      "learning_rate": 1.5212976743783047e-05,
+      "loss": 1.0241,
+      "step": 41995
+    },
+    {
+      "epoch": 12.57,
+      "grad_norm": 1.3212940692901611,
+      "learning_rate": 1.520757072618548e-05,
+      "loss": 1.0562,
+      "step": 42000
+    },
+    {
+      "epoch": 12.57,
+      "grad_norm": 1.8071925640106201,
+      "learning_rate": 1.5202165249413592e-05,
+      "loss": 1.0923,
+      "step": 42005
+    },
+    {
+      "epoch": 12.57,
+      "grad_norm": 2.4288082122802734,
+      "learning_rate": 1.5196760313765946e-05,
+      "loss": 1.1725,
+      "step": 42010
+    },
+    {
+      "epoch": 12.57,
+      "grad_norm": 4.023059368133545,
+      "learning_rate": 1.5191355919541026e-05,
+      "loss": 1.1285,
+      "step": 42015
+    },
+    {
+      "epoch": 12.57,
+      "grad_norm": 3.815493583679199,
+      "learning_rate": 1.5185952067037335e-05,
+      "loss": 0.9849,
+      "step": 42020
+    },
+    {
+      "epoch": 12.57,
+      "grad_norm": 3.8521835803985596,
+      "learning_rate": 1.5180548756553306e-05,
+      "loss": 1.0811,
+      "step": 42025
+    },
+    {
+      "epoch": 12.57,
+      "grad_norm": 1.4135304689407349,
+      "learning_rate": 1.5175145988387353e-05,
+      "loss": 1.0513,
+      "step": 42030
+    },
+    {
+      "epoch": 12.58,
+      "grad_norm": 1.7829787731170654,
+      "learning_rate": 1.5169743762837887e-05,
+      "loss": 1.1091,
+      "step": 42035
+    },
+    {
+      "epoch": 12.58,
+      "grad_norm": 1.822656512260437,
+      "learning_rate": 1.5164342080203246e-05,
+      "loss": 1.2833,
+      "step": 42040
+    },
+    {
+      "epoch": 12.58,
+      "grad_norm": 4.21824312210083,
+      "learning_rate": 1.5158940940781777e-05,
+      "loss": 0.922,
+      "step": 42045
+    },
+    {
+      "epoch": 12.58,
+      "grad_norm": 3.6083362102508545,
+      "learning_rate": 1.5153540344871757e-05,
+      "loss": 1.0563,
+      "step": 42050
+    },
+    {
+      "epoch": 12.58,
+      "grad_norm": 2.541013479232788,
+      "learning_rate": 1.5148140292771484e-05,
+      "loss": 0.8514,
+      "step": 42055
+    },
+    {
+      "epoch": 12.58,
+      "grad_norm": 2.3995792865753174,
+      "learning_rate": 1.5142740784779174e-05,
+      "loss": 1.0454,
+      "step": 42060
+    },
+    {
+      "epoch": 12.59,
+      "grad_norm": 2.6729390621185303,
+      "learning_rate": 1.5137341821193045e-05,
+      "loss": 0.9045,
+      "step": 42065
+    },
+    {
+      "epoch": 12.59,
+      "grad_norm": 3.4425437450408936,
+      "learning_rate": 1.5131943402311279e-05,
+      "loss": 1.0334,
+      "step": 42070
+    },
+    {
+      "epoch": 12.59,
+      "grad_norm": 2.506922483444214,
+      "learning_rate": 1.5126545528432017e-05,
+      "loss": 1.0794,
+      "step": 42075
+    },
+    {
+      "epoch": 12.59,
+      "grad_norm": 1.7686995267868042,
+      "learning_rate": 1.5121148199853385e-05,
+      "loss": 1.046,
+      "step": 42080
+    },
+    {
+      "epoch": 12.59,
+      "grad_norm": 1.438901424407959,
+      "learning_rate": 1.5115751416873472e-05,
+      "loss": 1.118,
+      "step": 42085
+    },
+    {
+      "epoch": 12.59,
+      "grad_norm": 1.8655344247817993,
+      "learning_rate": 1.5110355179790331e-05,
+      "loss": 0.9908,
+      "step": 42090
+    },
+    {
+      "epoch": 12.59,
+      "grad_norm": 3.9504570960998535,
+      "learning_rate": 1.5104959488902e-05,
+      "loss": 0.9724,
+      "step": 42095
+    },
+    {
+      "epoch": 12.6,
+      "grad_norm": 2.1039040088653564,
+      "learning_rate": 1.5099564344506458e-05,
+      "loss": 1.0601,
+      "step": 42100
+    },
+    {
+      "epoch": 12.6,
+      "grad_norm": 3.9319233894348145,
+      "learning_rate": 1.5094169746901698e-05,
+      "loss": 0.9059,
+      "step": 42105
+    },
+    {
+      "epoch": 12.6,
+      "grad_norm": 1.384982943534851,
+      "learning_rate": 1.5088775696385637e-05,
+      "loss": 1.0329,
+      "step": 42110
+    },
+    {
+      "epoch": 12.6,
+      "grad_norm": 3.824261426925659,
+      "learning_rate": 1.50833821932562e-05,
+      "loss": 0.9907,
+      "step": 42115
+    },
+    {
+      "epoch": 12.6,
+      "grad_norm": 2.2953989505767822,
+      "learning_rate": 1.5077989237811253e-05,
+      "loss": 0.8469,
+      "step": 42120
+    },
+    {
+      "epoch": 12.6,
+      "grad_norm": 6.605740070343018,
+      "learning_rate": 1.507259683034865e-05,
+      "loss": 1.0258,
+      "step": 42125
+    },
+    {
+      "epoch": 12.6,
+      "grad_norm": 4.182356357574463,
+      "learning_rate": 1.506720497116621e-05,
+      "loss": 0.8448,
+      "step": 42130
+    },
+    {
+      "epoch": 12.61,
+      "grad_norm": 3.7092835903167725,
+      "learning_rate": 1.5061813660561703e-05,
+      "loss": 0.906,
+      "step": 42135
+    },
+    {
+      "epoch": 12.61,
+      "grad_norm": 3.1926708221435547,
+      "learning_rate": 1.5056422898832906e-05,
+      "loss": 0.8871,
+      "step": 42140
+    },
+    {
+      "epoch": 12.61,
+      "grad_norm": 3.0576462745666504,
+      "learning_rate": 1.5051032686277527e-05,
+      "loss": 1.0512,
+      "step": 42145
+    },
+    {
+      "epoch": 12.61,
+      "grad_norm": 5.683520793914795,
+      "learning_rate": 1.5045643023193281e-05,
+      "loss": 1.0064,
+      "step": 42150
+    },
+    {
+      "epoch": 12.61,
+      "grad_norm": 1.615805983543396,
+      "learning_rate": 1.5040253909877822e-05,
+      "loss": 0.8095,
+      "step": 42155
+    },
+    {
+      "epoch": 12.61,
+      "grad_norm": 1.38075852394104,
+      "learning_rate": 1.5034865346628782e-05,
+      "loss": 1.1425,
+      "step": 42160
+    },
+    {
+      "epoch": 12.62,
+      "grad_norm": 6.462873935699463,
+      "learning_rate": 1.5029477333743774e-05,
+      "loss": 1.0516,
+      "step": 42165
+    },
+    {
+      "epoch": 12.62,
+      "grad_norm": 2.051048755645752,
+      "learning_rate": 1.5024089871520369e-05,
+      "loss": 0.8301,
+      "step": 42170
+    },
+    {
+      "epoch": 12.62,
+      "grad_norm": 2.33416485786438,
+      "learning_rate": 1.5018702960256108e-05,
+      "loss": 1.1115,
+      "step": 42175
+    },
+    {
+      "epoch": 12.62,
+      "grad_norm": 1.324583888053894,
+      "learning_rate": 1.5013316600248507e-05,
+      "loss": 1.1135,
+      "step": 42180
+    },
+    {
+      "epoch": 12.62,
+      "grad_norm": 1.6215559244155884,
+      "learning_rate": 1.5007930791795055e-05,
+      "loss": 0.9414,
+      "step": 42185
+    },
+    {
+      "epoch": 12.62,
+      "grad_norm": 1.7824453115463257,
+      "learning_rate": 1.5002545535193203e-05,
+      "loss": 1.1032,
+      "step": 42190
+    },
+    {
+      "epoch": 12.62,
+      "grad_norm": 4.399259090423584,
+      "learning_rate": 1.4997160830740358e-05,
+      "loss": 0.9014,
+      "step": 42195
+    },
+    {
+      "epoch": 12.63,
+      "grad_norm": 3.2295663356781006,
+      "learning_rate": 1.4991776678733935e-05,
+      "loss": 0.7915,
+      "step": 42200
+    },
+    {
+      "epoch": 12.63,
+      "grad_norm": 6.502789497375488,
+      "learning_rate": 1.4986393079471269e-05,
+      "loss": 0.9355,
+      "step": 42205
+    },
+    {
+      "epoch": 12.63,
+      "grad_norm": 2.078927516937256,
+      "learning_rate": 1.4981010033249718e-05,
+      "loss": 0.9986,
+      "step": 42210
+    },
+    {
+      "epoch": 12.63,
+      "grad_norm": 3.382509231567383,
+      "learning_rate": 1.4975627540366558e-05,
+      "loss": 1.0191,
+      "step": 42215
+    },
+    {
+      "epoch": 12.63,
+      "grad_norm": 2.477400302886963,
+      "learning_rate": 1.4970245601119077e-05,
+      "loss": 0.9213,
+      "step": 42220
+    },
+    {
+      "epoch": 12.63,
+      "grad_norm": 2.6311304569244385,
+      "learning_rate": 1.4964864215804508e-05,
+      "loss": 1.083,
+      "step": 42225
+    },
+    {
+      "epoch": 12.63,
+      "grad_norm": 5.062621116638184,
+      "learning_rate": 1.4959483384720047e-05,
+      "loss": 0.922,
+      "step": 42230
+    },
+    {
+      "epoch": 12.64,
+      "grad_norm": 0.9835122227668762,
+      "learning_rate": 1.4954103108162892e-05,
+      "loss": 0.9226,
+      "step": 42235
+    },
+    {
+      "epoch": 12.64,
+      "grad_norm": 1.5001591444015503,
+      "learning_rate": 1.4948723386430172e-05,
+      "loss": 0.9836,
+      "step": 42240
+    },
+    {
+      "epoch": 12.64,
+      "grad_norm": 3.599999189376831,
+      "learning_rate": 1.494334421981902e-05,
+      "loss": 0.9332,
+      "step": 42245
+    },
+    {
+      "epoch": 12.64,
+      "grad_norm": 2.7632405757904053,
+      "learning_rate": 1.4937965608626503e-05,
+      "loss": 0.869,
+      "step": 42250
+    },
+    {
+      "epoch": 12.64,
+      "grad_norm": 2.2371535301208496,
+      "learning_rate": 1.493258755314969e-05,
+      "loss": 0.9492,
+      "step": 42255
+    },
+    {
+      "epoch": 12.64,
+      "grad_norm": 3.195711851119995,
+      "learning_rate": 1.4927210053685611e-05,
+      "loss": 1.1248,
+      "step": 42260
+    },
+    {
+      "epoch": 12.65,
+      "grad_norm": 3.760953903198242,
+      "learning_rate": 1.4921833110531239e-05,
+      "loss": 1.0165,
+      "step": 42265
+    },
+    {
+      "epoch": 12.65,
+      "grad_norm": 2.693413019180298,
+      "learning_rate": 1.4916456723983558e-05,
+      "loss": 0.8719,
+      "step": 42270
+    },
+    {
+      "epoch": 12.65,
+      "grad_norm": 1.946767807006836,
+      "learning_rate": 1.4911080894339482e-05,
+      "loss": 0.9631,
+      "step": 42275
+    },
+    {
+      "epoch": 12.65,
+      "grad_norm": 3.6406733989715576,
+      "learning_rate": 1.4905705621895932e-05,
+      "loss": 0.937,
+      "step": 42280
+    },
+    {
+      "epoch": 12.65,
+      "grad_norm": 2.393702745437622,
+      "learning_rate": 1.4900330906949767e-05,
+      "loss": 1.0413,
+      "step": 42285
+    },
+    {
+      "epoch": 12.65,
+      "grad_norm": 2.1976375579833984,
+      "learning_rate": 1.4894956749797817e-05,
+      "loss": 0.9671,
+      "step": 42290
+    },
+    {
+      "epoch": 12.65,
+      "grad_norm": 1.965586543083191,
+      "learning_rate": 1.4889583150736913e-05,
+      "loss": 0.9733,
+      "step": 42295
+    },
+    {
+      "epoch": 12.66,
+      "grad_norm": 2.103189468383789,
+      "learning_rate": 1.4884210110063811e-05,
+      "loss": 1.0441,
+      "step": 42300
+    },
+    {
+      "epoch": 12.66,
+      "grad_norm": 3.943366050720215,
+      "learning_rate": 1.4878837628075281e-05,
+      "loss": 0.8933,
+      "step": 42305
+    },
+    {
+      "epoch": 12.66,
+      "grad_norm": 2.3624773025512695,
+      "learning_rate": 1.4873465705068018e-05,
+      "loss": 0.9805,
+      "step": 42310
+    },
+    {
+      "epoch": 12.66,
+      "grad_norm": 4.398625373840332,
+      "learning_rate": 1.4868094341338729e-05,
+      "loss": 1.0953,
+      "step": 42315
+    },
+    {
+      "epoch": 12.66,
+      "grad_norm": 2.5565319061279297,
+      "learning_rate": 1.4862723537184059e-05,
+      "loss": 0.8852,
+      "step": 42320
+    },
+    {
+      "epoch": 12.66,
+      "grad_norm": 3.775041341781616,
+      "learning_rate": 1.4857353292900616e-05,
+      "loss": 0.9722,
+      "step": 42325
+    },
+    {
+      "epoch": 12.66,
+      "grad_norm": 5.716676235198975,
+      "learning_rate": 1.485198360878502e-05,
+      "loss": 0.9083,
+      "step": 42330
+    },
+    {
+      "epoch": 12.67,
+      "grad_norm": 2.962364435195923,
+      "learning_rate": 1.4846614485133814e-05,
+      "loss": 1.1341,
+      "step": 42335
+    },
+    {
+      "epoch": 12.67,
+      "grad_norm": 2.1964805126190186,
+      "learning_rate": 1.4841245922243541e-05,
+      "loss": 0.9692,
+      "step": 42340
+    },
+    {
+      "epoch": 12.67,
+      "grad_norm": 1.980453372001648,
+      "learning_rate": 1.4835877920410695e-05,
+      "loss": 1.13,
+      "step": 42345
+    },
+    {
+      "epoch": 12.67,
+      "grad_norm": 16.400785446166992,
+      "learning_rate": 1.4830510479931747e-05,
+      "loss": 1.0367,
+      "step": 42350
+    },
+    {
+      "epoch": 12.67,
+      "grad_norm": 1.031053066253662,
+      "learning_rate": 1.4825143601103137e-05,
+      "loss": 0.9072,
+      "step": 42355
+    },
+    {
+      "epoch": 12.67,
+      "grad_norm": 8.164375305175781,
+      "learning_rate": 1.4819777284221264e-05,
+      "loss": 0.9322,
+      "step": 42360
+    },
+    {
+      "epoch": 12.68,
+      "grad_norm": 2.8158113956451416,
+      "learning_rate": 1.4814411529582517e-05,
+      "loss": 0.9522,
+      "step": 42365
+    },
+    {
+      "epoch": 12.68,
+      "grad_norm": 2.234344244003296,
+      "learning_rate": 1.4809046337483223e-05,
+      "loss": 0.9949,
+      "step": 42370
+    },
+    {
+      "epoch": 12.68,
+      "grad_norm": 7.7483720779418945,
+      "learning_rate": 1.4803681708219719e-05,
+      "loss": 0.9843,
+      "step": 42375
+    },
+    {
+      "epoch": 12.68,
+      "grad_norm": 2.1835453510284424,
+      "learning_rate": 1.4798317642088274e-05,
+      "loss": 0.9065,
+      "step": 42380
+    },
+    {
+      "epoch": 12.68,
+      "grad_norm": 1.4759970903396606,
+      "learning_rate": 1.4792954139385134e-05,
+      "loss": 1.0872,
+      "step": 42385
+    },
+    {
+      "epoch": 12.68,
+      "grad_norm": 2.038754940032959,
+      "learning_rate": 1.4787591200406536e-05,
+      "loss": 1.1076,
+      "step": 42390
+    },
+    {
+      "epoch": 12.68,
+      "grad_norm": 2.0084173679351807,
+      "learning_rate": 1.4782228825448653e-05,
+      "loss": 1.0353,
+      "step": 42395
+    },
+    {
+      "epoch": 12.69,
+      "grad_norm": 3.7430591583251953,
+      "learning_rate": 1.4776867014807654e-05,
+      "loss": 1.1075,
+      "step": 42400
+    },
+    {
+      "epoch": 12.69,
+      "grad_norm": 3.881999969482422,
+      "learning_rate": 1.4771505768779659e-05,
+      "loss": 0.9953,
+      "step": 42405
+    },
+    {
+      "epoch": 12.69,
+      "grad_norm": 2.164416790008545,
+      "learning_rate": 1.4766145087660779e-05,
+      "loss": 1.1684,
+      "step": 42410
+    },
+    {
+      "epoch": 12.69,
+      "grad_norm": 3.810238838195801,
+      "learning_rate": 1.476078497174706e-05,
+      "loss": 1.067,
+      "step": 42415
+    },
+    {
+      "epoch": 12.69,
+      "grad_norm": 2.1892411708831787,
+      "learning_rate": 1.4755425421334548e-05,
+      "loss": 1.0355,
+      "step": 42420
+    },
+    {
+      "epoch": 12.69,
+      "grad_norm": 1.6944376230239868,
+      "learning_rate": 1.475006643671924e-05,
+      "loss": 1.0403,
+      "step": 42425
+    },
+    {
+      "epoch": 12.69,
+      "grad_norm": 1.0041096210479736,
+      "learning_rate": 1.4744708018197107e-05,
+      "loss": 1.1216,
+      "step": 42430
+    },
+    {
+      "epoch": 12.7,
+      "grad_norm": 1.6599242687225342,
+      "learning_rate": 1.4739350166064092e-05,
+      "loss": 1.0584,
+      "step": 42435
+    },
+    {
+      "epoch": 12.7,
+      "grad_norm": 1.151971459388733,
+      "learning_rate": 1.4733992880616099e-05,
+      "loss": 1.0802,
+      "step": 42440
+    },
+    {
+      "epoch": 12.7,
+      "grad_norm": 3.420637845993042,
+      "learning_rate": 1.4728636162149011e-05,
+      "loss": 1.0013,
+      "step": 42445
+    },
+    {
+      "epoch": 12.7,
+      "grad_norm": 3.075683832168579,
+      "learning_rate": 1.4723280010958676e-05,
+      "loss": 1.0202,
+      "step": 42450
+    },
+    {
+      "epoch": 12.7,
+      "grad_norm": 4.555771350860596,
+      "learning_rate": 1.4717924427340896e-05,
+      "loss": 1.2088,
+      "step": 42455
+    },
+    {
+      "epoch": 12.7,
+      "grad_norm": 1.3353490829467773,
+      "learning_rate": 1.4712569411591468e-05,
+      "loss": 1.031,
+      "step": 42460
+    },
+    {
+      "epoch": 12.71,
+      "grad_norm": 2.8824167251586914,
+      "learning_rate": 1.4707214964006131e-05,
+      "loss": 1.1876,
+      "step": 42465
+    },
+    {
+      "epoch": 12.71,
+      "grad_norm": 4.97543478012085,
+      "learning_rate": 1.470186108488062e-05,
+      "loss": 1.1018,
+      "step": 42470
+    },
+    {
+      "epoch": 12.71,
+      "grad_norm": 1.2788058519363403,
+      "learning_rate": 1.469650777451061e-05,
+      "loss": 1.0444,
+      "step": 42475
+    },
+    {
+      "epoch": 12.71,
+      "grad_norm": 4.054439067840576,
+      "learning_rate": 1.4691155033191773e-05,
+      "loss": 0.7675,
+      "step": 42480
+    },
+    {
+      "epoch": 12.71,
+      "grad_norm": 4.729781627655029,
+      "learning_rate": 1.4685802861219732e-05,
+      "loss": 1.1607,
+      "step": 42485
+    },
+    {
+      "epoch": 12.71,
+      "grad_norm": 3.289395570755005,
+      "learning_rate": 1.4680451258890066e-05,
+      "loss": 1.0532,
+      "step": 42490
+    },
+    {
+      "epoch": 12.71,
+      "grad_norm": 3.1528773307800293,
+      "learning_rate": 1.467510022649836e-05,
+      "loss": 1.1017,
+      "step": 42495
+    },
+    {
+      "epoch": 12.72,
+      "grad_norm": 1.4691784381866455,
+      "learning_rate": 1.4669749764340124e-05,
+      "loss": 0.9924,
+      "step": 42500
+    },
+    {
+      "epoch": 12.72,
+      "grad_norm": 4.324456691741943,
+      "learning_rate": 1.4664399872710885e-05,
+      "loss": 1.0135,
+      "step": 42505
+    },
+    {
+      "epoch": 12.72,
+      "grad_norm": 4.2510294914245605,
+      "learning_rate": 1.4659050551906089e-05,
+      "loss": 1.0921,
+      "step": 42510
+    },
+    {
+      "epoch": 12.72,
+      "grad_norm": 10.014864921569824,
+      "learning_rate": 1.4653701802221187e-05,
+      "loss": 0.9695,
+      "step": 42515
+    },
+    {
+      "epoch": 12.72,
+      "grad_norm": 6.1459455490112305,
+      "learning_rate": 1.4648353623951577e-05,
+      "loss": 0.9502,
+      "step": 42520
+    },
+    {
+      "epoch": 12.72,
+      "grad_norm": 2.0565860271453857,
+      "learning_rate": 1.4643006017392636e-05,
+      "loss": 0.827,
+      "step": 42525
+    },
+    {
+      "epoch": 12.72,
+      "grad_norm": 2.4255897998809814,
+      "learning_rate": 1.4637658982839709e-05,
+      "loss": 0.9423,
+      "step": 42530
+    },
+    {
+      "epoch": 12.73,
+      "grad_norm": 2.0831713676452637,
+      "learning_rate": 1.4632312520588104e-05,
+      "loss": 0.8922,
+      "step": 42535
+    },
+    {
+      "epoch": 12.73,
+      "grad_norm": 2.779167890548706,
+      "learning_rate": 1.46269666309331e-05,
+      "loss": 0.9137,
+      "step": 42540
+    },
+    {
+      "epoch": 12.73,
+      "grad_norm": 2.2856433391571045,
+      "learning_rate": 1.4621621314169953e-05,
+      "loss": 0.7789,
+      "step": 42545
+    },
+    {
+      "epoch": 12.73,
+      "grad_norm": 1.7285444736480713,
+      "learning_rate": 1.4616276570593865e-05,
+      "loss": 1.0144,
+      "step": 42550
+    },
+    {
+      "epoch": 12.73,
+      "grad_norm": 4.265864372253418,
+      "learning_rate": 1.4610932400500037e-05,
+      "loss": 1.0229,
+      "step": 42555
+    },
+    {
+      "epoch": 12.73,
+      "grad_norm": 3.695251941680908,
+      "learning_rate": 1.4605588804183607e-05,
+      "loss": 0.9748,
+      "step": 42560
+    },
+    {
+      "epoch": 12.73,
+      "grad_norm": 6.5001749992370605,
+      "learning_rate": 1.4600245781939708e-05,
+      "loss": 1.1294,
+      "step": 42565
+    },
+    {
+      "epoch": 12.74,
+      "grad_norm": 2.0487442016601562,
+      "learning_rate": 1.4594903334063423e-05,
+      "loss": 1.0273,
+      "step": 42570
+    },
+    {
+      "epoch": 12.74,
+      "grad_norm": 2.534780502319336,
+      "learning_rate": 1.458956146084981e-05,
+      "loss": 1.0041,
+      "step": 42575
+    },
+    {
+      "epoch": 12.74,
+      "grad_norm": 4.824891567230225,
+      "learning_rate": 1.4584220162593898e-05,
+      "loss": 1.0993,
+      "step": 42580
+    },
+    {
+      "epoch": 12.74,
+      "grad_norm": 1.7529278993606567,
+      "learning_rate": 1.4578879439590675e-05,
+      "loss": 1.1054,
+      "step": 42585
+    },
+    {
+      "epoch": 12.74,
+      "grad_norm": 1.254698634147644,
+      "learning_rate": 1.4573539292135113e-05,
+      "loss": 0.8739,
+      "step": 42590
+    },
+    {
+      "epoch": 12.74,
+      "grad_norm": 4.246768951416016,
+      "learning_rate": 1.4568199720522135e-05,
+      "loss": 1.1456,
+      "step": 42595
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 4.16953182220459,
+      "learning_rate": 1.4562860725046648e-05,
+      "loss": 0.839,
+      "step": 42600
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 1.6719205379486084,
+      "learning_rate": 1.4557522306003521e-05,
+      "loss": 0.9557,
+      "step": 42605
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 2.465372085571289,
+      "learning_rate": 1.4552184463687562e-05,
+      "loss": 1.0845,
+      "step": 42610
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 3.8194258213043213,
+      "learning_rate": 1.4546847198393618e-05,
+      "loss": 1.0781,
+      "step": 42615
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 1.8574891090393066,
+      "learning_rate": 1.4541510510416418e-05,
+      "loss": 0.9392,
+      "step": 42620
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 2.715480089187622,
+      "learning_rate": 1.4536174400050739e-05,
+      "loss": 1.1462,
+      "step": 42625
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 1.606964349746704,
+      "learning_rate": 1.453083886759125e-05,
+      "loss": 0.9121,
+      "step": 42630
+    },
+    {
+      "epoch": 12.76,
+      "grad_norm": 3.806936740875244,
+      "learning_rate": 1.4525503913332672e-05,
+      "loss": 0.973,
+      "step": 42635
+    },
+    {
+      "epoch": 12.76,
+      "grad_norm": 3.890246629714966,
+      "learning_rate": 1.4520169537569615e-05,
+      "loss": 1.1849,
+      "step": 42640
+    },
+    {
+      "epoch": 12.76,
+      "grad_norm": 3.9114744663238525,
+      "learning_rate": 1.4514835740596699e-05,
+      "loss": 1.0194,
+      "step": 42645
+    },
+    {
+      "epoch": 12.76,
+      "grad_norm": 1.8596765995025635,
+      "learning_rate": 1.4509502522708507e-05,
+      "loss": 1.057,
+      "step": 42650
+    },
+    {
+      "epoch": 12.76,
+      "grad_norm": 3.2842726707458496,
+      "learning_rate": 1.4504169884199587e-05,
+      "loss": 0.9596,
+      "step": 42655
+    },
+    {
+      "epoch": 12.76,
+      "grad_norm": 2.9906461238861084,
+      "learning_rate": 1.4498837825364452e-05,
+      "loss": 0.8944,
+      "step": 42660
+    },
+    {
+      "epoch": 12.76,
+      "grad_norm": 1.7279856204986572,
+      "learning_rate": 1.4493506346497593e-05,
+      "loss": 1.0216,
+      "step": 42665
+    },
+    {
+      "epoch": 12.77,
+      "grad_norm": 1.639946460723877,
+      "learning_rate": 1.4488175447893455e-05,
+      "loss": 1.097,
+      "step": 42670
+    },
+    {
+      "epoch": 12.77,
+      "grad_norm": 9.285032272338867,
+      "learning_rate": 1.4482845129846459e-05,
+      "loss": 1.0578,
+      "step": 42675
+    },
+    {
+      "epoch": 12.77,
+      "grad_norm": 1.8126649856567383,
+      "learning_rate": 1.4477515392650997e-05,
+      "loss": 0.999,
+      "step": 42680
+    },
+    {
+      "epoch": 12.77,
+      "grad_norm": 2.5790581703186035,
+      "learning_rate": 1.447218623660142e-05,
+      "loss": 0.9432,
+      "step": 42685
+    },
+    {
+      "epoch": 12.77,
+      "grad_norm": 1.561050295829773,
+      "learning_rate": 1.4466857661992055e-05,
+      "loss": 0.9825,
+      "step": 42690
+    },
+    {
+      "epoch": 12.77,
+      "grad_norm": 3.258485794067383,
+      "learning_rate": 1.4461529669117197e-05,
+      "loss": 1.1113,
+      "step": 42695
+    },
+    {
+      "epoch": 12.78,
+      "grad_norm": 1.9323861598968506,
+      "learning_rate": 1.4456202258271096e-05,
+      "loss": 1.1198,
+      "step": 42700
+    },
+    {
+      "epoch": 12.78,
+      "grad_norm": 3.977754831314087,
+      "learning_rate": 1.4450875429747987e-05,
+      "loss": 1.0889,
+      "step": 42705
+    },
+    {
+      "epoch": 12.78,
+      "grad_norm": 2.0819084644317627,
+      "learning_rate": 1.4445549183842072e-05,
+      "loss": 1.2944,
+      "step": 42710
+    },
+    {
+      "epoch": 12.78,
+      "grad_norm": 2.5505735874176025,
+      "learning_rate": 1.4440223520847485e-05,
+      "loss": 0.9569,
+      "step": 42715
+    },
+    {
+      "epoch": 12.78,
+      "grad_norm": 1.7122985124588013,
+      "learning_rate": 1.4434898441058398e-05,
+      "loss": 1.0941,
+      "step": 42720
+    },
+    {
+      "epoch": 12.78,
+      "grad_norm": 2.2943074703216553,
+      "learning_rate": 1.4429573944768865e-05,
+      "loss": 1.0611,
+      "step": 42725
+    },
+    {
+      "epoch": 12.78,
+      "grad_norm": 5.206298828125,
+      "learning_rate": 1.4424250032272999e-05,
+      "loss": 1.1535,
+      "step": 42730
+    },
+    {
+      "epoch": 12.79,
+      "grad_norm": 1.9564958810806274,
+      "learning_rate": 1.4418926703864799e-05,
+      "loss": 1.0421,
+      "step": 42735
+    },
+    {
+      "epoch": 12.79,
+      "grad_norm": 1.8103880882263184,
+      "learning_rate": 1.4413603959838278e-05,
+      "loss": 0.8809,
+      "step": 42740
+    },
+    {
+      "epoch": 12.79,
+      "grad_norm": 2.9786670207977295,
+      "learning_rate": 1.4408281800487411e-05,
+      "loss": 1.0065,
+      "step": 42745
+    },
+    {
+      "epoch": 12.79,
+      "grad_norm": 5.209625720977783,
+      "learning_rate": 1.4402960226106128e-05,
+      "loss": 1.0411,
+      "step": 42750
+    },
+    {
+      "epoch": 12.79,
+      "grad_norm": 2.3398683071136475,
+      "learning_rate": 1.4397639236988334e-05,
+      "loss": 1.0483,
+      "step": 42755
+    },
+    {
+      "epoch": 12.79,
+      "grad_norm": 1.2769595384597778,
+      "learning_rate": 1.4392318833427896e-05,
+      "loss": 1.0679,
+      "step": 42760
+    },
+    {
+      "epoch": 12.79,
+      "grad_norm": 8.292903900146484,
+      "learning_rate": 1.4386999015718684e-05,
+      "loss": 1.0761,
+      "step": 42765
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 2.5282647609710693,
+      "learning_rate": 1.4381679784154472e-05,
+      "loss": 1.041,
+      "step": 42770
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 2.1609020233154297,
+      "learning_rate": 1.4376361139029052e-05,
+      "loss": 1.123,
+      "step": 42775
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 1.4257582426071167,
+      "learning_rate": 1.4371043080636159e-05,
+      "loss": 1.1743,
+      "step": 42780
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 3.5058562755584717,
+      "learning_rate": 1.4365725609269515e-05,
+      "loss": 1.0082,
+      "step": 42785
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 2.25451397895813,
+      "learning_rate": 1.4360408725222785e-05,
+      "loss": 0.9842,
+      "step": 42790
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 2.5415477752685547,
+      "learning_rate": 1.4355092428789627e-05,
+      "loss": 0.9281,
+      "step": 42795
+    },
+    {
+      "epoch": 12.81,
+      "grad_norm": 2.6299939155578613,
+      "learning_rate": 1.4349776720263647e-05,
+      "loss": 0.9657,
+      "step": 42800
+    },
+    {
+      "epoch": 12.81,
+      "grad_norm": 3.0297467708587646,
+      "learning_rate": 1.4344461599938439e-05,
+      "loss": 0.9558,
+      "step": 42805
+    },
+    {
+      "epoch": 12.81,
+      "grad_norm": 1.9421048164367676,
+      "learning_rate": 1.4339147068107522e-05,
+      "loss": 1.0252,
+      "step": 42810
+    },
+    {
+      "epoch": 12.81,
+      "grad_norm": 3.981034517288208,
+      "learning_rate": 1.433383312506445e-05,
+      "loss": 1.0214,
+      "step": 42815
+    },
+    {
+      "epoch": 12.81,
+      "grad_norm": 5.4559736251831055,
+      "learning_rate": 1.4328519771102666e-05,
+      "loss": 0.9897,
+      "step": 42820
+    },
+    {
+      "epoch": 12.81,
+      "grad_norm": 3.1577250957489014,
+      "learning_rate": 1.4323207006515665e-05,
+      "loss": 0.8926,
+      "step": 42825
+    },
+    {
+      "epoch": 12.81,
+      "grad_norm": 11.30273723602295,
+      "learning_rate": 1.4317894831596823e-05,
+      "loss": 0.9951,
+      "step": 42830
+    },
+    {
+      "epoch": 12.82,
+      "grad_norm": 1.4776180982589722,
+      "learning_rate": 1.4312583246639566e-05,
+      "loss": 1.1993,
+      "step": 42835
+    },
+    {
+      "epoch": 12.82,
+      "grad_norm": 1.8206647634506226,
+      "learning_rate": 1.4307272251937215e-05,
+      "loss": 1.1282,
+      "step": 42840
+    },
+    {
+      "epoch": 12.82,
+      "grad_norm": 4.272174835205078,
+      "learning_rate": 1.4301961847783096e-05,
+      "loss": 1.0822,
+      "step": 42845
+    },
+    {
+      "epoch": 12.82,
+      "grad_norm": 1.5834592580795288,
+      "learning_rate": 1.4296652034470523e-05,
+      "loss": 1.0474,
+      "step": 42850
+    },
+    {
+      "epoch": 12.82,
+      "grad_norm": 3.7897226810455322,
+      "learning_rate": 1.4291342812292712e-05,
+      "loss": 1.0012,
+      "step": 42855
+    },
+    {
+      "epoch": 12.82,
+      "grad_norm": 1.9825385808944702,
+      "learning_rate": 1.4286034181542928e-05,
+      "loss": 0.9521,
+      "step": 42860
+    },
+    {
+      "epoch": 12.82,
+      "grad_norm": 2.32985782623291,
+      "learning_rate": 1.4280726142514328e-05,
+      "loss": 1.189,
+      "step": 42865
+    },
+    {
+      "epoch": 12.83,
+      "grad_norm": 2.0926525592803955,
+      "learning_rate": 1.4275418695500081e-05,
+      "loss": 0.8125,
+      "step": 42870
+    },
+    {
+      "epoch": 12.83,
+      "grad_norm": 3.3241047859191895,
+      "learning_rate": 1.4270111840793316e-05,
+      "loss": 0.8291,
+      "step": 42875
+    },
+    {
+      "epoch": 12.83,
+      "grad_norm": 2.867534637451172,
+      "learning_rate": 1.4264805578687116e-05,
+      "loss": 0.9589,
+      "step": 42880
+    },
+    {
+      "epoch": 12.83,
+      "grad_norm": 1.9998924732208252,
+      "learning_rate": 1.4259499909474548e-05,
+      "loss": 1.176,
+      "step": 42885
+    },
+    {
+      "epoch": 12.83,
+      "grad_norm": 3.46891188621521,
+      "learning_rate": 1.4254194833448634e-05,
+      "loss": 1.0184,
+      "step": 42890
+    },
+    {
+      "epoch": 12.83,
+      "grad_norm": 2.473839044570923,
+      "learning_rate": 1.4248890350902371e-05,
+      "loss": 1.104,
+      "step": 42895
+    },
+    {
+      "epoch": 12.84,
+      "grad_norm": 1.8417174816131592,
+      "learning_rate": 1.4243586462128727e-05,
+      "loss": 1.0769,
+      "step": 42900
+    },
+    {
+      "epoch": 12.84,
+      "grad_norm": 3.319664716720581,
+      "learning_rate": 1.4238283167420601e-05,
+      "loss": 0.9912,
+      "step": 42905
+    },
+    {
+      "epoch": 12.84,
+      "grad_norm": 1.5683692693710327,
+      "learning_rate": 1.4232980467070933e-05,
+      "loss": 0.8247,
+      "step": 42910
+    },
+    {
+      "epoch": 12.84,
+      "grad_norm": 1.6574991941452026,
+      "learning_rate": 1.4227678361372544e-05,
+      "loss": 1.074,
+      "step": 42915
+    },
+    {
+      "epoch": 12.84,
+      "grad_norm": 4.681344032287598,
+      "learning_rate": 1.4222376850618286e-05,
+      "loss": 1.0578,
+      "step": 42920
+    },
+    {
+      "epoch": 12.84,
+      "grad_norm": 1.7636343240737915,
+      "learning_rate": 1.4217075935100955e-05,
+      "loss": 1.0217,
+      "step": 42925
+    },
+    {
+      "epoch": 12.84,
+      "grad_norm": 3.016009569168091,
+      "learning_rate": 1.4211775615113312e-05,
+      "loss": 1.0764,
+      "step": 42930
+    },
+    {
+      "epoch": 12.85,
+      "grad_norm": 2.3196792602539062,
+      "learning_rate": 1.4206475890948099e-05,
+      "loss": 0.9541,
+      "step": 42935
+    },
+    {
+      "epoch": 12.85,
+      "grad_norm": 1.9939143657684326,
+      "learning_rate": 1.4201176762897981e-05,
+      "loss": 1.1661,
+      "step": 42940
+    },
+    {
+      "epoch": 12.85,
+      "grad_norm": 2.861206531524658,
+      "learning_rate": 1.4195878231255665e-05,
+      "loss": 0.9937,
+      "step": 42945
+    },
+    {
+      "epoch": 12.85,
+      "grad_norm": 3.518277883529663,
+      "learning_rate": 1.4190580296313746e-05,
+      "loss": 0.9096,
+      "step": 42950
+    },
+    {
+      "epoch": 12.85,
+      "grad_norm": 5.832763671875,
+      "learning_rate": 1.4185282958364863e-05,
+      "loss": 0.982,
+      "step": 42955
+    },
+    {
+      "epoch": 12.85,
+      "grad_norm": 1.4618480205535889,
+      "learning_rate": 1.4179986217701547e-05,
+      "loss": 1.1446,
+      "step": 42960
+    },
+    {
+      "epoch": 12.85,
+      "grad_norm": 2.2054197788238525,
+      "learning_rate": 1.4174690074616348e-05,
+      "loss": 1.0362,
+      "step": 42965
+    },
+    {
+      "epoch": 12.86,
+      "grad_norm": 2.494736909866333,
+      "learning_rate": 1.4169394529401764e-05,
+      "loss": 1.151,
+      "step": 42970
+    },
+    {
+      "epoch": 12.86,
+      "grad_norm": 2.238530397415161,
+      "learning_rate": 1.416409958235026e-05,
+      "loss": 0.9931,
+      "step": 42975
+    },
+    {
+      "epoch": 12.86,
+      "grad_norm": 2.8167457580566406,
+      "learning_rate": 1.4158805233754274e-05,
+      "loss": 0.9729,
+      "step": 42980
+    },
+    {
+      "epoch": 12.86,
+      "grad_norm": 2.740593433380127,
+      "learning_rate": 1.4153511483906207e-05,
+      "loss": 0.894,
+      "step": 42985
+    },
+    {
+      "epoch": 12.86,
+      "grad_norm": 2.3016016483306885,
+      "learning_rate": 1.414821833309843e-05,
+      "loss": 1.0095,
+      "step": 42990
+    },
+    {
+      "epoch": 12.86,
+      "grad_norm": 4.366925239562988,
+      "learning_rate": 1.4142925781623279e-05,
+      "loss": 1.0067,
+      "step": 42995
+    },
+    {
+      "epoch": 12.87,
+      "grad_norm": 2.0582058429718018,
+      "learning_rate": 1.4137633829773032e-05,
+      "loss": 1.0384,
+      "step": 43000
+    },
+    {
+      "epoch": 12.87,
+      "grad_norm": 2.0724596977233887,
+      "learning_rate": 1.4132342477839988e-05,
+      "loss": 1.0962,
+      "step": 43005
+    },
+    {
+      "epoch": 12.87,
+      "grad_norm": 4.274909496307373,
+      "learning_rate": 1.4127051726116375e-05,
+      "loss": 0.9295,
+      "step": 43010
+    },
+    {
+      "epoch": 12.87,
+      "grad_norm": 4.048436164855957,
+      "learning_rate": 1.4121761574894393e-05,
+      "loss": 0.7961,
+      "step": 43015
+    },
+    {
+      "epoch": 12.87,
+      "grad_norm": 2.6834957599639893,
+      "learning_rate": 1.4116472024466209e-05,
+      "loss": 1.1797,
+      "step": 43020
+    },
+    {
+      "epoch": 12.87,
+      "grad_norm": 1.8873329162597656,
+      "learning_rate": 1.4111183075123965e-05,
+      "loss": 0.8737,
+      "step": 43025
+    },
+    {
+      "epoch": 12.87,
+      "grad_norm": 3.20086407661438,
+      "learning_rate": 1.410589472715977e-05,
+      "loss": 0.9586,
+      "step": 43030
+    },
+    {
+      "epoch": 12.88,
+      "grad_norm": 2.417081117630005,
+      "learning_rate": 1.4100606980865666e-05,
+      "loss": 1.0065,
+      "step": 43035
+    },
+    {
+      "epoch": 12.88,
+      "grad_norm": 7.702847003936768,
+      "learning_rate": 1.4095319836533732e-05,
+      "loss": 1.0683,
+      "step": 43040
+    },
+    {
+      "epoch": 12.88,
+      "grad_norm": 2.575244665145874,
+      "learning_rate": 1.4090033294455923e-05,
+      "loss": 0.743,
+      "step": 43045
+    },
+    {
+      "epoch": 12.88,
+      "grad_norm": 2.828678607940674,
+      "learning_rate": 1.408474735492426e-05,
+      "loss": 1.0292,
+      "step": 43050
+    },
+    {
+      "epoch": 12.88,
+      "grad_norm": 2.4835140705108643,
+      "learning_rate": 1.4079462018230633e-05,
+      "loss": 1.0267,
+      "step": 43055
+    },
+    {
+      "epoch": 12.88,
+      "grad_norm": 3.8808891773223877,
+      "learning_rate": 1.4074177284666986e-05,
+      "loss": 0.7873,
+      "step": 43060
+    },
+    {
+      "epoch": 12.88,
+      "grad_norm": 2.5032103061676025,
+      "learning_rate": 1.4068893154525165e-05,
+      "loss": 1.1472,
+      "step": 43065
+    },
+    {
+      "epoch": 12.89,
+      "grad_norm": 7.541705131530762,
+      "learning_rate": 1.4063609628097007e-05,
+      "loss": 1.0327,
+      "step": 43070
+    },
+    {
+      "epoch": 12.89,
+      "grad_norm": 1.81223726272583,
+      "learning_rate": 1.4058326705674324e-05,
+      "loss": 1.0865,
+      "step": 43075
+    },
+    {
+      "epoch": 12.89,
+      "grad_norm": 1.7020493745803833,
+      "learning_rate": 1.4053044387548886e-05,
+      "loss": 1.1599,
+      "step": 43080
+    },
+    {
+      "epoch": 12.89,
+      "grad_norm": 1.811834454536438,
+      "learning_rate": 1.4047762674012426e-05,
+      "loss": 1.1018,
+      "step": 43085
+    },
+    {
+      "epoch": 12.89,
+      "grad_norm": 2.5735673904418945,
+      "learning_rate": 1.404248156535665e-05,
+      "loss": 1.164,
+      "step": 43090
+    },
+    {
+      "epoch": 12.89,
+      "grad_norm": 5.103279113769531,
+      "learning_rate": 1.4037201061873225e-05,
+      "loss": 1.0598,
+      "step": 43095
+    },
+    {
+      "epoch": 12.9,
+      "grad_norm": 2.6463088989257812,
+      "learning_rate": 1.4031921163853791e-05,
+      "loss": 1.1851,
+      "step": 43100
+    },
+    {
+      "epoch": 12.9,
+      "grad_norm": 3.571685314178467,
+      "learning_rate": 1.4026641871589951e-05,
+      "loss": 1.1104,
+      "step": 43105
+    },
+    {
+      "epoch": 12.9,
+      "grad_norm": 2.5913071632385254,
+      "learning_rate": 1.4021363185373277e-05,
+      "loss": 1.11,
+      "step": 43110
+    },
+    {
+      "epoch": 12.9,
+      "grad_norm": 1.9859071969985962,
+      "learning_rate": 1.4016085105495299e-05,
+      "loss": 1.0636,
+      "step": 43115
+    },
+    {
+      "epoch": 12.9,
+      "grad_norm": 1.736292839050293,
+      "learning_rate": 1.4010807632247525e-05,
+      "loss": 1.1104,
+      "step": 43120
+    },
+    {
+      "epoch": 12.9,
+      "grad_norm": 4.8815836906433105,
+      "learning_rate": 1.4005530765921432e-05,
+      "loss": 0.8693,
+      "step": 43125
+    },
+    {
+      "epoch": 12.9,
+      "grad_norm": 3.071505069732666,
+      "learning_rate": 1.4000254506808428e-05,
+      "loss": 1.0174,
+      "step": 43130
+    },
+    {
+      "epoch": 12.91,
+      "grad_norm": 6.106081485748291,
+      "learning_rate": 1.3994978855199953e-05,
+      "loss": 1.0536,
+      "step": 43135
+    },
+    {
+      "epoch": 12.91,
+      "grad_norm": 3.1881868839263916,
+      "learning_rate": 1.3989703811387337e-05,
+      "loss": 1.0537,
+      "step": 43140
+    },
+    {
+      "epoch": 12.91,
+      "grad_norm": 9.79444694519043,
+      "learning_rate": 1.3984429375661955e-05,
+      "loss": 0.8779,
+      "step": 43145
+    },
+    {
+      "epoch": 12.91,
+      "grad_norm": 3.3935022354125977,
+      "learning_rate": 1.3979155548315067e-05,
+      "loss": 0.9611,
+      "step": 43150
+    },
+    {
+      "epoch": 12.91,
+      "grad_norm": 2.132736921310425,
+      "learning_rate": 1.3973882329637983e-05,
+      "loss": 0.9938,
+      "step": 43155
+    },
+    {
+      "epoch": 12.91,
+      "grad_norm": 3.71537446975708,
+      "learning_rate": 1.396860971992191e-05,
+      "loss": 0.8078,
+      "step": 43160
+    },
+    {
+      "epoch": 12.91,
+      "grad_norm": 1.3954871892929077,
+      "learning_rate": 1.3963337719458052e-05,
+      "loss": 0.8549,
+      "step": 43165
+    },
+    {
+      "epoch": 12.92,
+      "grad_norm": 4.286108016967773,
+      "learning_rate": 1.3958066328537583e-05,
+      "loss": 1.1669,
+      "step": 43170
+    },
+    {
+      "epoch": 12.92,
+      "grad_norm": 1.1374285221099854,
+      "learning_rate": 1.395279554745163e-05,
+      "loss": 1.1838,
+      "step": 43175
+    },
+    {
+      "epoch": 12.92,
+      "grad_norm": 4.613668441772461,
+      "learning_rate": 1.39475253764913e-05,
+      "loss": 0.9838,
+      "step": 43180
+    },
+    {
+      "epoch": 12.92,
+      "grad_norm": 4.321890830993652,
+      "learning_rate": 1.3942255815947652e-05,
+      "loss": 0.972,
+      "step": 43185
+    },
+    {
+      "epoch": 12.92,
+      "grad_norm": 2.9430012702941895,
+      "learning_rate": 1.393698686611172e-05,
+      "loss": 0.9042,
+      "step": 43190
+    },
+    {
+      "epoch": 12.92,
+      "grad_norm": 3.068352222442627,
+      "learning_rate": 1.3931718527274506e-05,
+      "loss": 1.0785,
+      "step": 43195
+    },
+    {
+      "epoch": 12.92,
+      "grad_norm": 1.4933991432189941,
+      "learning_rate": 1.3926450799726975e-05,
+      "loss": 0.9176,
+      "step": 43200
+    },
+    {
+      "epoch": 12.93,
+      "grad_norm": 3.601362943649292,
+      "learning_rate": 1.3921183683760054e-05,
+      "loss": 1.0315,
+      "step": 43205
+    },
+    {
+      "epoch": 12.93,
+      "grad_norm": 4.306023120880127,
+      "learning_rate": 1.391591717966464e-05,
+      "loss": 1.1656,
+      "step": 43210
+    },
+    {
+      "epoch": 12.93,
+      "grad_norm": 1.5193262100219727,
+      "learning_rate": 1.3910651287731601e-05,
+      "loss": 1.0546,
+      "step": 43215
+    },
+    {
+      "epoch": 12.93,
+      "grad_norm": 1.7641685009002686,
+      "learning_rate": 1.3905386008251775e-05,
+      "loss": 1.1402,
+      "step": 43220
+    },
+    {
+      "epoch": 12.93,
+      "grad_norm": 3.4775922298431396,
+      "learning_rate": 1.3900121341515929e-05,
+      "loss": 1.0022,
+      "step": 43225
+    },
+    {
+      "epoch": 12.93,
+      "grad_norm": 2.7105138301849365,
+      "learning_rate": 1.389485728781486e-05,
+      "loss": 1.0017,
+      "step": 43230
+    },
+    {
+      "epoch": 12.94,
+      "grad_norm": 4.041553497314453,
+      "learning_rate": 1.3889593847439265e-05,
+      "loss": 0.9267,
+      "step": 43235
+    },
+    {
+      "epoch": 12.94,
+      "grad_norm": 3.279675006866455,
+      "learning_rate": 1.3884331020679869e-05,
+      "loss": 0.9009,
+      "step": 43240
+    },
+    {
+      "epoch": 12.94,
+      "grad_norm": 1.1497808694839478,
+      "learning_rate": 1.3879068807827295e-05,
+      "loss": 1.0816,
+      "step": 43245
+    },
+    {
+      "epoch": 12.94,
+      "grad_norm": 3.0773322582244873,
+      "learning_rate": 1.3873807209172215e-05,
+      "loss": 1.0265,
+      "step": 43250
+    },
+    {
+      "epoch": 12.94,
+      "grad_norm": 3.1439836025238037,
+      "learning_rate": 1.3868546225005185e-05,
+      "loss": 1.0566,
+      "step": 43255
+    },
+    {
+      "epoch": 12.94,
+      "grad_norm": 4.710149765014648,
+      "learning_rate": 1.3863285855616781e-05,
+      "loss": 1.1018,
+      "step": 43260
+    },
+    {
+      "epoch": 12.94,
+      "grad_norm": 4.115421295166016,
+      "learning_rate": 1.385802610129752e-05,
+      "loss": 1.0303,
+      "step": 43265
+    },
+    {
+      "epoch": 12.95,
+      "grad_norm": 1.8717732429504395,
+      "learning_rate": 1.3852766962337897e-05,
+      "loss": 1.152,
+      "step": 43270
+    },
+    {
+      "epoch": 12.95,
+      "grad_norm": 1.6938856840133667,
+      "learning_rate": 1.3847508439028367e-05,
+      "loss": 1.1106,
+      "step": 43275
+    },
+    {
+      "epoch": 12.95,
+      "grad_norm": 2.244997024536133,
+      "learning_rate": 1.3842250531659351e-05,
+      "loss": 1.0663,
+      "step": 43280
+    },
+    {
+      "epoch": 12.95,
+      "grad_norm": 1.7450692653656006,
+      "learning_rate": 1.3836993240521245e-05,
+      "loss": 0.9527,
+      "step": 43285
+    },
+    {
+      "epoch": 12.95,
+      "grad_norm": 2.384140968322754,
+      "learning_rate": 1.3831736565904396e-05,
+      "loss": 0.8981,
+      "step": 43290
+    },
+    {
+      "epoch": 12.95,
+      "grad_norm": 3.3110265731811523,
+      "learning_rate": 1.382648050809913e-05,
+      "loss": 0.9716,
+      "step": 43295
+    },
+    {
+      "epoch": 12.95,
+      "grad_norm": 1.9519492387771606,
+      "learning_rate": 1.3821225067395729e-05,
+      "loss": 1.0456,
+      "step": 43300
+    },
+    {
+      "epoch": 12.96,
+      "grad_norm": 3.2571592330932617,
+      "learning_rate": 1.3815970244084447e-05,
+      "loss": 1.3004,
+      "step": 43305
+    },
+    {
+      "epoch": 12.96,
+      "grad_norm": 3.3342976570129395,
+      "learning_rate": 1.3810716038455506e-05,
+      "loss": 0.9739,
+      "step": 43310
+    },
+    {
+      "epoch": 12.96,
+      "grad_norm": 2.1756532192230225,
+      "learning_rate": 1.3805462450799098e-05,
+      "loss": 0.9842,
+      "step": 43315
+    },
+    {
+      "epoch": 12.96,
+      "grad_norm": 3.6889872550964355,
+      "learning_rate": 1.3800209481405341e-05,
+      "loss": 1.0174,
+      "step": 43320
+    },
+    {
+      "epoch": 12.96,
+      "grad_norm": 2.0239474773406982,
+      "learning_rate": 1.3794957130564393e-05,
+      "loss": 1.0647,
+      "step": 43325
+    },
+    {
+      "epoch": 12.96,
+      "grad_norm": 7.694988250732422,
+      "learning_rate": 1.3789705398566296e-05,
+      "loss": 0.9891,
+      "step": 43330
+    },
+    {
+      "epoch": 12.97,
+      "grad_norm": 1.3746068477630615,
+      "learning_rate": 1.3784454285701137e-05,
+      "loss": 1.0001,
+      "step": 43335
+    },
+    {
+      "epoch": 12.97,
+      "grad_norm": 3.215383291244507,
+      "learning_rate": 1.3779203792258891e-05,
+      "loss": 1.1611,
+      "step": 43340
+    },
+    {
+      "epoch": 12.97,
+      "grad_norm": 2.7879562377929688,
+      "learning_rate": 1.3773953918529576e-05,
+      "loss": 1.0483,
+      "step": 43345
+    },
+    {
+      "epoch": 12.97,
+      "grad_norm": 4.000429630279541,
+      "learning_rate": 1.3768704664803106e-05,
+      "loss": 1.0986,
+      "step": 43350
+    },
+    {
+      "epoch": 12.97,
+      "grad_norm": 3.5371413230895996,
+      "learning_rate": 1.3763456031369404e-05,
+      "loss": 1.0504,
+      "step": 43355
+    },
+    {
+      "epoch": 12.97,
+      "grad_norm": 2.885620594024658,
+      "learning_rate": 1.3758208018518346e-05,
+      "loss": 0.8533,
+      "step": 43360
+    },
+    {
+      "epoch": 12.97,
+      "grad_norm": 1.5964106321334839,
+      "learning_rate": 1.3752960626539774e-05,
+      "loss": 1.0267,
+      "step": 43365
+    },
+    {
+      "epoch": 12.98,
+      "grad_norm": 2.651956558227539,
+      "learning_rate": 1.3747713855723493e-05,
+      "loss": 1.0612,
+      "step": 43370
+    },
+    {
+      "epoch": 12.98,
+      "grad_norm": 1.7484973669052124,
+      "learning_rate": 1.3742467706359283e-05,
+      "loss": 1.1684,
+      "step": 43375
+    },
+    {
+      "epoch": 12.98,
+      "grad_norm": 2.3421952724456787,
+      "learning_rate": 1.373722217873688e-05,
+      "loss": 1.1648,
+      "step": 43380
+    },
+    {
+      "epoch": 12.98,
+      "grad_norm": 2.6561176776885986,
+      "learning_rate": 1.3731977273145984e-05,
+      "loss": 0.9651,
+      "step": 43385
+    },
+    {
+      "epoch": 12.98,
+      "grad_norm": 2.5251173973083496,
+      "learning_rate": 1.3726732989876278e-05,
+      "loss": 0.9813,
+      "step": 43390
+    },
+    {
+      "epoch": 12.98,
+      "grad_norm": 2.7231156826019287,
+      "learning_rate": 1.3721489329217385e-05,
+      "loss": 1.1203,
+      "step": 43395
+    },
+    {
+      "epoch": 12.98,
+      "grad_norm": 2.138787269592285,
+      "learning_rate": 1.3716246291458918e-05,
+      "loss": 1.0665,
+      "step": 43400
+    },
+    {
+      "epoch": 12.99,
+      "grad_norm": 7.916914939880371,
+      "learning_rate": 1.3711003876890438e-05,
+      "loss": 1.1231,
+      "step": 43405
+    },
+    {
+      "epoch": 12.99,
+      "grad_norm": 2.818751811981201,
+      "learning_rate": 1.3705762085801478e-05,
+      "loss": 1.101,
+      "step": 43410
+    },
+    {
+      "epoch": 12.99,
+      "grad_norm": 2.751744031906128,
+      "learning_rate": 1.370052091848154e-05,
+      "loss": 0.9642,
+      "step": 43415
+    },
+    {
+      "epoch": 12.99,
+      "grad_norm": 1.3935418128967285,
+      "learning_rate": 1.3695280375220093e-05,
+      "loss": 0.9676,
+      "step": 43420
+    },
+    {
+      "epoch": 12.99,
+      "grad_norm": 2.9495058059692383,
+      "learning_rate": 1.3690040456306546e-05,
+      "loss": 1.0966,
+      "step": 43425
+    },
+    {
+      "epoch": 12.99,
+      "grad_norm": 3.748474359512329,
+      "learning_rate": 1.3684801162030326e-05,
+      "loss": 1.0841,
+      "step": 43430
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 1.8227459192276,
+      "learning_rate": 1.3679562492680754e-05,
+      "loss": 0.9915,
+      "step": 43435
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 4.090573310852051,
+      "learning_rate": 1.3674324448547201e-05,
+      "loss": 0.8671,
+      "step": 43440
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 3.219416856765747,
+      "learning_rate": 1.3669087029918925e-05,
+      "loss": 1.0274,
+      "step": 43445
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 1.7496566772460938,
+      "learning_rate": 1.3663850237085196e-05,
+      "loss": 1.1035,
+      "step": 43450
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 1.8964442014694214,
+      "learning_rate": 1.3658614070335236e-05,
+      "loss": 0.9436,
+      "step": 43455
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 1.2473254203796387,
+      "learning_rate": 1.365337852995823e-05,
+      "loss": 0.9611,
+      "step": 43460
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 2.2625179290771484,
+      "learning_rate": 1.3648143616243334e-05,
+      "loss": 0.8075,
+      "step": 43465
+    },
+    {
+      "epoch": 13.01,
+      "grad_norm": 1.278442144393921,
+      "learning_rate": 1.3642909329479666e-05,
+      "loss": 0.8595,
+      "step": 43470
+    },
+    {
+      "epoch": 13.01,
+      "grad_norm": 3.522331953048706,
+      "learning_rate": 1.3637675669956312e-05,
+      "loss": 1.0262,
+      "step": 43475
+    },
+    {
+      "epoch": 13.01,
+      "grad_norm": 2.990205764770508,
+      "learning_rate": 1.363244263796232e-05,
+      "loss": 1.0063,
+      "step": 43480
+    },
+    {
+      "epoch": 13.01,
+      "grad_norm": 1.9036306142807007,
+      "learning_rate": 1.3627210233786705e-05,
+      "loss": 0.8788,
+      "step": 43485
+    },
+    {
+      "epoch": 13.01,
+      "grad_norm": 3.0223844051361084,
+      "learning_rate": 1.3621978457718449e-05,
+      "loss": 0.9906,
+      "step": 43490
+    },
+    {
+      "epoch": 13.01,
+      "grad_norm": 2.851306676864624,
+      "learning_rate": 1.3616747310046493e-05,
+      "loss": 1.1554,
+      "step": 43495
+    },
+    {
+      "epoch": 13.01,
+      "grad_norm": 1.1327463388442993,
+      "learning_rate": 1.3611516791059754e-05,
+      "loss": 0.9168,
+      "step": 43500
+    },
+    {
+      "epoch": 13.02,
+      "grad_norm": 2.258399248123169,
+      "learning_rate": 1.3606286901047099e-05,
+      "loss": 0.9594,
+      "step": 43505
+    },
+    {
+      "epoch": 13.02,
+      "grad_norm": 2.515151262283325,
+      "learning_rate": 1.3601057640297382e-05,
+      "loss": 0.961,
+      "step": 43510
+    },
+    {
+      "epoch": 13.02,
+      "grad_norm": 5.121065616607666,
+      "learning_rate": 1.359582900909941e-05,
+      "loss": 0.8494,
+      "step": 43515
+    },
+    {
+      "epoch": 13.02,
+      "grad_norm": 2.3708105087280273,
+      "learning_rate": 1.3590601007741926e-05,
+      "loss": 0.9847,
+      "step": 43520
+    },
+    {
+      "epoch": 13.02,
+      "grad_norm": 1.6514267921447754,
+      "learning_rate": 1.3585373636513715e-05,
+      "loss": 1.1194,
+      "step": 43525
+    },
+    {
+      "epoch": 13.02,
+      "grad_norm": 4.888957977294922,
+      "learning_rate": 1.3580146895703428e-05,
+      "loss": 1.012,
+      "step": 43530
+    },
+    {
+      "epoch": 13.03,
+      "grad_norm": 3.775286912918091,
+      "learning_rate": 1.357492078559978e-05,
+      "loss": 1.0716,
+      "step": 43535
+    },
+    {
+      "epoch": 13.03,
+      "grad_norm": 2.8520710468292236,
+      "learning_rate": 1.3569695306491358e-05,
+      "loss": 0.9447,
+      "step": 43540
+    },
+    {
+      "epoch": 13.03,
+      "grad_norm": 1.2906252145767212,
+      "learning_rate": 1.3564470458666806e-05,
+      "loss": 0.7763,
+      "step": 43545
+    },
+    {
+      "epoch": 13.03,
+      "grad_norm": 3.1408872604370117,
+      "learning_rate": 1.3559246242414653e-05,
+      "loss": 0.9982,
+      "step": 43550
+    },
+    {
+      "epoch": 13.03,
+      "grad_norm": 3.5548839569091797,
+      "learning_rate": 1.3554022658023442e-05,
+      "loss": 1.1682,
+      "step": 43555
+    },
+    {
+      "epoch": 13.03,
+      "grad_norm": 2.9345786571502686,
+      "learning_rate": 1.3548799705781655e-05,
+      "loss": 1.0853,
+      "step": 43560
+    },
+    {
+      "epoch": 13.03,
+      "grad_norm": 1.7587555646896362,
+      "learning_rate": 1.3543577385977763e-05,
+      "loss": 0.9033,
+      "step": 43565
+    },
+    {
+      "epoch": 13.04,
+      "grad_norm": 2.7659850120544434,
+      "learning_rate": 1.3538355698900182e-05,
+      "loss": 0.9808,
+      "step": 43570
+    },
+    {
+      "epoch": 13.04,
+      "grad_norm": 3.0085887908935547,
+      "learning_rate": 1.3533134644837303e-05,
+      "loss": 1.2206,
+      "step": 43575
+    },
+    {
+      "epoch": 13.04,
+      "grad_norm": 2.224841594696045,
+      "learning_rate": 1.3527914224077475e-05,
+      "loss": 1.0777,
+      "step": 43580
+    },
+    {
+      "epoch": 13.04,
+      "grad_norm": 2.03397536277771,
+      "learning_rate": 1.3522694436909022e-05,
+      "loss": 0.9373,
+      "step": 43585
+    },
+    {
+      "epoch": 13.04,
+      "grad_norm": 1.9126800298690796,
+      "learning_rate": 1.3517475283620226e-05,
+      "loss": 0.9797,
+      "step": 43590
+    },
+    {
+      "epoch": 13.04,
+      "grad_norm": 1.9172559976577759,
+      "learning_rate": 1.3512256764499335e-05,
+      "loss": 1.051,
+      "step": 43595
+    },
+    {
+      "epoch": 13.04,
+      "grad_norm": 3.562412738800049,
+      "learning_rate": 1.3507038879834561e-05,
+      "loss": 0.9013,
+      "step": 43600
+    },
+    {
+      "epoch": 13.05,
+      "grad_norm": 3.7953248023986816,
+      "learning_rate": 1.3501821629914082e-05,
+      "loss": 0.9603,
+      "step": 43605
+    },
+    {
+      "epoch": 13.05,
+      "grad_norm": 5.259886741638184,
+      "learning_rate": 1.3496605015026054e-05,
+      "loss": 0.9906,
+      "step": 43610
+    },
+    {
+      "epoch": 13.05,
+      "grad_norm": 1.6214237213134766,
+      "learning_rate": 1.3491389035458557e-05,
+      "loss": 0.8656,
+      "step": 43615
+    },
+    {
+      "epoch": 13.05,
+      "grad_norm": 2.001491069793701,
+      "learning_rate": 1.3486173691499698e-05,
+      "loss": 0.8376,
+      "step": 43620
+    },
+    {
+      "epoch": 13.05,
+      "grad_norm": 2.1933164596557617,
+      "learning_rate": 1.3480958983437475e-05,
+      "loss": 1.0552,
+      "step": 43625
+    },
+    {
+      "epoch": 13.05,
+      "grad_norm": 1.1674703359603882,
+      "learning_rate": 1.347574491155994e-05,
+      "loss": 0.9067,
+      "step": 43630
+    },
+    {
+      "epoch": 13.06,
+      "grad_norm": 3.31339693069458,
+      "learning_rate": 1.347053147615501e-05,
+      "loss": 0.9645,
+      "step": 43635
+    },
+    {
+      "epoch": 13.06,
+      "grad_norm": 2.1307213306427,
+      "learning_rate": 1.3465318677510664e-05,
+      "loss": 1.1741,
+      "step": 43640
+    },
+    {
+      "epoch": 13.06,
+      "grad_norm": 1.4840748310089111,
+      "learning_rate": 1.346010651591477e-05,
+      "loss": 0.8708,
+      "step": 43645
+    },
+    {
+      "epoch": 13.06,
+      "grad_norm": 1.7846823930740356,
+      "learning_rate": 1.3454894991655196e-05,
+      "loss": 0.8823,
+      "step": 43650
+    },
+    {
+      "epoch": 13.06,
+      "grad_norm": 16.712650299072266,
+      "learning_rate": 1.3449684105019772e-05,
+      "loss": 1.0997,
+      "step": 43655
+    },
+    {
+      "epoch": 13.06,
+      "grad_norm": 3.8388795852661133,
+      "learning_rate": 1.3444473856296277e-05,
+      "loss": 0.992,
+      "step": 43660
+    },
+    {
+      "epoch": 13.06,
+      "grad_norm": 1.0792344808578491,
+      "learning_rate": 1.3439264245772502e-05,
+      "loss": 0.9544,
+      "step": 43665
+    },
+    {
+      "epoch": 13.07,
+      "grad_norm": 2.6744980812072754,
+      "learning_rate": 1.3434055273736135e-05,
+      "loss": 0.9327,
+      "step": 43670
+    },
+    {
+      "epoch": 13.07,
+      "grad_norm": 4.98563289642334,
+      "learning_rate": 1.3428846940474871e-05,
+      "loss": 1.0448,
+      "step": 43675
+    },
+    {
+      "epoch": 13.07,
+      "grad_norm": 1.9465887546539307,
+      "learning_rate": 1.3423639246276366e-05,
+      "loss": 1.1599,
+      "step": 43680
+    },
+    {
+      "epoch": 13.07,
+      "grad_norm": 3.1894278526306152,
+      "learning_rate": 1.3418432191428232e-05,
+      "loss": 1.129,
+      "step": 43685
+    },
+    {
+      "epoch": 13.07,
+      "grad_norm": 1.8298451900482178,
+      "learning_rate": 1.3413225776218048e-05,
+      "loss": 1.0085,
+      "step": 43690
+    },
+    {
+      "epoch": 13.07,
+      "grad_norm": 2.9106414318084717,
+      "learning_rate": 1.3408020000933363e-05,
+      "loss": 1.0084,
+      "step": 43695
+    },
+    {
+      "epoch": 13.07,
+      "grad_norm": 1.1771775484085083,
+      "learning_rate": 1.340281486586168e-05,
+      "loss": 0.889,
+      "step": 43700
+    },
+    {
+      "epoch": 13.08,
+      "grad_norm": 15.508944511413574,
+      "learning_rate": 1.3397610371290492e-05,
+      "loss": 0.9712,
+      "step": 43705
+    },
+    {
+      "epoch": 13.08,
+      "grad_norm": 3.3548052310943604,
+      "learning_rate": 1.33924065175072e-05,
+      "loss": 0.9766,
+      "step": 43710
+    },
+    {
+      "epoch": 13.08,
+      "grad_norm": 1.3726935386657715,
+      "learning_rate": 1.338720330479925e-05,
+      "loss": 1.0355,
+      "step": 43715
+    },
+    {
+      "epoch": 13.08,
+      "grad_norm": 4.749921798706055,
+      "learning_rate": 1.3382000733453976e-05,
+      "loss": 1.154,
+      "step": 43720
+    },
+    {
+      "epoch": 13.08,
+      "grad_norm": 3.290040969848633,
+      "learning_rate": 1.3376798803758742e-05,
+      "loss": 1.101,
+      "step": 43725
+    },
+    {
+      "epoch": 13.08,
+      "grad_norm": 2.6318535804748535,
+      "learning_rate": 1.3371597516000809e-05,
+      "loss": 1.0034,
+      "step": 43730
+    },
+    {
+      "epoch": 13.09,
+      "grad_norm": 2.853180170059204,
+      "learning_rate": 1.3366396870467468e-05,
+      "loss": 1.0768,
+      "step": 43735
+    },
+    {
+      "epoch": 13.09,
+      "grad_norm": 4.42519474029541,
+      "learning_rate": 1.336119686744594e-05,
+      "loss": 0.9629,
+      "step": 43740
+    },
+    {
+      "epoch": 13.09,
+      "grad_norm": 3.0336861610412598,
+      "learning_rate": 1.33559975072234e-05,
+      "loss": 0.9401,
+      "step": 43745
+    },
+    {
+      "epoch": 13.09,
+      "grad_norm": 3.247124433517456,
+      "learning_rate": 1.335079879008703e-05,
+      "loss": 0.874,
+      "step": 43750
+    },
+    {
+      "epoch": 13.09,
+      "grad_norm": 2.6699001789093018,
+      "learning_rate": 1.3345600716323913e-05,
+      "loss": 1.1099,
+      "step": 43755
+    },
+    {
+      "epoch": 13.09,
+      "grad_norm": 6.743714332580566,
+      "learning_rate": 1.3340403286221176e-05,
+      "loss": 1.1685,
+      "step": 43760
+    },
+    {
+      "epoch": 13.09,
+      "grad_norm": 1.998405933380127,
+      "learning_rate": 1.3335206500065828e-05,
+      "loss": 1.0679,
+      "step": 43765
+    },
+    {
+      "epoch": 13.1,
+      "grad_norm": 4.789541721343994,
+      "learning_rate": 1.3330010358144917e-05,
+      "loss": 0.8997,
+      "step": 43770
+    },
+    {
+      "epoch": 13.1,
+      "grad_norm": 1.4575897455215454,
+      "learning_rate": 1.3324814860745394e-05,
+      "loss": 1.0083,
+      "step": 43775
+    },
+    {
+      "epoch": 13.1,
+      "grad_norm": 3.014540672302246,
+      "learning_rate": 1.331962000815421e-05,
+      "loss": 0.9282,
+      "step": 43780
+    },
+    {
+      "epoch": 13.1,
+      "grad_norm": 2.8595590591430664,
+      "learning_rate": 1.3314425800658275e-05,
+      "loss": 1.037,
+      "step": 43785
+    },
+    {
+      "epoch": 13.1,
+      "grad_norm": 1.8672281503677368,
+      "learning_rate": 1.3309232238544457e-05,
+      "loss": 1.0224,
+      "step": 43790
+    },
+    {
+      "epoch": 13.1,
+      "grad_norm": 1.25929594039917,
+      "learning_rate": 1.3304039322099588e-05,
+      "loss": 0.9784,
+      "step": 43795
+    },
+    {
+      "epoch": 13.1,
+      "grad_norm": 2.0091841220855713,
+      "learning_rate": 1.329884705161048e-05,
+      "loss": 0.9481,
+      "step": 43800
+    },
+    {
+      "epoch": 13.11,
+      "grad_norm": 3.789120674133301,
+      "learning_rate": 1.3293655427363866e-05,
+      "loss": 0.8126,
+      "step": 43805
+    },
+    {
+      "epoch": 13.11,
+      "grad_norm": 2.5179672241210938,
+      "learning_rate": 1.3288464449646503e-05,
+      "loss": 1.1027,
+      "step": 43810
+    },
+    {
+      "epoch": 13.11,
+      "grad_norm": 3.7937817573547363,
+      "learning_rate": 1.3283274118745079e-05,
+      "loss": 1.04,
+      "step": 43815
+    },
+    {
+      "epoch": 13.11,
+      "grad_norm": 3.776716947555542,
+      "learning_rate": 1.3278084434946247e-05,
+      "loss": 1.1246,
+      "step": 43820
+    },
+    {
+      "epoch": 13.11,
+      "grad_norm": 3.676992654800415,
+      "learning_rate": 1.3272895398536623e-05,
+      "loss": 0.9363,
+      "step": 43825
+    },
+    {
+      "epoch": 13.11,
+      "grad_norm": 2.452056407928467,
+      "learning_rate": 1.32677070098028e-05,
+      "loss": 1.1244,
+      "step": 43830
+    },
+    {
+      "epoch": 13.11,
+      "grad_norm": 4.2442307472229,
+      "learning_rate": 1.3262519269031331e-05,
+      "loss": 1.0792,
+      "step": 43835
+    },
+    {
+      "epoch": 13.12,
+      "grad_norm": 4.072299957275391,
+      "learning_rate": 1.3257332176508708e-05,
+      "loss": 0.9724,
+      "step": 43840
+    },
+    {
+      "epoch": 13.12,
+      "grad_norm": 5.6725311279296875,
+      "learning_rate": 1.3252145732521438e-05,
+      "loss": 0.9115,
+      "step": 43845
+    },
+    {
+      "epoch": 13.12,
+      "grad_norm": 4.147802829742432,
+      "learning_rate": 1.324695993735593e-05,
+      "loss": 1.1873,
+      "step": 43850
+    },
+    {
+      "epoch": 13.12,
+      "grad_norm": 2.995326042175293,
+      "learning_rate": 1.3241774791298628e-05,
+      "loss": 1.0866,
+      "step": 43855
+    },
+    {
+      "epoch": 13.12,
+      "grad_norm": 1.8452882766723633,
+      "learning_rate": 1.323659029463586e-05,
+      "loss": 1.1149,
+      "step": 43860
+    },
+    {
+      "epoch": 13.12,
+      "grad_norm": 3.185358762741089,
+      "learning_rate": 1.3231406447654004e-05,
+      "loss": 1.03,
+      "step": 43865
+    },
+    {
+      "epoch": 13.13,
+      "grad_norm": 5.549739360809326,
+      "learning_rate": 1.3226223250639328e-05,
+      "loss": 0.9462,
+      "step": 43870
+    },
+    {
+      "epoch": 13.13,
+      "grad_norm": 1.393759846687317,
+      "learning_rate": 1.3221040703878101e-05,
+      "loss": 1.0435,
+      "step": 43875
+    },
+    {
+      "epoch": 13.13,
+      "grad_norm": 1.9850372076034546,
+      "learning_rate": 1.3215858807656556e-05,
+      "loss": 1.0667,
+      "step": 43880
+    },
+    {
+      "epoch": 13.13,
+      "grad_norm": 2.1288838386535645,
+      "learning_rate": 1.3210677562260878e-05,
+      "loss": 0.9594,
+      "step": 43885
+    },
+    {
+      "epoch": 13.13,
+      "grad_norm": 1.302878975868225,
+      "learning_rate": 1.3205496967977226e-05,
+      "loss": 0.9986,
+      "step": 43890
+    },
+    {
+      "epoch": 13.13,
+      "grad_norm": 2.470869302749634,
+      "learning_rate": 1.3200317025091712e-05,
+      "loss": 0.9677,
+      "step": 43895
+    },
+    {
+      "epoch": 13.13,
+      "grad_norm": 3.7706472873687744,
+      "learning_rate": 1.3195137733890428e-05,
+      "loss": 0.9866,
+      "step": 43900
+    },
+    {
+      "epoch": 13.14,
+      "grad_norm": 1.7888362407684326,
+      "learning_rate": 1.3189959094659415e-05,
+      "loss": 1.061,
+      "step": 43905
+    },
+    {
+      "epoch": 13.14,
+      "grad_norm": 2.0343103408813477,
+      "learning_rate": 1.3184781107684687e-05,
+      "loss": 1.1638,
+      "step": 43910
+    },
+    {
+      "epoch": 13.14,
+      "grad_norm": 3.1971328258514404,
+      "learning_rate": 1.3179603773252217e-05,
+      "loss": 1.0671,
+      "step": 43915
+    },
+    {
+      "epoch": 13.14,
+      "grad_norm": 3.4805822372436523,
+      "learning_rate": 1.3174427091647943e-05,
+      "loss": 1.0129,
+      "step": 43920
+    },
+    {
+      "epoch": 13.14,
+      "grad_norm": 1.6715178489685059,
+      "learning_rate": 1.3169251063157767e-05,
+      "loss": 0.8938,
+      "step": 43925
+    },
+    {
+      "epoch": 13.14,
+      "grad_norm": 3.162388324737549,
+      "learning_rate": 1.3164075688067567e-05,
+      "loss": 0.8423,
+      "step": 43930
+    },
+    {
+      "epoch": 13.14,
+      "grad_norm": 2.0230519771575928,
+      "learning_rate": 1.3158900966663148e-05,
+      "loss": 1.0128,
+      "step": 43935
+    },
+    {
+      "epoch": 13.15,
+      "grad_norm": 2.7525076866149902,
+      "learning_rate": 1.315372689923034e-05,
+      "loss": 1.0519,
+      "step": 43940
+    },
+    {
+      "epoch": 13.15,
+      "grad_norm": 2.027358055114746,
+      "learning_rate": 1.314855348605486e-05,
+      "loss": 0.9151,
+      "step": 43945
+    },
+    {
+      "epoch": 13.15,
+      "grad_norm": 1.6696336269378662,
+      "learning_rate": 1.3143380727422472e-05,
+      "loss": 0.9351,
+      "step": 43950
+    },
+    {
+      "epoch": 13.15,
+      "grad_norm": 3.3159263134002686,
+      "learning_rate": 1.3138208623618823e-05,
+      "loss": 1.0519,
+      "step": 43955
+    },
+    {
+      "epoch": 13.15,
+      "grad_norm": 12.850335121154785,
+      "learning_rate": 1.3133037174929602e-05,
+      "loss": 1.0565,
+      "step": 43960
+    },
+    {
+      "epoch": 13.15,
+      "grad_norm": 2.337259531021118,
+      "learning_rate": 1.3127866381640392e-05,
+      "loss": 1.0833,
+      "step": 43965
+    },
+    {
+      "epoch": 13.16,
+      "grad_norm": 2.980786085128784,
+      "learning_rate": 1.312269624403678e-05,
+      "loss": 1.0833,
+      "step": 43970
+    },
+    {
+      "epoch": 13.16,
+      "grad_norm": 3.550776958465576,
+      "learning_rate": 1.311752676240431e-05,
+      "loss": 0.792,
+      "step": 43975
+    },
+    {
+      "epoch": 13.16,
+      "grad_norm": 5.096564292907715,
+      "learning_rate": 1.3112357937028488e-05,
+      "loss": 0.8951,
+      "step": 43980
+    },
+    {
+      "epoch": 13.16,
+      "grad_norm": 3.9432437419891357,
+      "learning_rate": 1.3107189768194777e-05,
+      "loss": 1.0846,
+      "step": 43985
+    },
+    {
+      "epoch": 13.16,
+      "grad_norm": 3.766063928604126,
+      "learning_rate": 1.3102022256188618e-05,
+      "loss": 1.151,
+      "step": 43990
+    },
+    {
+      "epoch": 13.16,
+      "grad_norm": 1.3759208917617798,
+      "learning_rate": 1.30968554012954e-05,
+      "loss": 0.8187,
+      "step": 43995
+    },
+    {
+      "epoch": 13.16,
+      "grad_norm": 3.239544153213501,
+      "learning_rate": 1.3091689203800483e-05,
+      "loss": 0.8463,
+      "step": 44000
+    },
+    {
+      "epoch": 13.17,
+      "grad_norm": 4.143579959869385,
+      "learning_rate": 1.3086523663989197e-05,
+      "loss": 1.0706,
+      "step": 44005
+    },
+    {
+      "epoch": 13.17,
+      "grad_norm": 1.883518934249878,
+      "learning_rate": 1.3081358782146823e-05,
+      "loss": 0.9982,
+      "step": 44010
+    },
+    {
+      "epoch": 13.17,
+      "grad_norm": 1.332044243812561,
+      "learning_rate": 1.3076194558558617e-05,
+      "loss": 0.9293,
+      "step": 44015
+    },
+    {
+      "epoch": 13.17,
+      "grad_norm": 3.111307382583618,
+      "learning_rate": 1.3071030993509788e-05,
+      "loss": 0.9561,
+      "step": 44020
+    },
+    {
+      "epoch": 13.17,
+      "grad_norm": 2.8684184551239014,
+      "learning_rate": 1.3065868087285533e-05,
+      "loss": 0.9597,
+      "step": 44025
+    },
+    {
+      "epoch": 13.17,
+      "grad_norm": 2.7545359134674072,
+      "learning_rate": 1.3060705840170953e-05,
+      "loss": 0.7545,
+      "step": 44030
+    },
+    {
+      "epoch": 13.17,
+      "grad_norm": 2.23874831199646,
+      "learning_rate": 1.3055544252451202e-05,
+      "loss": 0.9002,
+      "step": 44035
+    },
+    {
+      "epoch": 13.18,
+      "grad_norm": 3.11095929145813,
+      "learning_rate": 1.3050383324411308e-05,
+      "loss": 0.9736,
+      "step": 44040
+    },
+    {
+      "epoch": 13.18,
+      "grad_norm": 2.461932897567749,
+      "learning_rate": 1.3045223056336339e-05,
+      "loss": 1.1343,
+      "step": 44045
+    },
+    {
+      "epoch": 13.18,
+      "grad_norm": 3.571146011352539,
+      "learning_rate": 1.3040063448511255e-05,
+      "loss": 0.9675,
+      "step": 44050
+    },
+    {
+      "epoch": 13.18,
+      "grad_norm": 3.9316980838775635,
+      "learning_rate": 1.3034904501221057e-05,
+      "loss": 0.9565,
+      "step": 44055
+    },
+    {
+      "epoch": 13.18,
+      "grad_norm": 5.21784782409668,
+      "learning_rate": 1.3029746214750636e-05,
+      "loss": 0.9655,
+      "step": 44060
+    },
+    {
+      "epoch": 13.18,
+      "grad_norm": 3.174137830734253,
+      "learning_rate": 1.3024588589384887e-05,
+      "loss": 1.0195,
+      "step": 44065
+    },
+    {
+      "epoch": 13.19,
+      "grad_norm": 3.539970636367798,
+      "learning_rate": 1.3019431625408663e-05,
+      "loss": 1.0647,
+      "step": 44070
+    },
+    {
+      "epoch": 13.19,
+      "grad_norm": 4.315314769744873,
+      "learning_rate": 1.3014275323106777e-05,
+      "loss": 1.1308,
+      "step": 44075
+    },
+    {
+      "epoch": 13.19,
+      "grad_norm": 2.9523892402648926,
+      "learning_rate": 1.300911968276401e-05,
+      "loss": 0.9636,
+      "step": 44080
+    },
+    {
+      "epoch": 13.19,
+      "grad_norm": 2.1555330753326416,
+      "learning_rate": 1.3003964704665096e-05,
+      "loss": 1.0895,
+      "step": 44085
+    },
+    {
+      "epoch": 13.19,
+      "grad_norm": 1.9076026678085327,
+      "learning_rate": 1.2998810389094743e-05,
+      "loss": 0.959,
+      "step": 44090
+    },
+    {
+      "epoch": 13.19,
+      "grad_norm": 2.8785440921783447,
+      "learning_rate": 1.2993656736337617e-05,
+      "loss": 0.9798,
+      "step": 44095
+    },
+    {
+      "epoch": 13.19,
+      "grad_norm": 7.201198101043701,
+      "learning_rate": 1.2988503746678348e-05,
+      "loss": 0.9449,
+      "step": 44100
+    },
+    {
+      "epoch": 13.2,
+      "grad_norm": 2.4313573837280273,
+      "learning_rate": 1.2983351420401535e-05,
+      "loss": 1.0374,
+      "step": 44105
+    },
+    {
+      "epoch": 13.2,
+      "grad_norm": 3.2586467266082764,
+      "learning_rate": 1.297819975779173e-05,
+      "loss": 0.9367,
+      "step": 44110
+    },
+    {
+      "epoch": 13.2,
+      "grad_norm": 2.9108598232269287,
+      "learning_rate": 1.297304875913346e-05,
+      "loss": 1.2829,
+      "step": 44115
+    },
+    {
+      "epoch": 13.2,
+      "grad_norm": 4.189149379730225,
+      "learning_rate": 1.2967898424711203e-05,
+      "loss": 0.8909,
+      "step": 44120
+    },
+    {
+      "epoch": 13.2,
+      "grad_norm": 2.829456090927124,
+      "learning_rate": 1.296274875480941e-05,
+      "loss": 1.0128,
+      "step": 44125
+    },
+    {
+      "epoch": 13.2,
+      "grad_norm": 1.993833303451538,
+      "learning_rate": 1.29575997497125e-05,
+      "loss": 0.9394,
+      "step": 44130
+    },
+    {
+      "epoch": 13.2,
+      "grad_norm": 1.8777223825454712,
+      "learning_rate": 1.2952451409704818e-05,
+      "loss": 0.9643,
+      "step": 44135
+    },
+    {
+      "epoch": 13.21,
+      "grad_norm": 2.4245407581329346,
+      "learning_rate": 1.2947303735070742e-05,
+      "loss": 0.9848,
+      "step": 44140
+    },
+    {
+      "epoch": 13.21,
+      "grad_norm": 3.382310628890991,
+      "learning_rate": 1.2942156726094534e-05,
+      "loss": 1.0159,
+      "step": 44145
+    },
+    {
+      "epoch": 13.21,
+      "grad_norm": 3.079493761062622,
+      "learning_rate": 1.29370103830605e-05,
+      "loss": 0.8951,
+      "step": 44150
+    },
+    {
+      "epoch": 13.21,
+      "grad_norm": 2.4276509284973145,
+      "learning_rate": 1.2931864706252828e-05,
+      "loss": 1.2272,
+      "step": 44155
+    },
+    {
+      "epoch": 13.21,
+      "grad_norm": 3.1140429973602295,
+      "learning_rate": 1.2926719695955727e-05,
+      "loss": 1.0597,
+      "step": 44160
+    },
+    {
+      "epoch": 13.21,
+      "grad_norm": 2.4898757934570312,
+      "learning_rate": 1.2921575352453347e-05,
+      "loss": 0.9542,
+      "step": 44165
+    },
+    {
+      "epoch": 13.22,
+      "grad_norm": 2.116917133331299,
+      "learning_rate": 1.2916431676029806e-05,
+      "loss": 0.9554,
+      "step": 44170
+    },
+    {
+      "epoch": 13.22,
+      "grad_norm": 3.3684239387512207,
+      "learning_rate": 1.2911288666969184e-05,
+      "loss": 1.057,
+      "step": 44175
+    },
+    {
+      "epoch": 13.22,
+      "grad_norm": 3.8344614505767822,
+      "learning_rate": 1.2906146325555522e-05,
+      "loss": 0.9452,
+      "step": 44180
+    },
+    {
+      "epoch": 13.22,
+      "grad_norm": 3.2199113368988037,
+      "learning_rate": 1.2901004652072826e-05,
+      "loss": 0.9491,
+      "step": 44185
+    },
+    {
+      "epoch": 13.22,
+      "grad_norm": 2.2717814445495605,
+      "learning_rate": 1.289586364680507e-05,
+      "loss": 1.0688,
+      "step": 44190
+    },
+    {
+      "epoch": 13.22,
+      "grad_norm": 2.989772319793701,
+      "learning_rate": 1.2890723310036181e-05,
+      "loss": 1.018,
+      "step": 44195
+    },
+    {
+      "epoch": 13.22,
+      "grad_norm": 9.16136360168457,
+      "learning_rate": 1.2885583642050058e-05,
+      "loss": 1.0984,
+      "step": 44200
+    },
+    {
+      "epoch": 13.23,
+      "grad_norm": 1.604578971862793,
+      "learning_rate": 1.2880444643130556e-05,
+      "loss": 1.0616,
+      "step": 44205
+    },
+    {
+      "epoch": 13.23,
+      "grad_norm": 2.7663040161132812,
+      "learning_rate": 1.2875306313561502e-05,
+      "loss": 0.9586,
+      "step": 44210
+    },
+    {
+      "epoch": 13.23,
+      "grad_norm": 1.7422765493392944,
+      "learning_rate": 1.2870168653626677e-05,
+      "loss": 1.0868,
+      "step": 44215
+    },
+    {
+      "epoch": 13.23,
+      "grad_norm": 3.091707944869995,
+      "learning_rate": 1.2865031663609827e-05,
+      "loss": 1.1116,
+      "step": 44220
+    },
+    {
+      "epoch": 13.23,
+      "grad_norm": 2.8757541179656982,
+      "learning_rate": 1.2859895343794676e-05,
+      "loss": 1.0587,
+      "step": 44225
+    },
+    {
+      "epoch": 13.23,
+      "grad_norm": 2.7807388305664062,
+      "learning_rate": 1.2854759694464868e-05,
+      "loss": 1.175,
+      "step": 44230
+    },
+    {
+      "epoch": 13.23,
+      "grad_norm": 1.115530252456665,
+      "learning_rate": 1.2849624715904074e-05,
+      "loss": 0.9452,
+      "step": 44235
+    },
+    {
+      "epoch": 13.24,
+      "grad_norm": 4.6550822257995605,
+      "learning_rate": 1.284449040839586e-05,
+      "loss": 1.0934,
+      "step": 44240
+    },
+    {
+      "epoch": 13.24,
+      "grad_norm": 2.822744369506836,
+      "learning_rate": 1.2839356772223823e-05,
+      "loss": 1.1822,
+      "step": 44245
+    },
+    {
+      "epoch": 13.24,
+      "grad_norm": 2.8846404552459717,
+      "learning_rate": 1.2834223807671453e-05,
+      "loss": 1.1075,
+      "step": 44250
+    },
+    {
+      "epoch": 13.24,
+      "grad_norm": 2.568181037902832,
+      "learning_rate": 1.2829091515022279e-05,
+      "loss": 0.876,
+      "step": 44255
+    },
+    {
+      "epoch": 13.24,
+      "grad_norm": 5.2756123542785645,
+      "learning_rate": 1.2823959894559718e-05,
+      "loss": 1.0383,
+      "step": 44260
+    },
+    {
+      "epoch": 13.24,
+      "grad_norm": 1.8959022760391235,
+      "learning_rate": 1.2818828946567194e-05,
+      "loss": 0.9319,
+      "step": 44265
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 2.6469838619232178,
+      "learning_rate": 1.2813698671328084e-05,
+      "loss": 1.0376,
+      "step": 44270
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 2.5506365299224854,
+      "learning_rate": 1.2808569069125734e-05,
+      "loss": 0.944,
+      "step": 44275
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 2.260101079940796,
+      "learning_rate": 1.280344014024344e-05,
+      "loss": 0.8573,
+      "step": 44280
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 1.153891921043396,
+      "learning_rate": 1.279831188496447e-05,
+      "loss": 1.0468,
+      "step": 44285
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 3.153810977935791,
+      "learning_rate": 1.279318430357205e-05,
+      "loss": 0.854,
+      "step": 44290
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 2.335554838180542,
+      "learning_rate": 1.2788057396349374e-05,
+      "loss": 0.9501,
+      "step": 44295
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 1.3886281251907349,
+      "learning_rate": 1.2782931163579593e-05,
+      "loss": 1.0305,
+      "step": 44300
+    },
+    {
+      "epoch": 13.26,
+      "grad_norm": 1.6027575731277466,
+      "learning_rate": 1.2777805605545828e-05,
+      "loss": 0.9369,
+      "step": 44305
+    },
+    {
+      "epoch": 13.26,
+      "grad_norm": 2.810401201248169,
+      "learning_rate": 1.2772680722531153e-05,
+      "loss": 0.8592,
+      "step": 44310
+    },
+    {
+      "epoch": 13.26,
+      "grad_norm": 4.480473518371582,
+      "learning_rate": 1.2767556514818613e-05,
+      "loss": 0.9118,
+      "step": 44315
+    },
+    {
+      "epoch": 13.26,
+      "grad_norm": 5.290129661560059,
+      "learning_rate": 1.2762432982691219e-05,
+      "loss": 0.8724,
+      "step": 44320
+    },
+    {
+      "epoch": 13.26,
+      "grad_norm": 3.318096160888672,
+      "learning_rate": 1.2757310126431915e-05,
+      "loss": 0.8357,
+      "step": 44325
+    },
+    {
+      "epoch": 13.26,
+      "grad_norm": 2.25472354888916,
+      "learning_rate": 1.2752187946323663e-05,
+      "loss": 0.9452,
+      "step": 44330
+    },
+    {
+      "epoch": 13.26,
+      "grad_norm": 3.4436185359954834,
+      "learning_rate": 1.2747066442649324e-05,
+      "loss": 0.9833,
+      "step": 44335
+    },
+    {
+      "epoch": 13.27,
+      "grad_norm": 5.141889572143555,
+      "learning_rate": 1.2741945615691785e-05,
+      "loss": 1.0254,
+      "step": 44340
+    },
+    {
+      "epoch": 13.27,
+      "grad_norm": 2.958003520965576,
+      "learning_rate": 1.2736825465733832e-05,
+      "loss": 0.8983,
+      "step": 44345
+    },
+    {
+      "epoch": 13.27,
+      "grad_norm": 1.62663996219635,
+      "learning_rate": 1.2731705993058279e-05,
+      "loss": 0.9047,
+      "step": 44350
+    },
+    {
+      "epoch": 13.27,
+      "grad_norm": 1.6435695886611938,
+      "learning_rate": 1.2726587197947843e-05,
+      "loss": 0.9762,
+      "step": 44355
+    },
+    {
+      "epoch": 13.27,
+      "grad_norm": 2.0215165615081787,
+      "learning_rate": 1.2721469080685238e-05,
+      "loss": 1.058,
+      "step": 44360
+    },
+    {
+      "epoch": 13.27,
+      "grad_norm": 7.587679386138916,
+      "learning_rate": 1.2716351641553137e-05,
+      "loss": 0.797,
+      "step": 44365
+    },
+    {
+      "epoch": 13.27,
+      "grad_norm": 4.353124141693115,
+      "learning_rate": 1.271123488083416e-05,
+      "loss": 0.9351,
+      "step": 44370
+    },
+    {
+      "epoch": 13.28,
+      "grad_norm": 1.8056886196136475,
+      "learning_rate": 1.2706118798810912e-05,
+      "loss": 1.0511,
+      "step": 44375
+    },
+    {
+      "epoch": 13.28,
+      "grad_norm": 2.1814777851104736,
+      "learning_rate": 1.2701003395765942e-05,
+      "loss": 0.9481,
+      "step": 44380
+    },
+    {
+      "epoch": 13.28,
+      "grad_norm": 2.314023017883301,
+      "learning_rate": 1.2695888671981771e-05,
+      "loss": 1.0634,
+      "step": 44385
+    },
+    {
+      "epoch": 13.28,
+      "grad_norm": 2.8826498985290527,
+      "learning_rate": 1.2690774627740875e-05,
+      "loss": 0.9873,
+      "step": 44390
+    },
+    {
+      "epoch": 13.28,
+      "grad_norm": 2.8844268321990967,
+      "learning_rate": 1.2685661263325707e-05,
+      "loss": 0.8023,
+      "step": 44395
+    },
+    {
+      "epoch": 13.28,
+      "grad_norm": 2.0241379737854004,
+      "learning_rate": 1.2680548579018666e-05,
+      "loss": 0.9862,
+      "step": 44400
+    },
+    {
+      "epoch": 13.29,
+      "grad_norm": 2.5573320388793945,
+      "learning_rate": 1.2676458921440643e-05,
+      "loss": 0.8519,
+      "step": 44405
+    },
+    {
+      "epoch": 13.29,
+      "grad_norm": 2.713150978088379,
+      "learning_rate": 1.2671347462039773e-05,
+      "loss": 1.0685,
+      "step": 44410
+    },
+    {
+      "epoch": 13.29,
+      "grad_norm": 2.36755633354187,
+      "learning_rate": 1.2666236683537571e-05,
+      "loss": 0.8845,
+      "step": 44415
+    },
+    {
+      "epoch": 13.29,
+      "grad_norm": 3.59240984916687,
+      "learning_rate": 1.26611265862163e-05,
+      "loss": 1.0012,
+      "step": 44420
+    },
+    {
+      "epoch": 13.29,
+      "grad_norm": 1.345497965812683,
+      "learning_rate": 1.2656017170358175e-05,
+      "loss": 1.0763,
+      "step": 44425
+    },
+    {
+      "epoch": 13.29,
+      "grad_norm": 2.690333604812622,
+      "learning_rate": 1.2650908436245395e-05,
+      "loss": 1.0471,
+      "step": 44430
+    },
+    {
+      "epoch": 13.29,
+      "grad_norm": 3.9363012313842773,
+      "learning_rate": 1.2645800384160103e-05,
+      "loss": 1.0104,
+      "step": 44435
+    },
+    {
+      "epoch": 13.3,
+      "grad_norm": 3.2774622440338135,
+      "learning_rate": 1.2640693014384417e-05,
+      "loss": 1.052,
+      "step": 44440
+    },
+    {
+      "epoch": 13.3,
+      "grad_norm": 5.962934494018555,
+      "learning_rate": 1.2635586327200408e-05,
+      "loss": 1.044,
+      "step": 44445
+    },
+    {
+      "epoch": 13.3,
+      "grad_norm": 1.6138075590133667,
+      "learning_rate": 1.2630480322890114e-05,
+      "loss": 1.0042,
+      "step": 44450
+    },
+    {
+      "epoch": 13.3,
+      "grad_norm": 4.094041347503662,
+      "learning_rate": 1.2625375001735534e-05,
+      "loss": 0.9887,
+      "step": 44455
+    },
+    {
+      "epoch": 13.3,
+      "grad_norm": 10.620294570922852,
+      "learning_rate": 1.2620270364018633e-05,
+      "loss": 0.9852,
+      "step": 44460
+    },
+    {
+      "epoch": 13.3,
+      "grad_norm": 1.635864496231079,
+      "learning_rate": 1.2615166410021329e-05,
+      "loss": 1.0061,
+      "step": 44465
+    },
+    {
+      "epoch": 13.3,
+      "grad_norm": 5.53233528137207,
+      "learning_rate": 1.2610063140025519e-05,
+      "loss": 1.0371,
+      "step": 44470
+    },
+    {
+      "epoch": 13.31,
+      "grad_norm": 3.4139328002929688,
+      "learning_rate": 1.2604960554313027e-05,
+      "loss": 1.0421,
+      "step": 44475
+    },
+    {
+      "epoch": 13.31,
+      "grad_norm": 4.036562919616699,
+      "learning_rate": 1.2599858653165698e-05,
+      "loss": 0.8257,
+      "step": 44480
+    },
+    {
+      "epoch": 13.31,
+      "grad_norm": 5.966806411743164,
+      "learning_rate": 1.2594757436865265e-05,
+      "loss": 1.1459,
+      "step": 44485
+    },
+    {
+      "epoch": 13.31,
+      "grad_norm": 2.0404038429260254,
+      "learning_rate": 1.2589656905693503e-05,
+      "loss": 1.0175,
+      "step": 44490
+    },
+    {
+      "epoch": 13.31,
+      "grad_norm": 2.6337130069732666,
+      "learning_rate": 1.2584557059932076e-05,
+      "loss": 1.0467,
+      "step": 44495
+    },
+    {
+      "epoch": 13.31,
+      "grad_norm": 5.764756679534912,
+      "learning_rate": 1.2579457899862673e-05,
+      "loss": 1.0118,
+      "step": 44500
+    },
+    {
+      "epoch": 13.32,
+      "grad_norm": 8.07419490814209,
+      "learning_rate": 1.257435942576689e-05,
+      "loss": 1.1198,
+      "step": 44505
+    },
+    {
+      "epoch": 13.32,
+      "grad_norm": 2.342231273651123,
+      "learning_rate": 1.2569261637926322e-05,
+      "loss": 1.1913,
+      "step": 44510
+    },
+    {
+      "epoch": 13.32,
+      "grad_norm": 2.863327980041504,
+      "learning_rate": 1.2564164536622513e-05,
+      "loss": 0.9799,
+      "step": 44515
+    },
+    {
+      "epoch": 13.32,
+      "grad_norm": 2.152618885040283,
+      "learning_rate": 1.255906812213697e-05,
+      "loss": 0.9915,
+      "step": 44520
+    },
+    {
+      "epoch": 13.32,
+      "grad_norm": 1.8867714405059814,
+      "learning_rate": 1.2553972394751162e-05,
+      "loss": 0.9104,
+      "step": 44525
+    },
+    {
+      "epoch": 13.32,
+      "grad_norm": 4.3180460929870605,
+      "learning_rate": 1.2548877354746519e-05,
+      "loss": 0.8839,
+      "step": 44530
+    },
+    {
+      "epoch": 13.32,
+      "grad_norm": 3.7813620567321777,
+      "learning_rate": 1.254378300240444e-05,
+      "loss": 1.0885,
+      "step": 44535
+    },
+    {
+      "epoch": 13.33,
+      "grad_norm": 2.9510982036590576,
+      "learning_rate": 1.2538689338006282e-05,
+      "loss": 1.0409,
+      "step": 44540
+    },
+    {
+      "epoch": 13.33,
+      "grad_norm": 2.6716909408569336,
+      "learning_rate": 1.2533596361833355e-05,
+      "loss": 1.0121,
+      "step": 44545
+    },
+    {
+      "epoch": 13.33,
+      "grad_norm": 3.946298837661743,
+      "learning_rate": 1.2528504074166941e-05,
+      "loss": 0.9055,
+      "step": 44550
+    },
+    {
+      "epoch": 13.33,
+      "grad_norm": 2.191622018814087,
+      "learning_rate": 1.2523412475288288e-05,
+      "loss": 1.0652,
+      "step": 44555
+    },
+    {
+      "epoch": 13.33,
+      "grad_norm": 4.9477858543396,
+      "learning_rate": 1.2518321565478593e-05,
+      "loss": 0.8954,
+      "step": 44560
+    },
+    {
+      "epoch": 13.33,
+      "grad_norm": 1.2523143291473389,
+      "learning_rate": 1.2513231345019032e-05,
+      "loss": 1.0451,
+      "step": 44565
+    },
+    {
+      "epoch": 13.33,
+      "grad_norm": 2.840963840484619,
+      "learning_rate": 1.250814181419071e-05,
+      "loss": 0.8473,
+      "step": 44570
+    },
+    {
+      "epoch": 13.34,
+      "grad_norm": 2.3389129638671875,
+      "learning_rate": 1.2503052973274749e-05,
+      "loss": 1.0362,
+      "step": 44575
+    },
+    {
+      "epoch": 13.34,
+      "grad_norm": 2.131584882736206,
+      "learning_rate": 1.2497964822552161e-05,
+      "loss": 1.0984,
+      "step": 44580
+    },
+    {
+      "epoch": 13.34,
+      "grad_norm": 6.629162311553955,
+      "learning_rate": 1.2492877362304e-05,
+      "loss": 1.0674,
+      "step": 44585
+    },
+    {
+      "epoch": 13.34,
+      "grad_norm": 4.698531627655029,
+      "learning_rate": 1.2487790592811202e-05,
+      "loss": 0.8879,
+      "step": 44590
+    },
+    {
+      "epoch": 13.34,
+      "grad_norm": 3.542203187942505,
+      "learning_rate": 1.2482704514354746e-05,
+      "loss": 0.916,
+      "step": 44595
+    },
+    {
+      "epoch": 13.34,
+      "grad_norm": 6.20295524597168,
+      "learning_rate": 1.2477619127215498e-05,
+      "loss": 1.0514,
+      "step": 44600
+    },
+    {
+      "epoch": 13.35,
+      "grad_norm": 7.1144022941589355,
+      "learning_rate": 1.2472534431674327e-05,
+      "loss": 0.9958,
+      "step": 44605
+    },
+    {
+      "epoch": 13.35,
+      "grad_norm": 1.586234211921692,
+      "learning_rate": 1.2467450428012059e-05,
+      "loss": 0.9424,
+      "step": 44610
+    },
+    {
+      "epoch": 13.35,
+      "grad_norm": 3.302457332611084,
+      "learning_rate": 1.246236711650948e-05,
+      "loss": 0.9997,
+      "step": 44615
+    },
+    {
+      "epoch": 13.35,
+      "grad_norm": 3.5869815349578857,
+      "learning_rate": 1.245728449744733e-05,
+      "loss": 1.0039,
+      "step": 44620
+    },
+    {
+      "epoch": 13.35,
+      "grad_norm": 4.133788108825684,
+      "learning_rate": 1.245220257110632e-05,
+      "loss": 0.8164,
+      "step": 44625
+    },
+    {
+      "epoch": 13.35,
+      "grad_norm": 1.5097733736038208,
+      "learning_rate": 1.2447121337767121e-05,
+      "loss": 0.7921,
+      "step": 44630
+    },
+    {
+      "epoch": 13.35,
+      "grad_norm": 1.8934462070465088,
+      "learning_rate": 1.2442040797710358e-05,
+      "loss": 0.9432,
+      "step": 44635
+    },
+    {
+      "epoch": 13.36,
+      "grad_norm": 7.073252201080322,
+      "learning_rate": 1.2436960951216633e-05,
+      "loss": 0.9477,
+      "step": 44640
+    },
+    {
+      "epoch": 13.36,
+      "grad_norm": 2.9802677631378174,
+      "learning_rate": 1.2431881798566495e-05,
+      "loss": 0.9598,
+      "step": 44645
+    },
+    {
+      "epoch": 13.36,
+      "grad_norm": 1.559698224067688,
+      "learning_rate": 1.2426803340040461e-05,
+      "loss": 0.9934,
+      "step": 44650
+    },
+    {
+      "epoch": 13.36,
+      "grad_norm": 1.7870979309082031,
+      "learning_rate": 1.242172557591901e-05,
+      "loss": 1.0398,
+      "step": 44655
+    },
+    {
+      "epoch": 13.36,
+      "grad_norm": 4.0721755027771,
+      "learning_rate": 1.2416648506482589e-05,
+      "loss": 1.0168,
+      "step": 44660
+    },
+    {
+      "epoch": 13.36,
+      "grad_norm": 4.374423980712891,
+      "learning_rate": 1.2411572132011572e-05,
+      "loss": 0.9884,
+      "step": 44665
+    },
+    {
+      "epoch": 13.36,
+      "grad_norm": 3.6331615447998047,
+      "learning_rate": 1.2406496452786364e-05,
+      "loss": 0.9356,
+      "step": 44670
+    },
+    {
+      "epoch": 13.37,
+      "grad_norm": 1.985280156135559,
+      "learning_rate": 1.2401421469087246e-05,
+      "loss": 1.0527,
+      "step": 44675
+    },
+    {
+      "epoch": 13.37,
+      "grad_norm": 3.937107563018799,
+      "learning_rate": 1.2396347181194542e-05,
+      "loss": 1.0093,
+      "step": 44680
+    },
+    {
+      "epoch": 13.37,
+      "grad_norm": 1.2847137451171875,
+      "learning_rate": 1.2391273589388463e-05,
+      "loss": 1.065,
+      "step": 44685
+    },
+    {
+      "epoch": 13.37,
+      "grad_norm": 2.778695583343506,
+      "learning_rate": 1.238620069394926e-05,
+      "loss": 1.0256,
+      "step": 44690
+    },
+    {
+      "epoch": 13.37,
+      "grad_norm": 2.4905941486358643,
+      "learning_rate": 1.2381128495157071e-05,
+      "loss": 0.9014,
+      "step": 44695
+    },
+    {
+      "epoch": 13.37,
+      "grad_norm": 3.061601400375366,
+      "learning_rate": 1.2376056993292037e-05,
+      "loss": 0.9241,
+      "step": 44700
+    },
+    {
+      "epoch": 13.38,
+      "grad_norm": 5.156485557556152,
+      "learning_rate": 1.2370986188634253e-05,
+      "loss": 1.0604,
+      "step": 44705
+    },
+    {
+      "epoch": 13.38,
+      "grad_norm": 2.8037447929382324,
+      "learning_rate": 1.2365916081463777e-05,
+      "loss": 0.852,
+      "step": 44710
+    },
+    {
+      "epoch": 13.38,
+      "grad_norm": 2.281531572341919,
+      "learning_rate": 1.2360846672060622e-05,
+      "loss": 1.1074,
+      "step": 44715
+    },
+    {
+      "epoch": 13.38,
+      "grad_norm": 3.54512357711792,
+      "learning_rate": 1.2355777960704767e-05,
+      "loss": 0.9447,
+      "step": 44720
+    },
+    {
+      "epoch": 13.38,
+      "grad_norm": 2.5671660900115967,
+      "learning_rate": 1.2350709947676153e-05,
+      "loss": 1.0479,
+      "step": 44725
+    },
+    {
+      "epoch": 13.38,
+      "grad_norm": 1.9921631813049316,
+      "learning_rate": 1.2345642633254682e-05,
+      "loss": 0.9497,
+      "step": 44730
+    },
+    {
+      "epoch": 13.38,
+      "grad_norm": 1.9253196716308594,
+      "learning_rate": 1.2340576017720212e-05,
+      "loss": 1.0235,
+      "step": 44735
+    },
+    {
+      "epoch": 13.39,
+      "grad_norm": 5.337843418121338,
+      "learning_rate": 1.233551010135257e-05,
+      "loss": 1.1129,
+      "step": 44740
+    },
+    {
+      "epoch": 13.39,
+      "grad_norm": 3.605689764022827,
+      "learning_rate": 1.2330444884431544e-05,
+      "loss": 1.0846,
+      "step": 44745
+    },
+    {
+      "epoch": 13.39,
+      "grad_norm": 2.8609673976898193,
+      "learning_rate": 1.2325380367236875e-05,
+      "loss": 1.0635,
+      "step": 44750
+    },
+    {
+      "epoch": 13.39,
+      "grad_norm": 2.364123582839966,
+      "learning_rate": 1.2320316550048287e-05,
+      "loss": 0.951,
+      "step": 44755
+    },
+    {
+      "epoch": 13.39,
+      "grad_norm": 7.171827793121338,
+      "learning_rate": 1.2315253433145412e-05,
+      "loss": 0.8352,
+      "step": 44760
+    },
+    {
+      "epoch": 13.39,
+      "grad_norm": 3.1451687812805176,
+      "learning_rate": 1.2310191016807925e-05,
+      "loss": 0.9133,
+      "step": 44765
+    },
+    {
+      "epoch": 13.39,
+      "grad_norm": 2.174330711364746,
+      "learning_rate": 1.2305129301315382e-05,
+      "loss": 1.0346,
+      "step": 44770
+    },
+    {
+      "epoch": 13.4,
+      "grad_norm": 2.454254150390625,
+      "learning_rate": 1.230006828694737e-05,
+      "loss": 1.0938,
+      "step": 44775
+    },
+    {
+      "epoch": 13.4,
+      "grad_norm": 2.6113619804382324,
+      "learning_rate": 1.2295007973983366e-05,
+      "loss": 0.9711,
+      "step": 44780
+    },
+    {
+      "epoch": 13.4,
+      "grad_norm": 2.4465274810791016,
+      "learning_rate": 1.2289948362702883e-05,
+      "loss": 1.0361,
+      "step": 44785
+    },
+    {
+      "epoch": 13.4,
+      "grad_norm": 2.1108946800231934,
+      "learning_rate": 1.2284889453385334e-05,
+      "loss": 1.0811,
+      "step": 44790
+    },
+    {
+      "epoch": 13.4,
+      "grad_norm": 3.172182083129883,
+      "learning_rate": 1.2279831246310122e-05,
+      "loss": 1.0407,
+      "step": 44795
+    },
+    {
+      "epoch": 13.4,
+      "grad_norm": 1.988529086112976,
+      "learning_rate": 1.2274773741756607e-05,
+      "loss": 1.0839,
+      "step": 44800
+    },
+    {
+      "epoch": 13.41,
+      "grad_norm": 2.424704074859619,
+      "learning_rate": 1.2269716940004103e-05,
+      "loss": 1.0702,
+      "step": 44805
+    },
+    {
+      "epoch": 13.41,
+      "grad_norm": 1.8845196962356567,
+      "learning_rate": 1.2264660841331921e-05,
+      "loss": 0.9947,
+      "step": 44810
+    },
+    {
+      "epoch": 13.41,
+      "grad_norm": 1.3340157270431519,
+      "learning_rate": 1.2259605446019261e-05,
+      "loss": 1.1926,
+      "step": 44815
+    },
+    {
+      "epoch": 13.41,
+      "grad_norm": 1.2041699886322021,
+      "learning_rate": 1.225455075434537e-05,
+      "loss": 1.128,
+      "step": 44820
+    },
+    {
+      "epoch": 13.41,
+      "grad_norm": 2.0719640254974365,
+      "learning_rate": 1.2249496766589382e-05,
+      "loss": 1.0323,
+      "step": 44825
+    },
+    {
+      "epoch": 13.41,
+      "grad_norm": 2.7494218349456787,
+      "learning_rate": 1.2244443483030438e-05,
+      "loss": 1.1221,
+      "step": 44830
+    },
+    {
+      "epoch": 13.41,
+      "grad_norm": 1.4055849313735962,
+      "learning_rate": 1.2239390903947618e-05,
+      "loss": 1.0946,
+      "step": 44835
+    },
+    {
+      "epoch": 13.42,
+      "grad_norm": 3.2642478942871094,
+      "learning_rate": 1.2234339029619974e-05,
+      "loss": 1.0766,
+      "step": 44840
+    },
+    {
+      "epoch": 13.42,
+      "grad_norm": 1.8189183473587036,
+      "learning_rate": 1.2229287860326519e-05,
+      "loss": 0.8681,
+      "step": 44845
+    },
+    {
+      "epoch": 13.42,
+      "grad_norm": 2.4354348182678223,
+      "learning_rate": 1.222423739634622e-05,
+      "loss": 0.722,
+      "step": 44850
+    },
+    {
+      "epoch": 13.42,
+      "grad_norm": 1.813585638999939,
+      "learning_rate": 1.2219187637958007e-05,
+      "loss": 1.1683,
+      "step": 44855
+    },
+    {
+      "epoch": 13.42,
+      "grad_norm": 3.655863046646118,
+      "learning_rate": 1.221413858544079e-05,
+      "loss": 1.067,
+      "step": 44860
+    },
+    {
+      "epoch": 13.42,
+      "grad_norm": 8.485624313354492,
+      "learning_rate": 1.2209090239073386e-05,
+      "loss": 1.0451,
+      "step": 44865
+    },
+    {
+      "epoch": 13.42,
+      "grad_norm": 4.247042655944824,
+      "learning_rate": 1.2204042599134654e-05,
+      "loss": 1.1318,
+      "step": 44870
+    },
+    {
+      "epoch": 13.43,
+      "grad_norm": 3.260227680206299,
+      "learning_rate": 1.219899566590333e-05,
+      "loss": 0.9493,
+      "step": 44875
+    },
+    {
+      "epoch": 13.43,
+      "grad_norm": 2.963711977005005,
+      "learning_rate": 1.2193949439658187e-05,
+      "loss": 0.919,
+      "step": 44880
+    },
+    {
+      "epoch": 13.43,
+      "grad_norm": 3.6867597103118896,
+      "learning_rate": 1.2188903920677896e-05,
+      "loss": 1.0995,
+      "step": 44885
+    },
+    {
+      "epoch": 13.43,
+      "grad_norm": 2.1366159915924072,
+      "learning_rate": 1.2183859109241116e-05,
+      "loss": 0.9898,
+      "step": 44890
+    },
+    {
+      "epoch": 13.43,
+      "grad_norm": 4.271877288818359,
+      "learning_rate": 1.2178815005626493e-05,
+      "loss": 1.0278,
+      "step": 44895
+    },
+    {
+      "epoch": 13.43,
+      "grad_norm": 7.151296615600586,
+      "learning_rate": 1.2173771610112574e-05,
+      "loss": 1.0398,
+      "step": 44900
+    },
+    {
+      "epoch": 13.44,
+      "grad_norm": 2.853313684463501,
+      "learning_rate": 1.2168728922977934e-05,
+      "loss": 1.0188,
+      "step": 44905
+    },
+    {
+      "epoch": 13.44,
+      "grad_norm": 1.6056421995162964,
+      "learning_rate": 1.216368694450104e-05,
+      "loss": 0.8771,
+      "step": 44910
+    },
+    {
+      "epoch": 13.44,
+      "grad_norm": 1.6957907676696777,
+      "learning_rate": 1.2158645674960392e-05,
+      "loss": 0.9153,
+      "step": 44915
+    },
+    {
+      "epoch": 13.44,
+      "grad_norm": 13.22221851348877,
+      "learning_rate": 1.2153605114634387e-05,
+      "loss": 1.1427,
+      "step": 44920
+    },
+    {
+      "epoch": 13.44,
+      "grad_norm": 4.49719762802124,
+      "learning_rate": 1.2148565263801417e-05,
+      "loss": 1.0046,
+      "step": 44925
+    },
+    {
+      "epoch": 13.44,
+      "grad_norm": 8.903334617614746,
+      "learning_rate": 1.2143526122739832e-05,
+      "loss": 0.9307,
+      "step": 44930
+    },
+    {
+      "epoch": 13.44,
+      "grad_norm": 2.33970046043396,
+      "learning_rate": 1.2138487691727932e-05,
+      "loss": 1.0706,
+      "step": 44935
+    },
+    {
+      "epoch": 13.45,
+      "grad_norm": 2.2916619777679443,
+      "learning_rate": 1.2133449971043992e-05,
+      "loss": 1.002,
+      "step": 44940
+    },
+    {
+      "epoch": 13.45,
+      "grad_norm": 3.4537885189056396,
+      "learning_rate": 1.2128412960966231e-05,
+      "loss": 1.0616,
+      "step": 44945
+    },
+    {
+      "epoch": 13.45,
+      "grad_norm": 2.979152202606201,
+      "learning_rate": 1.2123376661772848e-05,
+      "loss": 0.9152,
+      "step": 44950
+    },
+    {
+      "epoch": 13.45,
+      "grad_norm": 1.5841301679611206,
+      "learning_rate": 1.2118341073741993e-05,
+      "loss": 0.9443,
+      "step": 44955
+    },
+    {
+      "epoch": 13.45,
+      "grad_norm": 8.155377388000488,
+      "learning_rate": 1.2113306197151754e-05,
+      "loss": 0.9218,
+      "step": 44960
+    },
+    {
+      "epoch": 13.45,
+      "grad_norm": 3.3958740234375,
+      "learning_rate": 1.2108272032280226e-05,
+      "loss": 0.9962,
+      "step": 44965
+    },
+    {
+      "epoch": 13.45,
+      "grad_norm": 3.104916572570801,
+      "learning_rate": 1.2103238579405437e-05,
+      "loss": 1.0308,
+      "step": 44970
+    },
+    {
+      "epoch": 13.46,
+      "grad_norm": 1.5764812231063843,
+      "learning_rate": 1.2098205838805373e-05,
+      "loss": 0.9052,
+      "step": 44975
+    },
+    {
+      "epoch": 13.46,
+      "grad_norm": 1.8205498456954956,
+      "learning_rate": 1.2093173810757991e-05,
+      "loss": 1.1331,
+      "step": 44980
+    },
+    {
+      "epoch": 13.46,
+      "grad_norm": 3.6816673278808594,
+      "learning_rate": 1.2088142495541205e-05,
+      "loss": 0.9301,
+      "step": 44985
+    },
+    {
+      "epoch": 13.46,
+      "grad_norm": 3.829503059387207,
+      "learning_rate": 1.2083111893432893e-05,
+      "loss": 1.0146,
+      "step": 44990
+    },
+    {
+      "epoch": 13.46,
+      "grad_norm": 2.2091660499572754,
+      "learning_rate": 1.207808200471087e-05,
+      "loss": 0.9556,
+      "step": 44995
+    },
+    {
+      "epoch": 13.46,
+      "grad_norm": 4.129636287689209,
+      "learning_rate": 1.2074058607558082e-05,
+      "loss": 1.0866,
+      "step": 45000
+    },
+    {
+      "epoch": 13.46,
+      "grad_norm": 3.674562454223633,
+      "learning_rate": 1.206903000363143e-05,
+      "loss": 1.1933,
+      "step": 45005
+    },
+    {
+      "epoch": 13.47,
+      "grad_norm": 3.4031946659088135,
+      "learning_rate": 1.2064002113868822e-05,
+      "loss": 1.0472,
+      "step": 45010
+    },
+    {
+      "epoch": 13.47,
+      "grad_norm": 4.102842330932617,
+      "learning_rate": 1.2058974938547913e-05,
+      "loss": 1.0977,
+      "step": 45015
+    },
+    {
+      "epoch": 13.47,
+      "grad_norm": 1.0470472574234009,
+      "learning_rate": 1.205394847794638e-05,
+      "loss": 1.0491,
+      "step": 45020
+    },
+    {
+      "epoch": 13.47,
+      "grad_norm": 4.4615702629089355,
+      "learning_rate": 1.2048922732341802e-05,
+      "loss": 1.0362,
+      "step": 45025
+    },
+    {
+      "epoch": 13.47,
+      "grad_norm": 1.8083804845809937,
+      "learning_rate": 1.2043897702011777e-05,
+      "loss": 0.9582,
+      "step": 45030
+    },
+    {
+      "epoch": 13.47,
+      "grad_norm": 1.3587656021118164,
+      "learning_rate": 1.2038873387233806e-05,
+      "loss": 1.0904,
+      "step": 45035
+    },
+    {
+      "epoch": 13.48,
+      "grad_norm": 9.736397743225098,
+      "learning_rate": 1.2033849788285387e-05,
+      "loss": 1.0481,
+      "step": 45040
+    },
+    {
+      "epoch": 13.48,
+      "grad_norm": 1.1890517473220825,
+      "learning_rate": 1.2028826905443965e-05,
+      "loss": 1.0538,
+      "step": 45045
+    },
+    {
+      "epoch": 13.48,
+      "grad_norm": 1.4753776788711548,
+      "learning_rate": 1.2023804738986948e-05,
+      "loss": 1.0234,
+      "step": 45050
+    },
+    {
+      "epoch": 13.48,
+      "grad_norm": 2.6727852821350098,
+      "learning_rate": 1.2018783289191713e-05,
+      "loss": 1.0864,
+      "step": 45055
+    },
+    {
+      "epoch": 13.48,
+      "grad_norm": 3.8187108039855957,
+      "learning_rate": 1.2013762556335578e-05,
+      "loss": 1.0876,
+      "step": 45060
+    },
+    {
+      "epoch": 13.48,
+      "grad_norm": 2.2639684677124023,
+      "learning_rate": 1.2008742540695842e-05,
+      "loss": 1.1417,
+      "step": 45065
+    },
+    {
+      "epoch": 13.48,
+      "grad_norm": 1.7074878215789795,
+      "learning_rate": 1.2003723242549749e-05,
+      "loss": 0.9508,
+      "step": 45070
+    },
+    {
+      "epoch": 13.49,
+      "grad_norm": 3.0342020988464355,
+      "learning_rate": 1.1998704662174515e-05,
+      "loss": 0.8973,
+      "step": 45075
+    },
+    {
+      "epoch": 13.49,
+      "grad_norm": 2.464416980743408,
+      "learning_rate": 1.1993686799847307e-05,
+      "loss": 1.093,
+      "step": 45080
+    },
+    {
+      "epoch": 13.49,
+      "grad_norm": 2.7546801567077637,
+      "learning_rate": 1.1988669655845259e-05,
+      "loss": 1.1162,
+      "step": 45085
+    },
+    {
+      "epoch": 13.49,
+      "grad_norm": 7.984701156616211,
+      "learning_rate": 1.198365323044546e-05,
+      "loss": 0.9271,
+      "step": 45090
+    },
+    {
+      "epoch": 13.49,
+      "grad_norm": 2.5185933113098145,
+      "learning_rate": 1.1978637523924962e-05,
+      "loss": 0.9088,
+      "step": 45095
+    },
+    {
+      "epoch": 13.49,
+      "grad_norm": 5.400019645690918,
+      "learning_rate": 1.1973622536560783e-05,
+      "loss": 1.0487,
+      "step": 45100
+    },
+    {
+      "epoch": 13.49,
+      "grad_norm": 1.2262287139892578,
+      "learning_rate": 1.1968608268629897e-05,
+      "loss": 0.9563,
+      "step": 45105
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 1.9510408639907837,
+      "learning_rate": 1.1963594720409213e-05,
+      "loss": 0.9829,
+      "step": 45110
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 3.9547598361968994,
+      "learning_rate": 1.1958581892175664e-05,
+      "loss": 1.0245,
+      "step": 45115
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 2.1796655654907227,
+      "learning_rate": 1.1953569784206056e-05,
+      "loss": 0.8948,
+      "step": 45120
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 1.8676890134811401,
+      "learning_rate": 1.194855839677725e-05,
+      "loss": 0.9237,
+      "step": 45125
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 2.9657278060913086,
+      "learning_rate": 1.194354773016599e-05,
+      "loss": 0.9703,
+      "step": 45130
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 3.7803890705108643,
+      "learning_rate": 1.1938537784649015e-05,
+      "loss": 0.941,
+      "step": 45135
+    },
+    {
+      "epoch": 13.51,
+      "grad_norm": 4.787195205688477,
+      "learning_rate": 1.1933528560503021e-05,
+      "loss": 0.854,
+      "step": 45140
+    },
+    {
+      "epoch": 13.51,
+      "grad_norm": 2.3021395206451416,
+      "learning_rate": 1.1928520058004666e-05,
+      "loss": 0.9937,
+      "step": 45145
+    },
+    {
+      "epoch": 13.51,
+      "grad_norm": 3.430438995361328,
+      "learning_rate": 1.192351227743056e-05,
+      "loss": 1.0839,
+      "step": 45150
+    },
+    {
+      "epoch": 13.51,
+      "grad_norm": 3.5404064655303955,
+      "learning_rate": 1.1918505219057265e-05,
+      "loss": 1.0443,
+      "step": 45155
+    },
+    {
+      "epoch": 13.51,
+      "grad_norm": 1.529388427734375,
+      "learning_rate": 1.1913498883161354e-05,
+      "loss": 1.0485,
+      "step": 45160
+    },
+    {
+      "epoch": 13.51,
+      "grad_norm": 3.046905755996704,
+      "learning_rate": 1.1908493270019283e-05,
+      "loss": 1.1611,
+      "step": 45165
+    },
+    {
+      "epoch": 13.51,
+      "grad_norm": 2.4746837615966797,
+      "learning_rate": 1.1903488379907524e-05,
+      "loss": 0.9898,
+      "step": 45170
+    },
+    {
+      "epoch": 13.52,
+      "grad_norm": 3.2525274753570557,
+      "learning_rate": 1.1898484213102487e-05,
+      "loss": 1.036,
+      "step": 45175
+    },
+    {
+      "epoch": 13.52,
+      "grad_norm": 2.007542371749878,
+      "learning_rate": 1.1893480769880549e-05,
+      "loss": 1.0232,
+      "step": 45180
+    },
+    {
+      "epoch": 13.52,
+      "grad_norm": 3.283017873764038,
+      "learning_rate": 1.1888478050518046e-05,
+      "loss": 1.0205,
+      "step": 45185
+    },
+    {
+      "epoch": 13.52,
+      "grad_norm": 3.2330360412597656,
+      "learning_rate": 1.188347605529127e-05,
+      "loss": 0.9873,
+      "step": 45190
+    },
+    {
+      "epoch": 13.52,
+      "grad_norm": 3.1978647708892822,
+      "learning_rate": 1.1878474784476478e-05,
+      "loss": 1.0942,
+      "step": 45195
+    },
+    {
+      "epoch": 13.52,
+      "grad_norm": 3.234929084777832,
+      "learning_rate": 1.1873474238349894e-05,
+      "loss": 1.181,
+      "step": 45200
+    },
+    {
+      "epoch": 13.52,
+      "grad_norm": 3.867767333984375,
+      "learning_rate": 1.1868474417187666e-05,
+      "loss": 0.9687,
+      "step": 45205
+    },
+    {
+      "epoch": 13.53,
+      "grad_norm": 2.2236828804016113,
+      "learning_rate": 1.1863475321265962e-05,
+      "loss": 1.0287,
+      "step": 45210
+    },
+    {
+      "epoch": 13.53,
+      "grad_norm": 3.1465940475463867,
+      "learning_rate": 1.1858476950860848e-05,
+      "loss": 0.9651,
+      "step": 45215
+    },
+    {
+      "epoch": 13.53,
+      "grad_norm": 2.4534783363342285,
+      "learning_rate": 1.1853479306248404e-05,
+      "loss": 1.0121,
+      "step": 45220
+    },
+    {
+      "epoch": 13.53,
+      "grad_norm": 6.127681732177734,
+      "learning_rate": 1.1848482387704621e-05,
+      "loss": 1.1017,
+      "step": 45225
+    },
+    {
+      "epoch": 13.53,
+      "grad_norm": 2.688446283340454,
+      "learning_rate": 1.184348619550549e-05,
+      "loss": 1.0992,
+      "step": 45230
+    },
+    {
+      "epoch": 13.53,
+      "grad_norm": 2.3378570079803467,
+      "learning_rate": 1.1838490729926952e-05,
+      "loss": 1.0143,
+      "step": 45235
+    },
+    {
+      "epoch": 13.54,
+      "grad_norm": 2.69551944732666,
+      "learning_rate": 1.1833495991244873e-05,
+      "loss": 0.999,
+      "step": 45240
+    },
+    {
+      "epoch": 13.54,
+      "grad_norm": 3.029266357421875,
+      "learning_rate": 1.182850197973514e-05,
+      "loss": 1.0098,
+      "step": 45245
+    },
+    {
+      "epoch": 13.54,
+      "grad_norm": 8.008578300476074,
+      "learning_rate": 1.1823508695673535e-05,
+      "loss": 1.0435,
+      "step": 45250
+    },
+    {
+      "epoch": 13.54,
+      "grad_norm": 1.313415765762329,
+      "learning_rate": 1.1818516139335867e-05,
+      "loss": 1.2993,
+      "step": 45255
+    },
+    {
+      "epoch": 13.54,
+      "grad_norm": 4.881174087524414,
+      "learning_rate": 1.1813524310997842e-05,
+      "loss": 0.9305,
+      "step": 45260
+    },
+    {
+      "epoch": 13.54,
+      "grad_norm": 2.0278892517089844,
+      "learning_rate": 1.1808533210935164e-05,
+      "loss": 0.9142,
+      "step": 45265
+    },
+    {
+      "epoch": 13.54,
+      "grad_norm": 5.330128192901611,
+      "learning_rate": 1.180354283942348e-05,
+      "loss": 1.1023,
+      "step": 45270
+    },
+    {
+      "epoch": 13.55,
+      "grad_norm": 1.4603615999221802,
+      "learning_rate": 1.1798553196738415e-05,
+      "loss": 1.04,
+      "step": 45275
+    },
+    {
+      "epoch": 13.55,
+      "grad_norm": 2.043243408203125,
+      "learning_rate": 1.1793564283155532e-05,
+      "loss": 0.9853,
+      "step": 45280
+    },
+    {
+      "epoch": 13.55,
+      "grad_norm": 3.6421613693237305,
+      "learning_rate": 1.1788576098950365e-05,
+      "loss": 0.9557,
+      "step": 45285
+    },
+    {
+      "epoch": 13.55,
+      "grad_norm": 2.091878890991211,
+      "learning_rate": 1.1783588644398408e-05,
+      "loss": 1.1519,
+      "step": 45290
+    },
+    {
+      "epoch": 13.55,
+      "grad_norm": 1.9971786737442017,
+      "learning_rate": 1.1778601919775122e-05,
+      "loss": 1.0033,
+      "step": 45295
+    },
+    {
+      "epoch": 13.55,
+      "grad_norm": 0.9933714270591736,
+      "learning_rate": 1.1773615925355891e-05,
+      "loss": 0.9982,
+      "step": 45300
+    },
+    {
+      "epoch": 13.55,
+      "grad_norm": 5.131785869598389,
+      "learning_rate": 1.1768630661416114e-05,
+      "loss": 1.0143,
+      "step": 45305
+    },
+    {
+      "epoch": 13.56,
+      "grad_norm": 2.784761905670166,
+      "learning_rate": 1.176364612823111e-05,
+      "loss": 0.8749,
+      "step": 45310
+    },
+    {
+      "epoch": 13.56,
+      "grad_norm": 1.629541039466858,
+      "learning_rate": 1.1758662326076176e-05,
+      "loss": 1.1158,
+      "step": 45315
+    },
+    {
+      "epoch": 13.56,
+      "grad_norm": 3.978738784790039,
+      "learning_rate": 1.1753679255226555e-05,
+      "loss": 0.7877,
+      "step": 45320
+    },
+    {
+      "epoch": 13.56,
+      "grad_norm": 2.3555569648742676,
+      "learning_rate": 1.174869691595746e-05,
+      "loss": 1.0907,
+      "step": 45325
+    },
+    {
+      "epoch": 13.56,
+      "grad_norm": 2.2074787616729736,
+      "learning_rate": 1.174371530854407e-05,
+      "loss": 0.9258,
+      "step": 45330
+    },
+    {
+      "epoch": 13.56,
+      "grad_norm": 3.912524461746216,
+      "learning_rate": 1.1738734433261486e-05,
+      "loss": 0.9214,
+      "step": 45335
+    },
+    {
+      "epoch": 13.57,
+      "grad_norm": 3.4181976318359375,
+      "learning_rate": 1.1733754290384833e-05,
+      "loss": 0.9704,
+      "step": 45340
+    },
+    {
+      "epoch": 13.57,
+      "grad_norm": 2.531336545944214,
+      "learning_rate": 1.1728774880189123e-05,
+      "loss": 1.087,
+      "step": 45345
+    },
+    {
+      "epoch": 13.57,
+      "grad_norm": 3.3904941082000732,
+      "learning_rate": 1.17237962029494e-05,
+      "loss": 0.9672,
+      "step": 45350
+    },
+    {
+      "epoch": 13.57,
+      "grad_norm": 4.165085792541504,
+      "learning_rate": 1.1718818258940596e-05,
+      "loss": 0.9494,
+      "step": 45355
+    },
+    {
+      "epoch": 13.57,
+      "grad_norm": 3.3338541984558105,
+      "learning_rate": 1.171384104843767e-05,
+      "loss": 1.0561,
+      "step": 45360
+    },
+    {
+      "epoch": 13.57,
+      "grad_norm": 2.241774797439575,
+      "learning_rate": 1.1708864571715486e-05,
+      "loss": 1.0043,
+      "step": 45365
+    },
+    {
+      "epoch": 13.57,
+      "grad_norm": 3.3441967964172363,
+      "learning_rate": 1.1703888829048895e-05,
+      "loss": 1.0327,
+      "step": 45370
+    },
+    {
+      "epoch": 13.58,
+      "grad_norm": 3.085231065750122,
+      "learning_rate": 1.1698913820712704e-05,
+      "loss": 0.7949,
+      "step": 45375
+    },
+    {
+      "epoch": 13.58,
+      "grad_norm": 3.700368881225586,
+      "learning_rate": 1.1693939546981678e-05,
+      "loss": 1.0952,
+      "step": 45380
+    },
+    {
+      "epoch": 13.58,
+      "grad_norm": 3.532536745071411,
+      "learning_rate": 1.1688966008130539e-05,
+      "loss": 0.9898,
+      "step": 45385
+    },
+    {
+      "epoch": 13.58,
+      "grad_norm": 3.1939263343811035,
+      "learning_rate": 1.1683993204433971e-05,
+      "loss": 1.0335,
+      "step": 45390
+    },
+    {
+      "epoch": 13.58,
+      "grad_norm": 2.641188144683838,
+      "learning_rate": 1.1679021136166618e-05,
+      "loss": 0.9966,
+      "step": 45395
+    },
+    {
+      "epoch": 13.58,
+      "grad_norm": 12.577829360961914,
+      "learning_rate": 1.1674049803603079e-05,
+      "loss": 0.9728,
+      "step": 45400
+    },
+    {
+      "epoch": 13.58,
+      "grad_norm": 5.408492565155029,
+      "learning_rate": 1.166907920701792e-05,
+      "loss": 1.183,
+      "step": 45405
+    },
+    {
+      "epoch": 13.59,
+      "grad_norm": 15.173827171325684,
+      "learning_rate": 1.166410934668566e-05,
+      "loss": 1.0365,
+      "step": 45410
+    },
+    {
+      "epoch": 13.59,
+      "grad_norm": 1.8098371028900146,
+      "learning_rate": 1.1659140222880777e-05,
+      "loss": 1.0531,
+      "step": 45415
+    },
+    {
+      "epoch": 13.59,
+      "grad_norm": 3.362779378890991,
+      "learning_rate": 1.1654171835877715e-05,
+      "loss": 1.0122,
+      "step": 45420
+    },
+    {
+      "epoch": 13.59,
+      "grad_norm": 4.117199420928955,
+      "learning_rate": 1.1649204185950877e-05,
+      "loss": 0.7663,
+      "step": 45425
+    },
+    {
+      "epoch": 13.59,
+      "grad_norm": 4.3105621337890625,
+      "learning_rate": 1.1644237273374595e-05,
+      "loss": 1.0172,
+      "step": 45430
+    },
+    {
+      "epoch": 13.59,
+      "grad_norm": 2.2465403079986572,
+      "learning_rate": 1.1639271098423227e-05,
+      "loss": 0.9232,
+      "step": 45435
+    },
+    {
+      "epoch": 13.6,
+      "grad_norm": 1.8414210081100464,
+      "learning_rate": 1.163430566137101e-05,
+      "loss": 0.915,
+      "step": 45440
+    },
+    {
+      "epoch": 13.6,
+      "grad_norm": 2.6317269802093506,
+      "learning_rate": 1.1629340962492217e-05,
+      "loss": 0.8813,
+      "step": 45445
+    },
+    {
+      "epoch": 13.6,
+      "grad_norm": 2.4604814052581787,
+      "learning_rate": 1.1624377002061004e-05,
+      "loss": 1.111,
+      "step": 45450
+    },
+    {
+      "epoch": 13.6,
+      "grad_norm": 2.286057710647583,
+      "learning_rate": 1.1619413780351566e-05,
+      "loss": 1.0006,
+      "step": 45455
+    },
+    {
+      "epoch": 13.6,
+      "grad_norm": 3.36887264251709,
+      "learning_rate": 1.1614451297637985e-05,
+      "loss": 1.1177,
+      "step": 45460
+    },
+    {
+      "epoch": 13.6,
+      "grad_norm": 3.5638434886932373,
+      "learning_rate": 1.1609489554194348e-05,
+      "loss": 0.9992,
+      "step": 45465
+    },
+    {
+      "epoch": 13.6,
+      "grad_norm": 1.2167627811431885,
+      "learning_rate": 1.1604528550294685e-05,
+      "loss": 1.1973,
+      "step": 45470
+    },
+    {
+      "epoch": 13.61,
+      "grad_norm": 2.9576528072357178,
+      "learning_rate": 1.1599568286212984e-05,
+      "loss": 0.9278,
+      "step": 45475
+    },
+    {
+      "epoch": 13.61,
+      "grad_norm": 3.1728708744049072,
+      "learning_rate": 1.15946087622232e-05,
+      "loss": 1.1047,
+      "step": 45480
+    },
+    {
+      "epoch": 13.61,
+      "grad_norm": 5.425692558288574,
+      "learning_rate": 1.1589649978599237e-05,
+      "loss": 0.9143,
+      "step": 45485
+    },
+    {
+      "epoch": 13.61,
+      "grad_norm": 2.459667921066284,
+      "learning_rate": 1.158469193561497e-05,
+      "loss": 0.8807,
+      "step": 45490
+    },
+    {
+      "epoch": 13.61,
+      "grad_norm": 2.123094320297241,
+      "learning_rate": 1.157973463354422e-05,
+      "loss": 0.945,
+      "step": 45495
+    },
+    {
+      "epoch": 13.61,
+      "grad_norm": 2.2343719005584717,
+      "learning_rate": 1.1574778072660778e-05,
+      "loss": 1.0458,
+      "step": 45500
+    },
+    {
+      "epoch": 13.61,
+      "grad_norm": 6.470086097717285,
+      "learning_rate": 1.1569822253238385e-05,
+      "loss": 1.0621,
+      "step": 45505
+    },
+    {
+      "epoch": 13.62,
+      "grad_norm": 2.044508934020996,
+      "learning_rate": 1.1564867175550753e-05,
+      "loss": 1.0052,
+      "step": 45510
+    },
+    {
+      "epoch": 13.62,
+      "grad_norm": 2.4018311500549316,
+      "learning_rate": 1.1559912839871542e-05,
+      "loss": 1.0759,
+      "step": 45515
+    },
+    {
+      "epoch": 13.62,
+      "grad_norm": 3.800114154815674,
+      "learning_rate": 1.1554959246474382e-05,
+      "loss": 0.9216,
+      "step": 45520
+    },
+    {
+      "epoch": 13.62,
+      "grad_norm": 2.723524808883667,
+      "learning_rate": 1.155000639563283e-05,
+      "loss": 0.9357,
+      "step": 45525
+    },
+    {
+      "epoch": 13.62,
+      "grad_norm": 4.594392776489258,
+      "learning_rate": 1.154505428762046e-05,
+      "loss": 1.1448,
+      "step": 45530
+    },
+    {
+      "epoch": 13.62,
+      "grad_norm": 2.8049912452697754,
+      "learning_rate": 1.154010292271074e-05,
+      "loss": 1.1228,
+      "step": 45535
+    },
+    {
+      "epoch": 13.63,
+      "grad_norm": 2.90828013420105,
+      "learning_rate": 1.1535152301177165e-05,
+      "loss": 1.131,
+      "step": 45540
+    },
+    {
+      "epoch": 13.63,
+      "grad_norm": 2.163752317428589,
+      "learning_rate": 1.1530202423293113e-05,
+      "loss": 0.987,
+      "step": 45545
+    },
+    {
+      "epoch": 13.63,
+      "grad_norm": 2.9470038414001465,
+      "learning_rate": 1.1525253289331997e-05,
+      "loss": 0.8447,
+      "step": 45550
+    },
+    {
+      "epoch": 13.63,
+      "grad_norm": 5.131521224975586,
+      "learning_rate": 1.1520304899567128e-05,
+      "loss": 1.112,
+      "step": 45555
+    },
+    {
+      "epoch": 13.63,
+      "grad_norm": 2.858603000640869,
+      "learning_rate": 1.1515357254271805e-05,
+      "loss": 1.0796,
+      "step": 45560
+    },
+    {
+      "epoch": 13.63,
+      "grad_norm": 5.054248809814453,
+      "learning_rate": 1.1510410353719287e-05,
+      "loss": 0.8444,
+      "step": 45565
+    },
+    {
+      "epoch": 13.63,
+      "grad_norm": 1.9235126972198486,
+      "learning_rate": 1.1505464198182784e-05,
+      "loss": 1.1076,
+      "step": 45570
+    },
+    {
+      "epoch": 13.64,
+      "grad_norm": 3.045146942138672,
+      "learning_rate": 1.1500518787935466e-05,
+      "loss": 0.9475,
+      "step": 45575
+    },
+    {
+      "epoch": 13.64,
+      "grad_norm": 4.526063442230225,
+      "learning_rate": 1.1495574123250462e-05,
+      "loss": 0.8281,
+      "step": 45580
+    },
+    {
+      "epoch": 13.64,
+      "grad_norm": 2.535439968109131,
+      "learning_rate": 1.1490630204400863e-05,
+      "loss": 1.0387,
+      "step": 45585
+    },
+    {
+      "epoch": 13.64,
+      "grad_norm": 3.0743353366851807,
+      "learning_rate": 1.1485687031659717e-05,
+      "loss": 1.0645,
+      "step": 45590
+    },
+    {
+      "epoch": 13.64,
+      "grad_norm": 3.6751158237457275,
+      "learning_rate": 1.1480744605300026e-05,
+      "loss": 0.9101,
+      "step": 45595
+    },
+    {
+      "epoch": 13.64,
+      "grad_norm": 3.3130288124084473,
+      "learning_rate": 1.1475802925594762e-05,
+      "loss": 0.8764,
+      "step": 45600
+    },
+    {
+      "epoch": 13.64,
+      "grad_norm": 8.780059814453125,
+      "learning_rate": 1.1470861992816839e-05,
+      "loss": 1.0427,
+      "step": 45605
+    },
+    {
+      "epoch": 13.65,
+      "grad_norm": 3.3624203205108643,
+      "learning_rate": 1.146592180723915e-05,
+      "loss": 0.9371,
+      "step": 45610
+    },
+    {
+      "epoch": 13.65,
+      "grad_norm": 2.7508459091186523,
+      "learning_rate": 1.1460982369134541e-05,
+      "loss": 0.8532,
+      "step": 45615
+    },
+    {
+      "epoch": 13.65,
+      "grad_norm": 1.148845911026001,
+      "learning_rate": 1.1456043678775783e-05,
+      "loss": 0.992,
+      "step": 45620
+    },
+    {
+      "epoch": 13.65,
+      "grad_norm": 3.8979814052581787,
+      "learning_rate": 1.1451105736435672e-05,
+      "loss": 1.1124,
+      "step": 45625
+    },
+    {
+      "epoch": 13.65,
+      "grad_norm": 6.8672990798950195,
+      "learning_rate": 1.144616854238689e-05,
+      "loss": 0.8867,
+      "step": 45630
+    },
+    {
+      "epoch": 13.65,
+      "grad_norm": 3.3490543365478516,
+      "learning_rate": 1.1441232096902149e-05,
+      "loss": 1.044,
+      "step": 45635
+    },
+    {
+      "epoch": 13.65,
+      "grad_norm": 2.1857025623321533,
+      "learning_rate": 1.1436296400254049e-05,
+      "loss": 0.8137,
+      "step": 45640
+    },
+    {
+      "epoch": 13.66,
+      "grad_norm": 3.0062873363494873,
+      "learning_rate": 1.1431361452715218e-05,
+      "loss": 0.9847,
+      "step": 45645
+    },
+    {
+      "epoch": 13.66,
+      "grad_norm": 3.707160472869873,
+      "learning_rate": 1.1426427254558182e-05,
+      "loss": 1.1328,
+      "step": 45650
+    },
+    {
+      "epoch": 13.66,
+      "grad_norm": 3.175178289413452,
+      "learning_rate": 1.1421493806055455e-05,
+      "loss": 1.0651,
+      "step": 45655
+    },
+    {
+      "epoch": 13.66,
+      "grad_norm": 4.4922776222229,
+      "learning_rate": 1.1416561107479514e-05,
+      "loss": 1.1197,
+      "step": 45660
+    },
+    {
+      "epoch": 13.66,
+      "grad_norm": 2.236589193344116,
+      "learning_rate": 1.1411629159102785e-05,
+      "loss": 1.0603,
+      "step": 45665
+    },
+    {
+      "epoch": 13.66,
+      "grad_norm": 1.2973885536193848,
+      "learning_rate": 1.1406697961197654e-05,
+      "loss": 0.9378,
+      "step": 45670
+    },
+    {
+      "epoch": 13.67,
+      "grad_norm": 2.1828157901763916,
+      "learning_rate": 1.1401767514036463e-05,
+      "loss": 0.94,
+      "step": 45675
+    },
+    {
+      "epoch": 13.67,
+      "grad_norm": 4.644625186920166,
+      "learning_rate": 1.1396837817891518e-05,
+      "loss": 0.9499,
+      "step": 45680
+    },
+    {
+      "epoch": 13.67,
+      "grad_norm": 2.0859460830688477,
+      "learning_rate": 1.1391908873035082e-05,
+      "loss": 0.9002,
+      "step": 45685
+    },
+    {
+      "epoch": 13.67,
+      "grad_norm": 7.057641983032227,
+      "learning_rate": 1.1386980679739373e-05,
+      "loss": 0.917,
+      "step": 45690
+    },
+    {
+      "epoch": 13.67,
+      "grad_norm": 3.8702352046966553,
+      "learning_rate": 1.1382053238276569e-05,
+      "loss": 0.9209,
+      "step": 45695
+    },
+    {
+      "epoch": 13.67,
+      "grad_norm": 1.8291438817977905,
+      "learning_rate": 1.1377126548918811e-05,
+      "loss": 1.02,
+      "step": 45700
+    },
+    {
+      "epoch": 13.67,
+      "grad_norm": 6.240175247192383,
+      "learning_rate": 1.1372200611938196e-05,
+      "loss": 0.9408,
+      "step": 45705
+    },
+    {
+      "epoch": 13.68,
+      "grad_norm": 8.977832794189453,
+      "learning_rate": 1.1367275427606774e-05,
+      "loss": 0.8573,
+      "step": 45710
+    },
+    {
+      "epoch": 13.68,
+      "grad_norm": 2.6074182987213135,
+      "learning_rate": 1.1362350996196559e-05,
+      "loss": 0.9426,
+      "step": 45715
+    },
+    {
+      "epoch": 13.68,
+      "grad_norm": 3.5321543216705322,
+      "learning_rate": 1.1357427317979532e-05,
+      "loss": 1.0316,
+      "step": 45720
+    },
+    {
+      "epoch": 13.68,
+      "grad_norm": 1.716353178024292,
+      "learning_rate": 1.1352504393227598e-05,
+      "loss": 1.0807,
+      "step": 45725
+    },
+    {
+      "epoch": 13.68,
+      "grad_norm": 4.071742534637451,
+      "learning_rate": 1.1347582222212677e-05,
+      "loss": 0.951,
+      "step": 45730
+    },
+    {
+      "epoch": 13.68,
+      "grad_norm": 2.598299741744995,
+      "learning_rate": 1.1342660805206578e-05,
+      "loss": 0.9232,
+      "step": 45735
+    },
+    {
+      "epoch": 13.68,
+      "grad_norm": 3.115583658218384,
+      "learning_rate": 1.1337740142481148e-05,
+      "loss": 1.1328,
+      "step": 45740
+    },
+    {
+      "epoch": 13.69,
+      "grad_norm": 4.597363471984863,
+      "learning_rate": 1.1332820234308123e-05,
+      "loss": 0.9854,
+      "step": 45745
+    },
+    {
+      "epoch": 13.69,
+      "grad_norm": 4.575268745422363,
+      "learning_rate": 1.1327901080959224e-05,
+      "loss": 0.9848,
+      "step": 45750
+    },
+    {
+      "epoch": 13.69,
+      "grad_norm": 4.154250144958496,
+      "learning_rate": 1.132298268270614e-05,
+      "loss": 0.877,
+      "step": 45755
+    },
+    {
+      "epoch": 13.69,
+      "grad_norm": 2.12250018119812,
+      "learning_rate": 1.1318065039820505e-05,
+      "loss": 0.8998,
+      "step": 45760
+    },
+    {
+      "epoch": 13.69,
+      "grad_norm": 1.1914949417114258,
+      "learning_rate": 1.1313148152573915e-05,
+      "loss": 0.975,
+      "step": 45765
+    },
+    {
+      "epoch": 13.69,
+      "grad_norm": 2.145120143890381,
+      "learning_rate": 1.130823202123793e-05,
+      "loss": 1.1313,
+      "step": 45770
+    },
+    {
+      "epoch": 13.7,
+      "grad_norm": 6.92742395401001,
+      "learning_rate": 1.1303316646084055e-05,
+      "loss": 1.184,
+      "step": 45775
+    },
+    {
+      "epoch": 13.7,
+      "grad_norm": 4.461109638214111,
+      "learning_rate": 1.1298402027383768e-05,
+      "loss": 1.0285,
+      "step": 45780
+    },
+    {
+      "epoch": 13.7,
+      "grad_norm": 3.108293294906616,
+      "learning_rate": 1.1293488165408492e-05,
+      "loss": 1.0309,
+      "step": 45785
+    },
+    {
+      "epoch": 13.7,
+      "grad_norm": 2.175156593322754,
+      "learning_rate": 1.1288575060429618e-05,
+      "loss": 1.0424,
+      "step": 45790
+    },
+    {
+      "epoch": 13.7,
+      "grad_norm": 2.99977707862854,
+      "learning_rate": 1.1283662712718493e-05,
+      "loss": 0.8261,
+      "step": 45795
+    },
+    {
+      "epoch": 13.7,
+      "grad_norm": 3.166217565536499,
+      "learning_rate": 1.1278751122546417e-05,
+      "loss": 0.9703,
+      "step": 45800
+    },
+    {
+      "epoch": 13.7,
+      "grad_norm": 4.1235032081604,
+      "learning_rate": 1.1273840290184653e-05,
+      "loss": 1.0018,
+      "step": 45805
+    },
+    {
+      "epoch": 13.71,
+      "grad_norm": 2.862684726715088,
+      "learning_rate": 1.1268930215904424e-05,
+      "loss": 1.0977,
+      "step": 45810
+    },
+    {
+      "epoch": 13.71,
+      "grad_norm": 3.851482629776001,
+      "learning_rate": 1.1264020899976918e-05,
+      "loss": 0.9958,
+      "step": 45815
+    },
+    {
+      "epoch": 13.71,
+      "grad_norm": 4.1922197341918945,
+      "learning_rate": 1.125911234267324e-05,
+      "loss": 1.0227,
+      "step": 45820
+    },
+    {
+      "epoch": 13.71,
+      "grad_norm": 2.539138078689575,
+      "learning_rate": 1.1254204544264521e-05,
+      "loss": 1.1639,
+      "step": 45825
+    },
+    {
+      "epoch": 13.71,
+      "grad_norm": 5.336416721343994,
+      "learning_rate": 1.1249297505021778e-05,
+      "loss": 0.9809,
+      "step": 45830
+    },
+    {
+      "epoch": 13.71,
+      "grad_norm": 2.3755998611450195,
+      "learning_rate": 1.1244391225216061e-05,
+      "loss": 0.9196,
+      "step": 45835
+    },
+    {
+      "epoch": 13.71,
+      "grad_norm": 4.774353504180908,
+      "learning_rate": 1.123948570511831e-05,
+      "loss": 1.012,
+      "step": 45840
+    },
+    {
+      "epoch": 13.72,
+      "grad_norm": 4.135644435882568,
+      "learning_rate": 1.123458094499946e-05,
+      "loss": 1.0807,
+      "step": 45845
+    },
+    {
+      "epoch": 13.72,
+      "grad_norm": 2.5520973205566406,
+      "learning_rate": 1.1229676945130394e-05,
+      "loss": 1.0183,
+      "step": 45850
+    },
+    {
+      "epoch": 13.72,
+      "grad_norm": 4.4872870445251465,
+      "learning_rate": 1.122477370578196e-05,
+      "loss": 0.9953,
+      "step": 45855
+    },
+    {
+      "epoch": 13.72,
+      "grad_norm": 1.6755377054214478,
+      "learning_rate": 1.1219871227224953e-05,
+      "loss": 0.929,
+      "step": 45860
+    },
+    {
+      "epoch": 13.72,
+      "grad_norm": 3.4664084911346436,
+      "learning_rate": 1.1214969509730135e-05,
+      "loss": 0.9963,
+      "step": 45865
+    },
+    {
+      "epoch": 13.72,
+      "grad_norm": 4.792949199676514,
+      "learning_rate": 1.1210068553568224e-05,
+      "loss": 0.7811,
+      "step": 45870
+    },
+    {
+      "epoch": 13.73,
+      "grad_norm": 4.7333550453186035,
+      "learning_rate": 1.120516835900989e-05,
+      "loss": 0.9869,
+      "step": 45875
+    },
+    {
+      "epoch": 13.73,
+      "grad_norm": 2.1963367462158203,
+      "learning_rate": 1.1200268926325771e-05,
+      "loss": 1.0632,
+      "step": 45880
+    },
+    {
+      "epoch": 13.73,
+      "grad_norm": 4.561999797821045,
+      "learning_rate": 1.1195370255786455e-05,
+      "loss": 0.9137,
+      "step": 45885
+    },
+    {
+      "epoch": 13.73,
+      "grad_norm": 2.747818946838379,
+      "learning_rate": 1.119047234766249e-05,
+      "loss": 0.9463,
+      "step": 45890
+    },
+    {
+      "epoch": 13.73,
+      "grad_norm": 3.7482149600982666,
+      "learning_rate": 1.1185575202224383e-05,
+      "loss": 0.8418,
+      "step": 45895
+    },
+    {
+      "epoch": 13.73,
+      "grad_norm": 2.163506507873535,
+      "learning_rate": 1.1180678819742599e-05,
+      "loss": 1.2076,
+      "step": 45900
+    },
+    {
+      "epoch": 13.73,
+      "grad_norm": 1.5832871198654175,
+      "learning_rate": 1.1175783200487558e-05,
+      "loss": 1.1003,
+      "step": 45905
+    },
+    {
+      "epoch": 13.74,
+      "grad_norm": 4.2854413986206055,
+      "learning_rate": 1.1170888344729652e-05,
+      "loss": 1.0152,
+      "step": 45910
+    },
+    {
+      "epoch": 13.74,
+      "grad_norm": 2.954716920852661,
+      "learning_rate": 1.116599425273919e-05,
+      "loss": 1.0792,
+      "step": 45915
+    },
+    {
+      "epoch": 13.74,
+      "grad_norm": 2.61503005027771,
+      "learning_rate": 1.1161100924786502e-05,
+      "loss": 1.0473,
+      "step": 45920
+    },
+    {
+      "epoch": 13.74,
+      "grad_norm": 4.037346839904785,
+      "learning_rate": 1.115620836114181e-05,
+      "loss": 0.996,
+      "step": 45925
+    },
+    {
+      "epoch": 13.74,
+      "grad_norm": 1.7600798606872559,
+      "learning_rate": 1.1151316562075356e-05,
+      "loss": 0.942,
+      "step": 45930
+    },
+    {
+      "epoch": 13.74,
+      "grad_norm": 2.2943618297576904,
+      "learning_rate": 1.1146425527857274e-05,
+      "loss": 0.9601,
+      "step": 45935
+    },
+    {
+      "epoch": 13.74,
+      "grad_norm": 2.6054422855377197,
+      "learning_rate": 1.1141535258757733e-05,
+      "loss": 1.0357,
+      "step": 45940
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 2.9605374336242676,
+      "learning_rate": 1.1136645755046781e-05,
+      "loss": 0.9024,
+      "step": 45945
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 6.355667591094971,
+      "learning_rate": 1.1131757016994476e-05,
+      "loss": 1.1976,
+      "step": 45950
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 2.857518434524536,
+      "learning_rate": 1.1126869044870817e-05,
+      "loss": 1.0472,
+      "step": 45955
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 3.006471633911133,
+      "learning_rate": 1.1121981838945747e-05,
+      "loss": 1.1969,
+      "step": 45960
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 2.723449945449829,
+      "learning_rate": 1.1117095399489215e-05,
+      "loss": 0.8746,
+      "step": 45965
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 2.74454927444458,
+      "learning_rate": 1.1112209726771067e-05,
+      "loss": 1.1539,
+      "step": 45970
+    },
+    {
+      "epoch": 13.76,
+      "grad_norm": 3.3110272884368896,
+      "learning_rate": 1.1107324821061139e-05,
+      "loss": 0.9179,
+      "step": 45975
+    },
+    {
+      "epoch": 13.76,
+      "grad_norm": 1.4932845830917358,
+      "learning_rate": 1.1102440682629217e-05,
+      "loss": 1.0535,
+      "step": 45980
+    },
+    {
+      "epoch": 13.76,
+      "grad_norm": 2.4937548637390137,
+      "learning_rate": 1.1097557311745055e-05,
+      "loss": 1.027,
+      "step": 45985
+    },
+    {
+      "epoch": 13.76,
+      "grad_norm": 2.2536213397979736,
+      "learning_rate": 1.1092674708678349e-05,
+      "loss": 1.0212,
+      "step": 45990
+    },
+    {
+      "epoch": 13.76,
+      "grad_norm": 1.4273040294647217,
+      "learning_rate": 1.1087792873698763e-05,
+      "loss": 1.0554,
+      "step": 45995
+    },
+    {
+      "epoch": 13.76,
+      "grad_norm": 3.2895405292510986,
+      "learning_rate": 1.1082911807075917e-05,
+      "loss": 0.9926,
+      "step": 46000
+    },
+    {
+      "epoch": 13.76,
+      "grad_norm": 2.659669876098633,
+      "learning_rate": 1.1078031509079394e-05,
+      "loss": 1.0621,
+      "step": 46005
+    },
+    {
+      "epoch": 13.77,
+      "grad_norm": 6.099573135375977,
+      "learning_rate": 1.1073151979978702e-05,
+      "loss": 0.9202,
+      "step": 46010
+    },
+    {
+      "epoch": 13.77,
+      "grad_norm": 3.4382877349853516,
+      "learning_rate": 1.1068273220043367e-05,
+      "loss": 1.0081,
+      "step": 46015
+    },
+    {
+      "epoch": 13.77,
+      "grad_norm": 2.076596975326538,
+      "learning_rate": 1.1063395229542806e-05,
+      "loss": 1.0796,
+      "step": 46020
+    },
+    {
+      "epoch": 13.77,
+      "grad_norm": 3.9843549728393555,
+      "learning_rate": 1.1058518008746454e-05,
+      "loss": 0.8787,
+      "step": 46025
+    },
+    {
+      "epoch": 13.77,
+      "grad_norm": 4.303983688354492,
+      "learning_rate": 1.1053641557923647e-05,
+      "loss": 1.0372,
+      "step": 46030
+    },
+    {
+      "epoch": 13.77,
+      "grad_norm": 2.91055965423584,
+      "learning_rate": 1.1048765877343736e-05,
+      "loss": 0.9737,
+      "step": 46035
+    },
+    {
+      "epoch": 13.77,
+      "grad_norm": 3.4289169311523438,
+      "learning_rate": 1.1043890967275978e-05,
+      "loss": 1.1013,
+      "step": 46040
+    },
+    {
+      "epoch": 13.78,
+      "grad_norm": 18.518823623657227,
+      "learning_rate": 1.1039016827989601e-05,
+      "loss": 1.1429,
+      "step": 46045
+    },
+    {
+      "epoch": 13.78,
+      "grad_norm": 3.946131706237793,
+      "learning_rate": 1.1034143459753835e-05,
+      "loss": 0.9853,
+      "step": 46050
+    },
+    {
+      "epoch": 13.78,
+      "grad_norm": 1.882799506187439,
+      "learning_rate": 1.102927086283779e-05,
+      "loss": 0.9411,
+      "step": 46055
+    },
+    {
+      "epoch": 13.78,
+      "grad_norm": 4.6659393310546875,
+      "learning_rate": 1.102439903751061e-05,
+      "loss": 0.8427,
+      "step": 46060
+    },
+    {
+      "epoch": 13.78,
+      "grad_norm": 4.819268703460693,
+      "learning_rate": 1.1019527984041328e-05,
+      "loss": 1.0051,
+      "step": 46065
+    },
+    {
+      "epoch": 13.78,
+      "grad_norm": 1.3024616241455078,
+      "learning_rate": 1.1014657702699003e-05,
+      "loss": 1.1975,
+      "step": 46070
+    },
+    {
+      "epoch": 13.79,
+      "grad_norm": 5.274040222167969,
+      "learning_rate": 1.1009788193752584e-05,
+      "loss": 1.0944,
+      "step": 46075
+    },
+    {
+      "epoch": 13.79,
+      "grad_norm": 1.9087549448013306,
+      "learning_rate": 1.1004919457471022e-05,
+      "loss": 1.1598,
+      "step": 46080
+    },
+    {
+      "epoch": 13.79,
+      "grad_norm": 5.655836582183838,
+      "learning_rate": 1.1000051494123211e-05,
+      "loss": 0.8013,
+      "step": 46085
+    },
+    {
+      "epoch": 13.79,
+      "grad_norm": 4.382424831390381,
+      "learning_rate": 1.0995184303978004e-05,
+      "loss": 0.9066,
+      "step": 46090
+    },
+    {
+      "epoch": 13.79,
+      "grad_norm": 4.181612968444824,
+      "learning_rate": 1.0990317887304214e-05,
+      "loss": 0.8979,
+      "step": 46095
+    },
+    {
+      "epoch": 13.79,
+      "grad_norm": 3.6581287384033203,
+      "learning_rate": 1.0985452244370608e-05,
+      "loss": 0.9312,
+      "step": 46100
+    },
+    {
+      "epoch": 13.79,
+      "grad_norm": 3.3705477714538574,
+      "learning_rate": 1.0980587375445894e-05,
+      "loss": 0.9881,
+      "step": 46105
+    },
+    {
+      "epoch": 13.8,
+      "grad_norm": 1.2490730285644531,
+      "learning_rate": 1.0975723280798783e-05,
+      "loss": 0.9072,
+      "step": 46110
+    },
+    {
+      "epoch": 13.8,
+      "grad_norm": 2.7945122718811035,
+      "learning_rate": 1.0970859960697879e-05,
+      "loss": 0.9368,
+      "step": 46115
+    },
+    {
+      "epoch": 13.8,
+      "grad_norm": 1.166204810142517,
+      "learning_rate": 1.0965997415411808e-05,
+      "loss": 1.0581,
+      "step": 46120
+    },
+    {
+      "epoch": 13.8,
+      "grad_norm": 1.6771597862243652,
+      "learning_rate": 1.096113564520911e-05,
+      "loss": 0.8542,
+      "step": 46125
+    },
+    {
+      "epoch": 13.8,
+      "grad_norm": 4.801357269287109,
+      "learning_rate": 1.09562746503583e-05,
+      "loss": 1.0804,
+      "step": 46130
+    },
+    {
+      "epoch": 13.8,
+      "grad_norm": 3.126763105392456,
+      "learning_rate": 1.0951414431127852e-05,
+      "loss": 0.8767,
+      "step": 46135
+    },
+    {
+      "epoch": 13.8,
+      "grad_norm": 3.6435258388519287,
+      "learning_rate": 1.0946554987786162e-05,
+      "loss": 0.9254,
+      "step": 46140
+    },
+    {
+      "epoch": 13.81,
+      "grad_norm": 3.6946301460266113,
+      "learning_rate": 1.0941696320601652e-05,
+      "loss": 0.8884,
+      "step": 46145
+    },
+    {
+      "epoch": 13.81,
+      "grad_norm": 1.621283769607544,
+      "learning_rate": 1.0936838429842621e-05,
+      "loss": 1.1543,
+      "step": 46150
+    },
+    {
+      "epoch": 13.81,
+      "grad_norm": 2.5034942626953125,
+      "learning_rate": 1.0931981315777406e-05,
+      "loss": 1.0512,
+      "step": 46155
+    },
+    {
+      "epoch": 13.81,
+      "grad_norm": 2.1045899391174316,
+      "learning_rate": 1.0927124978674225e-05,
+      "loss": 1.0269,
+      "step": 46160
+    },
+    {
+      "epoch": 13.81,
+      "grad_norm": 1.1668440103530884,
+      "learning_rate": 1.0922269418801318e-05,
+      "loss": 0.8998,
+      "step": 46165
+    },
+    {
+      "epoch": 13.81,
+      "grad_norm": 1.8777698278427124,
+      "learning_rate": 1.091741463642683e-05,
+      "loss": 1.0598,
+      "step": 46170
+    },
+    {
+      "epoch": 13.82,
+      "grad_norm": 2.1455204486846924,
+      "learning_rate": 1.091256063181889e-05,
+      "loss": 1.0727,
+      "step": 46175
+    },
+    {
+      "epoch": 13.82,
+      "grad_norm": 3.745995283126831,
+      "learning_rate": 1.0907707405245588e-05,
+      "loss": 0.9028,
+      "step": 46180
+    },
+    {
+      "epoch": 13.82,
+      "grad_norm": 4.1767096519470215,
+      "learning_rate": 1.0902854956974957e-05,
+      "loss": 1.0149,
+      "step": 46185
+    },
+    {
+      "epoch": 13.82,
+      "grad_norm": 1.699058175086975,
+      "learning_rate": 1.0898003287274991e-05,
+      "loss": 0.9797,
+      "step": 46190
+    },
+    {
+      "epoch": 13.82,
+      "grad_norm": 2.886626958847046,
+      "learning_rate": 1.0893152396413655e-05,
+      "loss": 1.0911,
+      "step": 46195
+    },
+    {
+      "epoch": 13.82,
+      "grad_norm": 6.132671356201172,
+      "learning_rate": 1.0888302284658833e-05,
+      "loss": 0.9886,
+      "step": 46200
+    },
+    {
+      "epoch": 13.82,
+      "grad_norm": 1.1873294115066528,
+      "learning_rate": 1.0883452952278416e-05,
+      "loss": 1.1952,
+      "step": 46205
+    },
+    {
+      "epoch": 13.83,
+      "grad_norm": 2.7054784297943115,
+      "learning_rate": 1.0878604399540219e-05,
+      "loss": 1.0763,
+      "step": 46210
+    },
+    {
+      "epoch": 13.83,
+      "grad_norm": 2.0939934253692627,
+      "learning_rate": 1.087375662671202e-05,
+      "loss": 0.959,
+      "step": 46215
+    },
+    {
+      "epoch": 13.83,
+      "grad_norm": 2.734236717224121,
+      "learning_rate": 1.0868909634061561e-05,
+      "loss": 0.9568,
+      "step": 46220
+    },
+    {
+      "epoch": 13.83,
+      "grad_norm": 2.2614846229553223,
+      "learning_rate": 1.0864063421856535e-05,
+      "loss": 1.0809,
+      "step": 46225
+    },
+    {
+      "epoch": 13.83,
+      "grad_norm": 2.1451010704040527,
+      "learning_rate": 1.08592179903646e-05,
+      "loss": 1.1251,
+      "step": 46230
+    },
+    {
+      "epoch": 13.83,
+      "grad_norm": 1.8617221117019653,
+      "learning_rate": 1.085437333985334e-05,
+      "loss": 0.9847,
+      "step": 46235
+    },
+    {
+      "epoch": 13.83,
+      "grad_norm": 6.787776947021484,
+      "learning_rate": 1.0849529470590358e-05,
+      "loss": 0.777,
+      "step": 46240
+    },
+    {
+      "epoch": 13.84,
+      "grad_norm": 5.022008895874023,
+      "learning_rate": 1.0844686382843134e-05,
+      "loss": 1.2188,
+      "step": 46245
+    },
+    {
+      "epoch": 13.84,
+      "grad_norm": 3.551828622817993,
+      "learning_rate": 1.0839844076879185e-05,
+      "loss": 0.998,
+      "step": 46250
+    },
+    {
+      "epoch": 13.84,
+      "grad_norm": 2.8129985332489014,
+      "learning_rate": 1.0835002552965911e-05,
+      "loss": 1.0112,
+      "step": 46255
+    },
+    {
+      "epoch": 13.84,
+      "grad_norm": 5.713034629821777,
+      "learning_rate": 1.083016181137074e-05,
+      "loss": 0.8538,
+      "step": 46260
+    },
+    {
+      "epoch": 13.84,
+      "grad_norm": 7.516810417175293,
+      "learning_rate": 1.0825321852360995e-05,
+      "loss": 0.889,
+      "step": 46265
+    },
+    {
+      "epoch": 13.84,
+      "grad_norm": 1.1386287212371826,
+      "learning_rate": 1.0820482676203991e-05,
+      "loss": 1.0063,
+      "step": 46270
+    },
+    {
+      "epoch": 13.84,
+      "grad_norm": 1.6292932033538818,
+      "learning_rate": 1.081564428316699e-05,
+      "loss": 0.9629,
+      "step": 46275
+    },
+    {
+      "epoch": 13.85,
+      "grad_norm": 2.4698173999786377,
+      "learning_rate": 1.081080667351721e-05,
+      "loss": 0.9526,
+      "step": 46280
+    },
+    {
+      "epoch": 13.85,
+      "grad_norm": 5.473732948303223,
+      "learning_rate": 1.0805969847521829e-05,
+      "loss": 0.8786,
+      "step": 46285
+    },
+    {
+      "epoch": 13.85,
+      "grad_norm": 2.3145463466644287,
+      "learning_rate": 1.0801133805447979e-05,
+      "loss": 0.8604,
+      "step": 46290
+    },
+    {
+      "epoch": 13.85,
+      "grad_norm": 4.11014461517334,
+      "learning_rate": 1.0796298547562753e-05,
+      "loss": 1.0444,
+      "step": 46295
+    },
+    {
+      "epoch": 13.85,
+      "grad_norm": 2.2047159671783447,
+      "learning_rate": 1.0791464074133189e-05,
+      "loss": 1.0339,
+      "step": 46300
+    },
+    {
+      "epoch": 13.85,
+      "grad_norm": 5.002912521362305,
+      "learning_rate": 1.07866303854263e-05,
+      "loss": 1.0871,
+      "step": 46305
+    },
+    {
+      "epoch": 13.86,
+      "grad_norm": 2.067976713180542,
+      "learning_rate": 1.0781797481709039e-05,
+      "loss": 1.0389,
+      "step": 46310
+    },
+    {
+      "epoch": 13.86,
+      "grad_norm": 3.528301477432251,
+      "learning_rate": 1.0776965363248326e-05,
+      "loss": 1.061,
+      "step": 46315
+    },
+    {
+      "epoch": 13.86,
+      "grad_norm": 2.523347854614258,
+      "learning_rate": 1.077213403031103e-05,
+      "loss": 0.9864,
+      "step": 46320
+    },
+    {
+      "epoch": 13.86,
+      "grad_norm": 4.887777328491211,
+      "learning_rate": 1.0767303483163991e-05,
+      "loss": 1.0392,
+      "step": 46325
+    },
+    {
+      "epoch": 13.86,
+      "grad_norm": 4.290348052978516,
+      "learning_rate": 1.0762473722073968e-05,
+      "loss": 1.0014,
+      "step": 46330
+    },
+    {
+      "epoch": 13.86,
+      "grad_norm": 5.4099931716918945,
+      "learning_rate": 1.0757644747307744e-05,
+      "loss": 0.898,
+      "step": 46335
+    },
+    {
+      "epoch": 13.86,
+      "grad_norm": 3.631016254425049,
+      "learning_rate": 1.0752816559131976e-05,
+      "loss": 0.9292,
+      "step": 46340
+    },
+    {
+      "epoch": 13.87,
+      "grad_norm": 2.76509428024292,
+      "learning_rate": 1.0747989157813357e-05,
+      "loss": 1.0867,
+      "step": 46345
+    },
+    {
+      "epoch": 13.87,
+      "grad_norm": 1.9425288438796997,
+      "learning_rate": 1.0743162543618465e-05,
+      "loss": 1.0359,
+      "step": 46350
+    },
+    {
+      "epoch": 13.87,
+      "grad_norm": 1.2764666080474854,
+      "learning_rate": 1.0738336716813907e-05,
+      "loss": 1.122,
+      "step": 46355
+    },
+    {
+      "epoch": 13.87,
+      "grad_norm": 5.330173492431641,
+      "learning_rate": 1.0733511677666178e-05,
+      "loss": 1.1068,
+      "step": 46360
+    },
+    {
+      "epoch": 13.87,
+      "grad_norm": 4.102222442626953,
+      "learning_rate": 1.0728687426441769e-05,
+      "loss": 1.0624,
+      "step": 46365
+    },
+    {
+      "epoch": 13.87,
+      "grad_norm": 2.589555025100708,
+      "learning_rate": 1.0723863963407119e-05,
+      "loss": 0.9326,
+      "step": 46370
+    },
+    {
+      "epoch": 13.87,
+      "grad_norm": 2.6629140377044678,
+      "learning_rate": 1.0719041288828624e-05,
+      "loss": 0.8154,
+      "step": 46375
+    },
+    {
+      "epoch": 13.88,
+      "grad_norm": 3.4112234115600586,
+      "learning_rate": 1.0714219402972633e-05,
+      "loss": 1.1058,
+      "step": 46380
+    },
+    {
+      "epoch": 13.88,
+      "grad_norm": 2.342379093170166,
+      "learning_rate": 1.0709398306105456e-05,
+      "loss": 1.0584,
+      "step": 46385
+    },
+    {
+      "epoch": 13.88,
+      "grad_norm": 1.7157783508300781,
+      "learning_rate": 1.070457799849336e-05,
+      "loss": 1.0045,
+      "step": 46390
+    },
+    {
+      "epoch": 13.88,
+      "grad_norm": 1.9355769157409668,
+      "learning_rate": 1.0699758480402555e-05,
+      "loss": 0.8319,
+      "step": 46395
+    },
+    {
+      "epoch": 13.88,
+      "grad_norm": 2.4929144382476807,
+      "learning_rate": 1.0694939752099229e-05,
+      "loss": 0.9453,
+      "step": 46400
+    },
+    {
+      "epoch": 13.88,
+      "grad_norm": 3.4646151065826416,
+      "learning_rate": 1.069012181384951e-05,
+      "loss": 1.0055,
+      "step": 46405
+    },
+    {
+      "epoch": 13.89,
+      "grad_norm": 3.360816240310669,
+      "learning_rate": 1.068530466591949e-05,
+      "loss": 1.0103,
+      "step": 46410
+    },
+    {
+      "epoch": 13.89,
+      "grad_norm": 7.58369255065918,
+      "learning_rate": 1.0680488308575215e-05,
+      "loss": 1.0717,
+      "step": 46415
+    },
+    {
+      "epoch": 13.89,
+      "grad_norm": 3.2002921104431152,
+      "learning_rate": 1.0675672742082692e-05,
+      "loss": 0.9674,
+      "step": 46420
+    },
+    {
+      "epoch": 13.89,
+      "grad_norm": 3.446225643157959,
+      "learning_rate": 1.067085796670786e-05,
+      "loss": 1.1751,
+      "step": 46425
+    },
+    {
+      "epoch": 13.89,
+      "grad_norm": 1.5883591175079346,
+      "learning_rate": 1.0666043982716662e-05,
+      "loss": 1.177,
+      "step": 46430
+    },
+    {
+      "epoch": 13.89,
+      "grad_norm": 5.534438133239746,
+      "learning_rate": 1.066123079037494e-05,
+      "loss": 0.8887,
+      "step": 46435
+    },
+    {
+      "epoch": 13.89,
+      "grad_norm": 1.9987788200378418,
+      "learning_rate": 1.0656418389948556e-05,
+      "loss": 0.9892,
+      "step": 46440
+    },
+    {
+      "epoch": 13.9,
+      "grad_norm": 1.9698108434677124,
+      "learning_rate": 1.0651606781703256e-05,
+      "loss": 1.0445,
+      "step": 46445
+    },
+    {
+      "epoch": 13.9,
+      "grad_norm": 1.448635458946228,
+      "learning_rate": 1.0646795965904815e-05,
+      "loss": 1.0926,
+      "step": 46450
+    },
+    {
+      "epoch": 13.9,
+      "grad_norm": 1.830094337463379,
+      "learning_rate": 1.0641985942818907e-05,
+      "loss": 1.0397,
+      "step": 46455
+    },
+    {
+      "epoch": 13.9,
+      "grad_norm": 3.425884485244751,
+      "learning_rate": 1.0637176712711192e-05,
+      "loss": 1.1553,
+      "step": 46460
+    },
+    {
+      "epoch": 13.9,
+      "grad_norm": 3.424363136291504,
+      "learning_rate": 1.0632368275847276e-05,
+      "loss": 1.083,
+      "step": 46465
+    },
+    {
+      "epoch": 13.9,
+      "grad_norm": 4.42012357711792,
+      "learning_rate": 1.0627560632492726e-05,
+      "loss": 0.9418,
+      "step": 46470
+    },
+    {
+      "epoch": 13.9,
+      "grad_norm": 1.4648548364639282,
+      "learning_rate": 1.0622753782913062e-05,
+      "loss": 0.844,
+      "step": 46475
+    },
+    {
+      "epoch": 13.91,
+      "grad_norm": 2.15419340133667,
+      "learning_rate": 1.0617947727373764e-05,
+      "loss": 1.0659,
+      "step": 46480
+    },
+    {
+      "epoch": 13.91,
+      "grad_norm": 1.1197208166122437,
+      "learning_rate": 1.0613142466140264e-05,
+      "loss": 1.0441,
+      "step": 46485
+    },
+    {
+      "epoch": 13.91,
+      "grad_norm": 1.7040044069290161,
+      "learning_rate": 1.0608337999477949e-05,
+      "loss": 1.058,
+      "step": 46490
+    },
+    {
+      "epoch": 13.91,
+      "grad_norm": 3.230975389480591,
+      "learning_rate": 1.0603534327652167e-05,
+      "loss": 0.8304,
+      "step": 46495
+    },
+    {
+      "epoch": 13.91,
+      "grad_norm": 3.9677131175994873,
+      "learning_rate": 1.0598731450928224e-05,
+      "loss": 1.0955,
+      "step": 46500
+    },
+    {
+      "epoch": 13.91,
+      "grad_norm": 4.740390777587891,
+      "learning_rate": 1.059392936957137e-05,
+      "loss": 0.8264,
+      "step": 46505
+    },
+    {
+      "epoch": 13.92,
+      "grad_norm": 3.7869038581848145,
+      "learning_rate": 1.0589128083846822e-05,
+      "loss": 0.9756,
+      "step": 46510
+    },
+    {
+      "epoch": 13.92,
+      "grad_norm": 3.1534292697906494,
+      "learning_rate": 1.0584327594019752e-05,
+      "loss": 1.0928,
+      "step": 46515
+    },
+    {
+      "epoch": 13.92,
+      "grad_norm": 3.072333812713623,
+      "learning_rate": 1.0579527900355286e-05,
+      "loss": 0.8572,
+      "step": 46520
+    },
+    {
+      "epoch": 13.92,
+      "grad_norm": 1.7602797746658325,
+      "learning_rate": 1.0574729003118511e-05,
+      "loss": 0.8316,
+      "step": 46525
+    },
+    {
+      "epoch": 13.92,
+      "grad_norm": 2.0324742794036865,
+      "learning_rate": 1.0569930902574443e-05,
+      "loss": 0.9476,
+      "step": 46530
+    },
+    {
+      "epoch": 13.92,
+      "grad_norm": 1.305268406867981,
+      "learning_rate": 1.056513359898811e-05,
+      "loss": 1.0128,
+      "step": 46535
+    },
+    {
+      "epoch": 13.92,
+      "grad_norm": 1.4846941232681274,
+      "learning_rate": 1.0560337092624426e-05,
+      "loss": 1.087,
+      "step": 46540
+    },
+    {
+      "epoch": 13.93,
+      "grad_norm": 1.1681172847747803,
+      "learning_rate": 1.0555541383748332e-05,
+      "loss": 0.988,
+      "step": 46545
+    },
+    {
+      "epoch": 13.93,
+      "grad_norm": 4.223578929901123,
+      "learning_rate": 1.0550746472624665e-05,
+      "loss": 1.1653,
+      "step": 46550
+    },
+    {
+      "epoch": 13.93,
+      "grad_norm": 1.3550422191619873,
+      "learning_rate": 1.0545952359518251e-05,
+      "loss": 1.0737,
+      "step": 46555
+    },
+    {
+      "epoch": 13.93,
+      "grad_norm": 3.005021095275879,
+      "learning_rate": 1.0541159044693866e-05,
+      "loss": 0.8508,
+      "step": 46560
+    },
+    {
+      "epoch": 13.93,
+      "grad_norm": 3.721085786819458,
+      "learning_rate": 1.0536366528416236e-05,
+      "loss": 1.0313,
+      "step": 46565
+    },
+    {
+      "epoch": 13.93,
+      "grad_norm": 1.5910168886184692,
+      "learning_rate": 1.0531574810950048e-05,
+      "loss": 1.0548,
+      "step": 46570
+    },
+    {
+      "epoch": 13.93,
+      "grad_norm": 2.5519471168518066,
+      "learning_rate": 1.0526783892559944e-05,
+      "loss": 1.1611,
+      "step": 46575
+    },
+    {
+      "epoch": 13.94,
+      "grad_norm": 1.598804235458374,
+      "learning_rate": 1.0521993773510522e-05,
+      "loss": 1.0835,
+      "step": 46580
+    },
+    {
+      "epoch": 13.94,
+      "grad_norm": 4.5623273849487305,
+      "learning_rate": 1.0517204454066337e-05,
+      "loss": 1.0025,
+      "step": 46585
+    },
+    {
+      "epoch": 13.94,
+      "grad_norm": 2.7830114364624023,
+      "learning_rate": 1.0512415934491893e-05,
+      "loss": 0.9183,
+      "step": 46590
+    },
+    {
+      "epoch": 13.94,
+      "grad_norm": 2.62373685836792,
+      "learning_rate": 1.050762821505166e-05,
+      "loss": 0.8919,
+      "step": 46595
+    },
+    {
+      "epoch": 13.94,
+      "grad_norm": 8.385918617248535,
+      "learning_rate": 1.0502841296010055e-05,
+      "loss": 1.0263,
+      "step": 46600
+    },
+    {
+      "epoch": 13.94,
+      "grad_norm": 1.7394742965698242,
+      "learning_rate": 1.0498055177631458e-05,
+      "loss": 1.0657,
+      "step": 46605
+    },
+    {
+      "epoch": 13.95,
+      "grad_norm": 3.228177070617676,
+      "learning_rate": 1.0493269860180196e-05,
+      "loss": 0.856,
+      "step": 46610
+    },
+    {
+      "epoch": 13.95,
+      "grad_norm": 3.1336793899536133,
+      "learning_rate": 1.0488485343920564e-05,
+      "loss": 0.968,
+      "step": 46615
+    },
+    {
+      "epoch": 13.95,
+      "grad_norm": 3.7546615600585938,
+      "learning_rate": 1.0483701629116813e-05,
+      "loss": 0.9154,
+      "step": 46620
+    },
+    {
+      "epoch": 13.95,
+      "grad_norm": 4.517045497894287,
+      "learning_rate": 1.047891871603311e-05,
+      "loss": 1.031,
+      "step": 46625
+    },
+    {
+      "epoch": 13.95,
+      "grad_norm": 4.163939952850342,
+      "learning_rate": 1.0474136604933654e-05,
+      "loss": 0.9468,
+      "step": 46630
+    },
+    {
+      "epoch": 13.95,
+      "grad_norm": 1.9266438484191895,
+      "learning_rate": 1.0469355296082513e-05,
+      "loss": 1.1067,
+      "step": 46635
+    },
+    {
+      "epoch": 13.95,
+      "grad_norm": 1.395596981048584,
+      "learning_rate": 1.0464574789743798e-05,
+      "loss": 0.9645,
+      "step": 46640
+    },
+    {
+      "epoch": 13.96,
+      "grad_norm": 3.1779944896698,
+      "learning_rate": 1.0459795086181487e-05,
+      "loss": 0.9891,
+      "step": 46645
+    },
+    {
+      "epoch": 13.96,
+      "grad_norm": 7.933053493499756,
+      "learning_rate": 1.04550161856596e-05,
+      "loss": 1.0198,
+      "step": 46650
+    },
+    {
+      "epoch": 13.96,
+      "grad_norm": 3.6656312942504883,
+      "learning_rate": 1.0450238088442038e-05,
+      "loss": 0.8881,
+      "step": 46655
+    },
+    {
+      "epoch": 13.96,
+      "grad_norm": 3.141336679458618,
+      "learning_rate": 1.0445460794792706e-05,
+      "loss": 1.0471,
+      "step": 46660
+    },
+    {
+      "epoch": 13.96,
+      "grad_norm": 1.635010004043579,
+      "learning_rate": 1.0440684304975442e-05,
+      "loss": 0.8864,
+      "step": 46665
+    },
+    {
+      "epoch": 13.96,
+      "grad_norm": 3.4402453899383545,
+      "learning_rate": 1.0435908619254053e-05,
+      "loss": 1.0659,
+      "step": 46670
+    },
+    {
+      "epoch": 13.96,
+      "grad_norm": 2.2001841068267822,
+      "learning_rate": 1.0431133737892292e-05,
+      "loss": 0.9821,
+      "step": 46675
+    },
+    {
+      "epoch": 13.97,
+      "grad_norm": 2.5706429481506348,
+      "learning_rate": 1.0426359661153873e-05,
+      "loss": 0.8815,
+      "step": 46680
+    },
+    {
+      "epoch": 13.97,
+      "grad_norm": 5.090077877044678,
+      "learning_rate": 1.042158638930246e-05,
+      "loss": 0.9084,
+      "step": 46685
+    },
+    {
+      "epoch": 13.97,
+      "grad_norm": 4.350427150726318,
+      "learning_rate": 1.0416813922601677e-05,
+      "loss": 1.0049,
+      "step": 46690
+    },
+    {
+      "epoch": 13.97,
+      "grad_norm": 3.4876809120178223,
+      "learning_rate": 1.0412042261315103e-05,
+      "loss": 1.1069,
+      "step": 46695
+    },
+    {
+      "epoch": 13.97,
+      "grad_norm": 3.4440534114837646,
+      "learning_rate": 1.0407271405706271e-05,
+      "loss": 1.1091,
+      "step": 46700
+    },
+    {
+      "epoch": 13.97,
+      "grad_norm": 2.864044427871704,
+      "learning_rate": 1.0402501356038671e-05,
+      "loss": 0.8457,
+      "step": 46705
+    },
+    {
+      "epoch": 13.98,
+      "grad_norm": 1.3529998064041138,
+      "learning_rate": 1.0397732112575747e-05,
+      "loss": 1.0893,
+      "step": 46710
+    },
+    {
+      "epoch": 13.98,
+      "grad_norm": 1.4070992469787598,
+      "learning_rate": 1.0392963675580911e-05,
+      "loss": 1.1165,
+      "step": 46715
+    },
+    {
+      "epoch": 13.98,
+      "grad_norm": 3.6802985668182373,
+      "learning_rate": 1.0388196045317488e-05,
+      "loss": 1.1286,
+      "step": 46720
+    },
+    {
+      "epoch": 13.98,
+      "grad_norm": 1.862669587135315,
+      "learning_rate": 1.0383429222048829e-05,
+      "loss": 0.9559,
+      "step": 46725
+    },
+    {
+      "epoch": 13.98,
+      "grad_norm": 2.9887638092041016,
+      "learning_rate": 1.0378663206038164e-05,
+      "loss": 0.9677,
+      "step": 46730
+    },
+    {
+      "epoch": 13.98,
+      "grad_norm": 3.377013921737671,
+      "learning_rate": 1.0373897997548748e-05,
+      "loss": 0.9216,
+      "step": 46735
+    },
+    {
+      "epoch": 13.98,
+      "grad_norm": 3.8938591480255127,
+      "learning_rate": 1.0369133596843727e-05,
+      "loss": 1.0537,
+      "step": 46740
+    },
+    {
+      "epoch": 13.99,
+      "grad_norm": 3.0918357372283936,
+      "learning_rate": 1.0364370004186267e-05,
+      "loss": 1.0356,
+      "step": 46745
+    },
+    {
+      "epoch": 13.99,
+      "grad_norm": 4.587892055511475,
+      "learning_rate": 1.0359607219839427e-05,
+      "loss": 0.9786,
+      "step": 46750
+    },
+    {
+      "epoch": 13.99,
+      "grad_norm": 1.8486536741256714,
+      "learning_rate": 1.0354845244066263e-05,
+      "loss": 0.9638,
+      "step": 46755
+    },
+    {
+      "epoch": 13.99,
+      "grad_norm": 1.7416905164718628,
+      "learning_rate": 1.0350084077129776e-05,
+      "loss": 1.0862,
+      "step": 46760
+    },
+    {
+      "epoch": 13.99,
+      "grad_norm": 2.3283634185791016,
+      "learning_rate": 1.0345323719292915e-05,
+      "loss": 1.0598,
+      "step": 46765
+    },
+    {
+      "epoch": 13.99,
+      "grad_norm": 1.997711181640625,
+      "learning_rate": 1.0340564170818595e-05,
+      "loss": 1.2272,
+      "step": 46770
+    },
+    {
+      "epoch": 13.99,
+      "grad_norm": 2.0410749912261963,
+      "learning_rate": 1.0335805431969675e-05,
+      "loss": 1.1836,
+      "step": 46775
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 2.7848238945007324,
+      "learning_rate": 1.0331047503008981e-05,
+      "loss": 0.9318,
+      "step": 46780
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 1.9022867679595947,
+      "learning_rate": 1.0326290384199285e-05,
+      "loss": 1.0139,
+      "step": 46785
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 1.689530849456787,
+      "learning_rate": 1.0321534075803319e-05,
+      "loss": 1.0845,
+      "step": 46790
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 4.382483005523682,
+      "learning_rate": 1.0316778578083768e-05,
+      "loss": 0.8879,
+      "step": 46795
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 2.087588310241699,
+      "learning_rate": 1.0312023891303273e-05,
+      "loss": 0.9106,
+      "step": 46800
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 1.368632197380066,
+      "learning_rate": 1.0307270015724433e-05,
+      "loss": 1.122,
+      "step": 46805
+    },
+    {
+      "epoch": 14.01,
+      "grad_norm": 2.369845390319824,
+      "learning_rate": 1.0302516951609806e-05,
+      "loss": 0.9383,
+      "step": 46810
+    },
+    {
+      "epoch": 14.01,
+      "grad_norm": 1.6115524768829346,
+      "learning_rate": 1.0297764699221874e-05,
+      "loss": 0.9586,
+      "step": 46815
+    },
+    {
+      "epoch": 14.01,
+      "grad_norm": 4.875194072723389,
+      "learning_rate": 1.0293013258823131e-05,
+      "loss": 0.9979,
+      "step": 46820
+    },
+    {
+      "epoch": 14.01,
+      "grad_norm": 2.5877294540405273,
+      "learning_rate": 1.028826263067596e-05,
+      "loss": 0.8455,
+      "step": 46825
+    },
+    {
+      "epoch": 14.01,
+      "grad_norm": 2.146116018295288,
+      "learning_rate": 1.0283512815042773e-05,
+      "loss": 0.8959,
+      "step": 46830
+    },
+    {
+      "epoch": 14.01,
+      "grad_norm": 3.2791755199432373,
+      "learning_rate": 1.0278763812185857e-05,
+      "loss": 0.9684,
+      "step": 46835
+    },
+    {
+      "epoch": 14.01,
+      "grad_norm": 2.8382163047790527,
+      "learning_rate": 1.027401562236753e-05,
+      "loss": 1.033,
+      "step": 46840
+    },
+    {
+      "epoch": 14.02,
+      "grad_norm": 2.077665090560913,
+      "learning_rate": 1.0269268245850006e-05,
+      "loss": 0.9306,
+      "step": 46845
+    },
+    {
+      "epoch": 14.02,
+      "grad_norm": 3.7365541458129883,
+      "learning_rate": 1.0264521682895478e-05,
+      "loss": 0.9208,
+      "step": 46850
+    },
+    {
+      "epoch": 14.02,
+      "grad_norm": 3.4815821647644043,
+      "learning_rate": 1.0259775933766117e-05,
+      "loss": 0.9793,
+      "step": 46855
+    },
+    {
+      "epoch": 14.02,
+      "grad_norm": 3.6899566650390625,
+      "learning_rate": 1.0255030998723991e-05,
+      "loss": 0.9788,
+      "step": 46860
+    },
+    {
+      "epoch": 14.02,
+      "grad_norm": 2.341599702835083,
+      "learning_rate": 1.0250286878031193e-05,
+      "loss": 1.0963,
+      "step": 46865
+    },
+    {
+      "epoch": 14.02,
+      "grad_norm": 4.204500675201416,
+      "learning_rate": 1.0245543571949704e-05,
+      "loss": 0.9897,
+      "step": 46870
+    },
+    {
+      "epoch": 14.02,
+      "grad_norm": 2.09965443611145,
+      "learning_rate": 1.0240801080741525e-05,
+      "loss": 1.0211,
+      "step": 46875
+    },
+    {
+      "epoch": 14.03,
+      "grad_norm": 2.6598119735717773,
+      "learning_rate": 1.0236059404668549e-05,
+      "loss": 0.9235,
+      "step": 46880
+    },
+    {
+      "epoch": 14.03,
+      "grad_norm": 1.9532800912857056,
+      "learning_rate": 1.0231318543992669e-05,
+      "loss": 1.044,
+      "step": 46885
+    },
+    {
+      "epoch": 14.03,
+      "grad_norm": 2.8368186950683594,
+      "learning_rate": 1.0226578498975714e-05,
+      "loss": 0.9055,
+      "step": 46890
+    },
+    {
+      "epoch": 14.03,
+      "grad_norm": 1.4692111015319824,
+      "learning_rate": 1.0221839269879472e-05,
+      "loss": 1.0778,
+      "step": 46895
+    },
+    {
+      "epoch": 14.03,
+      "grad_norm": 2.8631324768066406,
+      "learning_rate": 1.0217100856965687e-05,
+      "loss": 0.9229,
+      "step": 46900
+    },
+    {
+      "epoch": 14.03,
+      "grad_norm": 1.9299060106277466,
+      "learning_rate": 1.0212363260496063e-05,
+      "loss": 1.0288,
+      "step": 46905
+    },
+    {
+      "epoch": 14.03,
+      "grad_norm": 2.9853992462158203,
+      "learning_rate": 1.0207626480732228e-05,
+      "loss": 0.952,
+      "step": 46910
+    },
+    {
+      "epoch": 14.04,
+      "grad_norm": 2.1383092403411865,
+      "learning_rate": 1.0202890517935824e-05,
+      "loss": 1.1462,
+      "step": 46915
+    },
+    {
+      "epoch": 14.04,
+      "grad_norm": 6.036135196685791,
+      "learning_rate": 1.019815537236838e-05,
+      "loss": 0.9865,
+      "step": 46920
+    },
+    {
+      "epoch": 14.04,
+      "grad_norm": 4.347772598266602,
+      "learning_rate": 1.0193421044291446e-05,
+      "loss": 0.9413,
+      "step": 46925
+    },
+    {
+      "epoch": 14.04,
+      "grad_norm": 5.3200860023498535,
+      "learning_rate": 1.0188687533966457e-05,
+      "loss": 0.9559,
+      "step": 46930
+    },
+    {
+      "epoch": 14.04,
+      "grad_norm": 1.9751383066177368,
+      "learning_rate": 1.0183954841654873e-05,
+      "loss": 1.0025,
+      "step": 46935
+    },
+    {
+      "epoch": 14.04,
+      "grad_norm": 2.211293935775757,
+      "learning_rate": 1.0179222967618068e-05,
+      "loss": 1.0879,
+      "step": 46940
+    },
+    {
+      "epoch": 14.05,
+      "grad_norm": 2.816315174102783,
+      "learning_rate": 1.0174491912117357e-05,
+      "loss": 1.1058,
+      "step": 46945
+    },
+    {
+      "epoch": 14.05,
+      "grad_norm": 3.3474440574645996,
+      "learning_rate": 1.0169761675414064e-05,
+      "loss": 0.8382,
+      "step": 46950
+    },
+    {
+      "epoch": 14.05,
+      "grad_norm": 1.9791414737701416,
+      "learning_rate": 1.0165032257769403e-05,
+      "loss": 1.1009,
+      "step": 46955
+    },
+    {
+      "epoch": 14.05,
+      "grad_norm": 3.0999069213867188,
+      "learning_rate": 1.0160303659444606e-05,
+      "loss": 0.9814,
+      "step": 46960
+    },
+    {
+      "epoch": 14.05,
+      "grad_norm": 1.7832539081573486,
+      "learning_rate": 1.0155575880700796e-05,
+      "loss": 0.9725,
+      "step": 46965
+    },
+    {
+      "epoch": 14.05,
+      "grad_norm": 2.467935800552368,
+      "learning_rate": 1.015084892179912e-05,
+      "loss": 0.9384,
+      "step": 46970
+    },
+    {
+      "epoch": 14.05,
+      "grad_norm": 1.9563651084899902,
+      "learning_rate": 1.0146122783000612e-05,
+      "loss": 1.039,
+      "step": 46975
+    },
+    {
+      "epoch": 14.06,
+      "grad_norm": 3.385742425918579,
+      "learning_rate": 1.0141397464566302e-05,
+      "loss": 1.0829,
+      "step": 46980
+    },
+    {
+      "epoch": 14.06,
+      "grad_norm": 6.85827112197876,
+      "learning_rate": 1.0136672966757166e-05,
+      "loss": 0.9445,
+      "step": 46985
+    },
+    {
+      "epoch": 14.06,
+      "grad_norm": 3.0008556842803955,
+      "learning_rate": 1.0131949289834133e-05,
+      "loss": 1.0382,
+      "step": 46990
+    },
+    {
+      "epoch": 14.06,
+      "grad_norm": 1.650694727897644,
+      "learning_rate": 1.0127226434058088e-05,
+      "loss": 0.9398,
+      "step": 46995
+    },
+    {
+      "epoch": 14.06,
+      "grad_norm": 2.1701951026916504,
+      "learning_rate": 1.0122504399689863e-05,
+      "loss": 1.0166,
+      "step": 47000
+    },
+    {
+      "epoch": 14.06,
+      "grad_norm": 3.96262526512146,
+      "learning_rate": 1.011778318699026e-05,
+      "loss": 0.8386,
+      "step": 47005
+    },
+    {
+      "epoch": 14.06,
+      "grad_norm": 3.1463615894317627,
+      "learning_rate": 1.0113062796220021e-05,
+      "loss": 1.0431,
+      "step": 47010
+    },
+    {
+      "epoch": 14.07,
+      "grad_norm": 3.198836326599121,
+      "learning_rate": 1.010834322763985e-05,
+      "loss": 1.0575,
+      "step": 47015
+    },
+    {
+      "epoch": 14.07,
+      "grad_norm": 2.9448750019073486,
+      "learning_rate": 1.0103624481510404e-05,
+      "loss": 0.9553,
+      "step": 47020
+    },
+    {
+      "epoch": 14.07,
+      "grad_norm": 1.3542437553405762,
+      "learning_rate": 1.0098906558092291e-05,
+      "loss": 1.0053,
+      "step": 47025
+    },
+    {
+      "epoch": 14.07,
+      "grad_norm": 3.5193848609924316,
+      "learning_rate": 1.0094189457646083e-05,
+      "loss": 0.9204,
+      "step": 47030
+    },
+    {
+      "epoch": 14.07,
+      "grad_norm": 1.792724609375,
+      "learning_rate": 1.0089473180432307e-05,
+      "loss": 1.0258,
+      "step": 47035
+    },
+    {
+      "epoch": 14.07,
+      "grad_norm": 1.6008864641189575,
+      "learning_rate": 1.0084757726711408e-05,
+      "loss": 0.9024,
+      "step": 47040
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 3.162278890609741,
+      "learning_rate": 1.0080043096743857e-05,
+      "loss": 0.9124,
+      "step": 47045
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 2.5094027519226074,
+      "learning_rate": 1.0075329290789998e-05,
+      "loss": 0.9806,
+      "step": 47050
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 3.3782403469085693,
+      "learning_rate": 1.0070616309110207e-05,
+      "loss": 1.0407,
+      "step": 47055
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 4.14473819732666,
+      "learning_rate": 1.0065904151964738e-05,
+      "loss": 0.8774,
+      "step": 47060
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 9.009183883666992,
+      "learning_rate": 1.0061192819613882e-05,
+      "loss": 1.0632,
+      "step": 47065
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 1.744873285293579,
+      "learning_rate": 1.0056482312317809e-05,
+      "loss": 1.0087,
+      "step": 47070
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 1.8300987482070923,
+      "learning_rate": 1.0051772630336684e-05,
+      "loss": 1.0234,
+      "step": 47075
+    },
+    {
+      "epoch": 14.09,
+      "grad_norm": 2.92419171333313,
+      "learning_rate": 1.0047063773930617e-05,
+      "loss": 0.8112,
+      "step": 47080
+    },
+    {
+      "epoch": 14.09,
+      "grad_norm": 2.440823793411255,
+      "learning_rate": 1.0042355743359676e-05,
+      "loss": 0.7468,
+      "step": 47085
+    },
+    {
+      "epoch": 14.09,
+      "grad_norm": 8.091914176940918,
+      "learning_rate": 1.0037648538883882e-05,
+      "loss": 0.9097,
+      "step": 47090
+    },
+    {
+      "epoch": 14.09,
+      "grad_norm": 5.099801063537598,
+      "learning_rate": 1.0032942160763206e-05,
+      "loss": 0.7116,
+      "step": 47095
+    },
+    {
+      "epoch": 14.09,
+      "grad_norm": 1.9223390817642212,
+      "learning_rate": 1.0028236609257574e-05,
+      "loss": 1.0824,
+      "step": 47100
+    },
+    {
+      "epoch": 14.09,
+      "grad_norm": 2.9645304679870605,
+      "learning_rate": 1.0023531884626875e-05,
+      "loss": 1.1287,
+      "step": 47105
+    },
+    {
+      "epoch": 14.09,
+      "grad_norm": 4.622707366943359,
+      "learning_rate": 1.0018827987130942e-05,
+      "loss": 0.9873,
+      "step": 47110
+    },
+    {
+      "epoch": 14.1,
+      "grad_norm": 2.609149217605591,
+      "learning_rate": 1.0014124917029567e-05,
+      "loss": 1.0403,
+      "step": 47115
+    },
+    {
+      "epoch": 14.1,
+      "grad_norm": 1.8755382299423218,
+      "learning_rate": 1.0009422674582497e-05,
+      "loss": 1.0787,
+      "step": 47120
+    },
+    {
+      "epoch": 14.1,
+      "grad_norm": 3.716013193130493,
+      "learning_rate": 1.000472126004943e-05,
+      "loss": 1.0361,
+      "step": 47125
+    },
+    {
+      "epoch": 14.1,
+      "grad_norm": 4.679911136627197,
+      "learning_rate": 1.0000020673690027e-05,
+      "loss": 0.9316,
+      "step": 47130
+    },
+    {
+      "epoch": 14.1,
+      "grad_norm": 4.787873268127441,
+      "learning_rate": 9.995320915763878e-06,
+      "loss": 0.9913,
+      "step": 47135
+    },
+    {
+      "epoch": 14.1,
+      "grad_norm": 4.154457092285156,
+      "learning_rate": 9.990621986530574e-06,
+      "loss": 0.8499,
+      "step": 47140
+    },
+    {
+      "epoch": 14.11,
+      "grad_norm": 2.3199286460876465,
+      "learning_rate": 9.985923886249599e-06,
+      "loss": 1.1649,
+      "step": 47145
+    },
+    {
+      "epoch": 14.11,
+      "grad_norm": 2.379757881164551,
+      "learning_rate": 9.981226615180456e-06,
+      "loss": 0.9088,
+      "step": 47150
+    },
+    {
+      "epoch": 14.11,
+      "grad_norm": 3.3448526859283447,
+      "learning_rate": 9.97653017358254e-06,
+      "loss": 1.0832,
+      "step": 47155
+    },
+    {
+      "epoch": 14.11,
+      "grad_norm": 1.5010708570480347,
+      "learning_rate": 9.971834561715265e-06,
+      "loss": 1.0648,
+      "step": 47160
+    },
+    {
+      "epoch": 14.11,
+      "grad_norm": 1.6477797031402588,
+      "learning_rate": 9.967139779837939e-06,
+      "loss": 0.9349,
+      "step": 47165
+    },
+    {
+      "epoch": 14.11,
+      "grad_norm": 4.819705009460449,
+      "learning_rate": 9.962445828209854e-06,
+      "loss": 0.9833,
+      "step": 47170
+    },
+    {
+      "epoch": 14.11,
+      "grad_norm": 6.116825580596924,
+      "learning_rate": 9.957752707090257e-06,
+      "loss": 0.9763,
+      "step": 47175
+    },
+    {
+      "epoch": 14.12,
+      "grad_norm": 2.0187926292419434,
+      "learning_rate": 9.953060416738343e-06,
+      "loss": 1.0569,
+      "step": 47180
+    },
+    {
+      "epoch": 14.12,
+      "grad_norm": 2.0881597995758057,
+      "learning_rate": 9.94836895741326e-06,
+      "loss": 0.8785,
+      "step": 47185
+    },
+    {
+      "epoch": 14.12,
+      "grad_norm": 4.898426055908203,
+      "learning_rate": 9.943678329374116e-06,
+      "loss": 0.7823,
+      "step": 47190
+    },
+    {
+      "epoch": 14.12,
+      "grad_norm": 3.7305986881256104,
+      "learning_rate": 9.938988532879967e-06,
+      "loss": 0.8913,
+      "step": 47195
+    },
+    {
+      "epoch": 14.12,
+      "grad_norm": 1.356601357460022,
+      "learning_rate": 9.934299568189826e-06,
+      "loss": 1.0196,
+      "step": 47200
+    },
+    {
+      "epoch": 14.12,
+      "grad_norm": 3.387441873550415,
+      "learning_rate": 9.929611435562658e-06,
+      "loss": 0.8262,
+      "step": 47205
+    },
+    {
+      "epoch": 14.12,
+      "grad_norm": 2.2645153999328613,
+      "learning_rate": 9.924924135257388e-06,
+      "loss": 0.9682,
+      "step": 47210
+    },
+    {
+      "epoch": 14.13,
+      "grad_norm": 3.1974477767944336,
+      "learning_rate": 9.92023766753289e-06,
+      "loss": 1.0503,
+      "step": 47215
+    },
+    {
+      "epoch": 14.13,
+      "grad_norm": 1.60183584690094,
+      "learning_rate": 9.915552032647988e-06,
+      "loss": 0.9126,
+      "step": 47220
+    },
+    {
+      "epoch": 14.13,
+      "grad_norm": 2.7414391040802,
+      "learning_rate": 9.910867230861467e-06,
+      "loss": 1.0307,
+      "step": 47225
+    },
+    {
+      "epoch": 14.13,
+      "grad_norm": 3.494734287261963,
+      "learning_rate": 9.906183262432068e-06,
+      "loss": 0.9636,
+      "step": 47230
+    },
+    {
+      "epoch": 14.13,
+      "grad_norm": 1.4926824569702148,
+      "learning_rate": 9.901500127618485e-06,
+      "loss": 0.8568,
+      "step": 47235
+    },
+    {
+      "epoch": 14.13,
+      "grad_norm": 1.8363795280456543,
+      "learning_rate": 9.896817826679338e-06,
+      "loss": 0.9843,
+      "step": 47240
+    },
+    {
+      "epoch": 14.14,
+      "grad_norm": 2.363992214202881,
+      "learning_rate": 9.892136359873264e-06,
+      "loss": 1.0686,
+      "step": 47245
+    },
+    {
+      "epoch": 14.14,
+      "grad_norm": 1.888505458831787,
+      "learning_rate": 9.887455727458775e-06,
+      "loss": 1.0122,
+      "step": 47250
+    },
+    {
+      "epoch": 14.14,
+      "grad_norm": 1.452127456665039,
+      "learning_rate": 9.882775929694416e-06,
+      "loss": 0.9163,
+      "step": 47255
+    },
+    {
+      "epoch": 14.14,
+      "grad_norm": 7.370832443237305,
+      "learning_rate": 9.878096966838617e-06,
+      "loss": 0.8539,
+      "step": 47260
+    },
+    {
+      "epoch": 14.14,
+      "grad_norm": 3.398411750793457,
+      "learning_rate": 9.873418839149808e-06,
+      "loss": 1.0393,
+      "step": 47265
+    },
+    {
+      "epoch": 14.14,
+      "grad_norm": 2.2340104579925537,
+      "learning_rate": 9.868741546886353e-06,
+      "loss": 0.9541,
+      "step": 47270
+    },
+    {
+      "epoch": 14.14,
+      "grad_norm": 2.5456645488739014,
+      "learning_rate": 9.864065090306574e-06,
+      "loss": 1.0563,
+      "step": 47275
+    },
+    {
+      "epoch": 14.15,
+      "grad_norm": 1.726277470588684,
+      "learning_rate": 9.859389469668745e-06,
+      "loss": 1.0858,
+      "step": 47280
+    },
+    {
+      "epoch": 14.15,
+      "grad_norm": 1.36215078830719,
+      "learning_rate": 9.8547146852311e-06,
+      "loss": 1.061,
+      "step": 47285
+    },
+    {
+      "epoch": 14.15,
+      "grad_norm": 3.9773330688476562,
+      "learning_rate": 9.85004073725182e-06,
+      "loss": 0.7896,
+      "step": 47290
+    },
+    {
+      "epoch": 14.15,
+      "grad_norm": 1.2395633459091187,
+      "learning_rate": 9.845367625989044e-06,
+      "loss": 1.0116,
+      "step": 47295
+    },
+    {
+      "epoch": 14.15,
+      "grad_norm": 2.955395460128784,
+      "learning_rate": 9.84069535170086e-06,
+      "loss": 1.0423,
+      "step": 47300
+    },
+    {
+      "epoch": 14.15,
+      "grad_norm": 2.663750171661377,
+      "learning_rate": 9.836023914645313e-06,
+      "loss": 0.9401,
+      "step": 47305
+    },
+    {
+      "epoch": 14.15,
+      "grad_norm": 3.750603437423706,
+      "learning_rate": 9.831353315080405e-06,
+      "loss": 0.9206,
+      "step": 47310
+    },
+    {
+      "epoch": 14.16,
+      "grad_norm": 4.748321056365967,
+      "learning_rate": 9.826683553264085e-06,
+      "loss": 1.0016,
+      "step": 47315
+    },
+    {
+      "epoch": 14.16,
+      "grad_norm": 4.10500955581665,
+      "learning_rate": 9.822014629454263e-06,
+      "loss": 0.9215,
+      "step": 47320
+    },
+    {
+      "epoch": 14.16,
+      "grad_norm": 5.371630668640137,
+      "learning_rate": 9.817346543908796e-06,
+      "loss": 0.9788,
+      "step": 47325
+    },
+    {
+      "epoch": 14.16,
+      "grad_norm": 1.5055346488952637,
+      "learning_rate": 9.812679296885505e-06,
+      "loss": 0.9854,
+      "step": 47330
+    },
+    {
+      "epoch": 14.16,
+      "grad_norm": 1.5663864612579346,
+      "learning_rate": 9.808012888642132e-06,
+      "loss": 0.9251,
+      "step": 47335
+    },
+    {
+      "epoch": 14.16,
+      "grad_norm": 2.7338407039642334,
+      "learning_rate": 9.803347319436435e-06,
+      "loss": 0.9418,
+      "step": 47340
+    },
+    {
+      "epoch": 14.17,
+      "grad_norm": 2.1328275203704834,
+      "learning_rate": 9.798682589526052e-06,
+      "loss": 0.8952,
+      "step": 47345
+    },
+    {
+      "epoch": 14.17,
+      "grad_norm": 3.0970237255096436,
+      "learning_rate": 9.794018699168644e-06,
+      "loss": 0.8826,
+      "step": 47350
+    },
+    {
+      "epoch": 14.17,
+      "grad_norm": 1.936923623085022,
+      "learning_rate": 9.789355648621764e-06,
+      "loss": 0.9455,
+      "step": 47355
+    },
+    {
+      "epoch": 14.17,
+      "grad_norm": 2.612597703933716,
+      "learning_rate": 9.784693438142975e-06,
+      "loss": 0.9279,
+      "step": 47360
+    },
+    {
+      "epoch": 14.17,
+      "grad_norm": 3.304236888885498,
+      "learning_rate": 9.780032067989744e-06,
+      "loss": 0.929,
+      "step": 47365
+    },
+    {
+      "epoch": 14.17,
+      "grad_norm": 1.589285135269165,
+      "learning_rate": 9.775371538419522e-06,
+      "loss": 1.154,
+      "step": 47370
+    },
+    {
+      "epoch": 14.17,
+      "grad_norm": 1.9944769144058228,
+      "learning_rate": 9.770711849689706e-06,
+      "loss": 0.8109,
+      "step": 47375
+    },
+    {
+      "epoch": 14.18,
+      "grad_norm": 3.4619359970092773,
+      "learning_rate": 9.766053002057646e-06,
+      "loss": 0.9818,
+      "step": 47380
+    },
+    {
+      "epoch": 14.18,
+      "grad_norm": 3.2043612003326416,
+      "learning_rate": 9.76139499578064e-06,
+      "loss": 1.0544,
+      "step": 47385
+    },
+    {
+      "epoch": 14.18,
+      "grad_norm": 7.88495397567749,
+      "learning_rate": 9.75673783111595e-06,
+      "loss": 0.891,
+      "step": 47390
+    },
+    {
+      "epoch": 14.18,
+      "grad_norm": 2.1868255138397217,
+      "learning_rate": 9.752081508320785e-06,
+      "loss": 0.9812,
+      "step": 47395
+    },
+    {
+      "epoch": 14.18,
+      "grad_norm": 2.6226940155029297,
+      "learning_rate": 9.747426027652309e-06,
+      "loss": 0.9334,
+      "step": 47400
+    },
+    {
+      "epoch": 14.18,
+      "grad_norm": 3.620140552520752,
+      "learning_rate": 9.74277138936764e-06,
+      "loss": 0.9,
+      "step": 47405
+    },
+    {
+      "epoch": 14.18,
+      "grad_norm": 7.102359771728516,
+      "learning_rate": 9.738117593723847e-06,
+      "loss": 0.8913,
+      "step": 47410
+    },
+    {
+      "epoch": 14.19,
+      "grad_norm": 3.3569910526275635,
+      "learning_rate": 9.733464640977957e-06,
+      "loss": 1.0191,
+      "step": 47415
+    },
+    {
+      "epoch": 14.19,
+      "grad_norm": 3.8863701820373535,
+      "learning_rate": 9.728812531386946e-06,
+      "loss": 1.025,
+      "step": 47420
+    },
+    {
+      "epoch": 14.19,
+      "grad_norm": 2.6565213203430176,
+      "learning_rate": 9.724161265207755e-06,
+      "loss": 1.0252,
+      "step": 47425
+    },
+    {
+      "epoch": 14.19,
+      "grad_norm": 3.354027032852173,
+      "learning_rate": 9.719510842697238e-06,
+      "loss": 0.8642,
+      "step": 47430
+    },
+    {
+      "epoch": 14.19,
+      "grad_norm": 3.7532005310058594,
+      "learning_rate": 9.714861264112273e-06,
+      "loss": 0.9739,
+      "step": 47435
+    },
+    {
+      "epoch": 14.19,
+      "grad_norm": 7.380504608154297,
+      "learning_rate": 9.710212529709618e-06,
+      "loss": 1.0136,
+      "step": 47440
+    },
+    {
+      "epoch": 14.19,
+      "grad_norm": 3.0353171825408936,
+      "learning_rate": 9.705564639746045e-06,
+      "loss": 1.1409,
+      "step": 47445
+    },
+    {
+      "epoch": 14.2,
+      "grad_norm": 4.069791793823242,
+      "learning_rate": 9.700917594478223e-06,
+      "loss": 0.8443,
+      "step": 47450
+    },
+    {
+      "epoch": 14.2,
+      "grad_norm": 2.922313928604126,
+      "learning_rate": 9.69627139416284e-06,
+      "loss": 0.9226,
+      "step": 47455
+    },
+    {
+      "epoch": 14.2,
+      "grad_norm": 2.804293632507324,
+      "learning_rate": 9.69162603905647e-06,
+      "loss": 1.0484,
+      "step": 47460
+    },
+    {
+      "epoch": 14.2,
+      "grad_norm": 1.4043070077896118,
+      "learning_rate": 9.68698152941568e-06,
+      "loss": 1.0361,
+      "step": 47465
+    },
+    {
+      "epoch": 14.2,
+      "grad_norm": 2.1712961196899414,
+      "learning_rate": 9.682337865496985e-06,
+      "loss": 1.0993,
+      "step": 47470
+    },
+    {
+      "epoch": 14.2,
+      "grad_norm": 3.498368740081787,
+      "learning_rate": 9.677695047556848e-06,
+      "loss": 0.9285,
+      "step": 47475
+    },
+    {
+      "epoch": 14.21,
+      "grad_norm": 2.01509428024292,
+      "learning_rate": 9.673053075851687e-06,
+      "loss": 1.0765,
+      "step": 47480
+    },
+    {
+      "epoch": 14.21,
+      "grad_norm": 1.8657152652740479,
+      "learning_rate": 9.668411950637874e-06,
+      "loss": 1.0225,
+      "step": 47485
+    },
+    {
+      "epoch": 14.21,
+      "grad_norm": 3.4774529933929443,
+      "learning_rate": 9.66377167217173e-06,
+      "loss": 0.9132,
+      "step": 47490
+    },
+    {
+      "epoch": 14.21,
+      "grad_norm": 3.322037696838379,
+      "learning_rate": 9.659132240709536e-06,
+      "loss": 0.9264,
+      "step": 47495
+    },
+    {
+      "epoch": 14.21,
+      "grad_norm": 3.845637321472168,
+      "learning_rate": 9.654493656507524e-06,
+      "loss": 0.883,
+      "step": 47500
+    },
+    {
+      "epoch": 14.21,
+      "grad_norm": 3.871750593185425,
+      "learning_rate": 9.649855919821874e-06,
+      "loss": 0.9087,
+      "step": 47505
+    },
+    {
+      "epoch": 14.21,
+      "grad_norm": 6.507814407348633,
+      "learning_rate": 9.645219030908728e-06,
+      "loss": 0.8509,
+      "step": 47510
+    },
+    {
+      "epoch": 14.22,
+      "grad_norm": 3.187875270843506,
+      "learning_rate": 9.640582990024175e-06,
+      "loss": 0.7997,
+      "step": 47515
+    },
+    {
+      "epoch": 14.22,
+      "grad_norm": 2.1788666248321533,
+      "learning_rate": 9.635947797424266e-06,
+      "loss": 0.8929,
+      "step": 47520
+    },
+    {
+      "epoch": 14.22,
+      "grad_norm": 1.6329278945922852,
+      "learning_rate": 9.63131345336497e-06,
+      "loss": 0.9739,
+      "step": 47525
+    },
+    {
+      "epoch": 14.22,
+      "grad_norm": 4.640608787536621,
+      "learning_rate": 9.62667995810228e-06,
+      "loss": 0.8712,
+      "step": 47530
+    },
+    {
+      "epoch": 14.22,
+      "grad_norm": 2.9405198097229004,
+      "learning_rate": 9.622047311892055e-06,
+      "loss": 0.9474,
+      "step": 47535
+    },
+    {
+      "epoch": 14.22,
+      "grad_norm": 1.338927149772644,
+      "learning_rate": 9.61741551499019e-06,
+      "loss": 1.0767,
+      "step": 47540
+    },
+    {
+      "epoch": 14.22,
+      "grad_norm": 1.2258890867233276,
+      "learning_rate": 9.61278456765246e-06,
+      "loss": 1.1003,
+      "step": 47545
+    },
+    {
+      "epoch": 14.23,
+      "grad_norm": 1.6150643825531006,
+      "learning_rate": 9.608154470134665e-06,
+      "loss": 1.1114,
+      "step": 47550
+    },
+    {
+      "epoch": 14.23,
+      "grad_norm": 4.458968639373779,
+      "learning_rate": 9.603525222692486e-06,
+      "loss": 0.9945,
+      "step": 47555
+    },
+    {
+      "epoch": 14.23,
+      "grad_norm": 4.157359600067139,
+      "learning_rate": 9.598896825581607e-06,
+      "loss": 0.9888,
+      "step": 47560
+    },
+    {
+      "epoch": 14.23,
+      "grad_norm": 3.2777342796325684,
+      "learning_rate": 9.59426927905765e-06,
+      "loss": 0.7707,
+      "step": 47565
+    },
+    {
+      "epoch": 14.23,
+      "grad_norm": 3.0390613079071045,
+      "learning_rate": 9.589642583376186e-06,
+      "loss": 0.9935,
+      "step": 47570
+    },
+    {
+      "epoch": 14.23,
+      "grad_norm": 3.542185068130493,
+      "learning_rate": 9.58501673879274e-06,
+      "loss": 0.8352,
+      "step": 47575
+    },
+    {
+      "epoch": 14.24,
+      "grad_norm": 3.9005398750305176,
+      "learning_rate": 9.580391745562802e-06,
+      "loss": 1.0942,
+      "step": 47580
+    },
+    {
+      "epoch": 14.24,
+      "grad_norm": 3.451244592666626,
+      "learning_rate": 9.575767603941798e-06,
+      "loss": 0.9602,
+      "step": 47585
+    },
+    {
+      "epoch": 14.24,
+      "grad_norm": 1.035033941268921,
+      "learning_rate": 9.571144314185116e-06,
+      "loss": 1.1651,
+      "step": 47590
+    },
+    {
+      "epoch": 14.24,
+      "grad_norm": 3.617086410522461,
+      "learning_rate": 9.566521876548096e-06,
+      "loss": 0.9759,
+      "step": 47595
+    },
+    {
+      "epoch": 14.24,
+      "grad_norm": 3.698251485824585,
+      "learning_rate": 9.561900291286032e-06,
+      "loss": 0.9557,
+      "step": 47600
+    },
+    {
+      "epoch": 14.24,
+      "grad_norm": 0.981471061706543,
+      "learning_rate": 9.557279558654167e-06,
+      "loss": 0.8647,
+      "step": 47605
+    },
+    {
+      "epoch": 14.24,
+      "grad_norm": 1.0644017457962036,
+      "learning_rate": 9.552659678907699e-06,
+      "loss": 0.8904,
+      "step": 47610
+    },
+    {
+      "epoch": 14.25,
+      "grad_norm": 2.2468714714050293,
+      "learning_rate": 9.548040652301792e-06,
+      "loss": 1.0655,
+      "step": 47615
+    },
+    {
+      "epoch": 14.25,
+      "grad_norm": 3.204252004623413,
+      "learning_rate": 9.543422479091518e-06,
+      "loss": 1.083,
+      "step": 47620
+    },
+    {
+      "epoch": 14.25,
+      "grad_norm": 3.293745279312134,
+      "learning_rate": 9.538805159531974e-06,
+      "loss": 1.0888,
+      "step": 47625
+    },
+    {
+      "epoch": 14.25,
+      "grad_norm": 2.6639957427978516,
+      "learning_rate": 9.534188693878131e-06,
+      "loss": 1.0,
+      "step": 47630
+    },
+    {
+      "epoch": 14.25,
+      "grad_norm": 3.4556894302368164,
+      "learning_rate": 9.529573082384988e-06,
+      "loss": 1.0495,
+      "step": 47635
+    },
+    {
+      "epoch": 14.25,
+      "grad_norm": 2.608086347579956,
+      "learning_rate": 9.524958325307426e-06,
+      "loss": 0.9045,
+      "step": 47640
+    },
+    {
+      "epoch": 14.25,
+      "grad_norm": 3.2534093856811523,
+      "learning_rate": 9.520344422900348e-06,
+      "loss": 1.1092,
+      "step": 47645
+    },
+    {
+      "epoch": 14.26,
+      "grad_norm": 1.683605670928955,
+      "learning_rate": 9.515731375418549e-06,
+      "loss": 0.9562,
+      "step": 47650
+    },
+    {
+      "epoch": 14.26,
+      "grad_norm": 4.434023857116699,
+      "learning_rate": 9.511119183116812e-06,
+      "loss": 0.8168,
+      "step": 47655
+    },
+    {
+      "epoch": 14.26,
+      "grad_norm": 2.663111448287964,
+      "learning_rate": 9.506507846249859e-06,
+      "loss": 0.95,
+      "step": 47660
+    },
+    {
+      "epoch": 14.26,
+      "grad_norm": 3.199991226196289,
+      "learning_rate": 9.501897365072367e-06,
+      "loss": 0.9858,
+      "step": 47665
+    },
+    {
+      "epoch": 14.26,
+      "grad_norm": 2.4415783882141113,
+      "learning_rate": 9.497287739838992e-06,
+      "loss": 1.1561,
+      "step": 47670
+    },
+    {
+      "epoch": 14.26,
+      "grad_norm": 4.541195392608643,
+      "learning_rate": 9.492678970804283e-06,
+      "loss": 0.9719,
+      "step": 47675
+    },
+    {
+      "epoch": 14.27,
+      "grad_norm": 3.981882095336914,
+      "learning_rate": 9.488071058222814e-06,
+      "loss": 0.984,
+      "step": 47680
+    },
+    {
+      "epoch": 14.27,
+      "grad_norm": 7.917405128479004,
+      "learning_rate": 9.483464002349046e-06,
+      "loss": 1.0035,
+      "step": 47685
+    },
+    {
+      "epoch": 14.27,
+      "grad_norm": 2.864999771118164,
+      "learning_rate": 9.478857803437433e-06,
+      "loss": 0.9655,
+      "step": 47690
+    },
+    {
+      "epoch": 14.27,
+      "grad_norm": 3.0915513038635254,
+      "learning_rate": 9.474252461742373e-06,
+      "loss": 1.0438,
+      "step": 47695
+    },
+    {
+      "epoch": 14.27,
+      "grad_norm": 2.601701021194458,
+      "learning_rate": 9.469647977518207e-06,
+      "loss": 0.9761,
+      "step": 47700
+    },
+    {
+      "epoch": 14.27,
+      "grad_norm": 4.57741641998291,
+      "learning_rate": 9.465044351019243e-06,
+      "loss": 1.025,
+      "step": 47705
+    },
+    {
+      "epoch": 14.27,
+      "grad_norm": 4.853657245635986,
+      "learning_rate": 9.460441582499729e-06,
+      "loss": 1.1026,
+      "step": 47710
+    },
+    {
+      "epoch": 14.28,
+      "grad_norm": 3.169487476348877,
+      "learning_rate": 9.455839672213878e-06,
+      "loss": 0.9469,
+      "step": 47715
+    },
+    {
+      "epoch": 14.28,
+      "grad_norm": 2.6972315311431885,
+      "learning_rate": 9.451238620415851e-06,
+      "loss": 1.1723,
+      "step": 47720
+    },
+    {
+      "epoch": 14.28,
+      "grad_norm": 2.2301697731018066,
+      "learning_rate": 9.446638427359736e-06,
+      "loss": 0.8984,
+      "step": 47725
+    },
+    {
+      "epoch": 14.28,
+      "grad_norm": 2.9678409099578857,
+      "learning_rate": 9.442039093299631e-06,
+      "loss": 0.9494,
+      "step": 47730
+    },
+    {
+      "epoch": 14.28,
+      "grad_norm": 1.5792745351791382,
+      "learning_rate": 9.437440618489518e-06,
+      "loss": 1.1286,
+      "step": 47735
+    },
+    {
+      "epoch": 14.28,
+      "grad_norm": 2.416245222091675,
+      "learning_rate": 9.432843003183392e-06,
+      "loss": 0.9913,
+      "step": 47740
+    },
+    {
+      "epoch": 14.28,
+      "grad_norm": 2.8561947345733643,
+      "learning_rate": 9.428246247635176e-06,
+      "loss": 0.8804,
+      "step": 47745
+    },
+    {
+      "epoch": 14.29,
+      "grad_norm": 3.78583025932312,
+      "learning_rate": 9.423650352098712e-06,
+      "loss": 0.9608,
+      "step": 47750
+    },
+    {
+      "epoch": 14.29,
+      "grad_norm": 2.2284066677093506,
+      "learning_rate": 9.419055316827869e-06,
+      "loss": 0.9273,
+      "step": 47755
+    },
+    {
+      "epoch": 14.29,
+      "grad_norm": 2.1891860961914062,
+      "learning_rate": 9.41446114207639e-06,
+      "loss": 0.8688,
+      "step": 47760
+    },
+    {
+      "epoch": 14.29,
+      "grad_norm": 7.929460048675537,
+      "learning_rate": 9.409867828098035e-06,
+      "loss": 0.9384,
+      "step": 47765
+    },
+    {
+      "epoch": 14.29,
+      "grad_norm": 1.693267822265625,
+      "learning_rate": 9.405275375146458e-06,
+      "loss": 1.0242,
+      "step": 47770
+    },
+    {
+      "epoch": 14.29,
+      "grad_norm": 4.085503578186035,
+      "learning_rate": 9.40068378347533e-06,
+      "loss": 1.039,
+      "step": 47775
+    },
+    {
+      "epoch": 14.3,
+      "grad_norm": 2.429783821105957,
+      "learning_rate": 9.396093053338218e-06,
+      "loss": 1.0925,
+      "step": 47780
+    },
+    {
+      "epoch": 14.3,
+      "grad_norm": 1.4568767547607422,
+      "learning_rate": 9.391503184988661e-06,
+      "loss": 0.906,
+      "step": 47785
+    },
+    {
+      "epoch": 14.3,
+      "grad_norm": 2.0767431259155273,
+      "learning_rate": 9.386914178680162e-06,
+      "loss": 0.9838,
+      "step": 47790
+    },
+    {
+      "epoch": 14.3,
+      "grad_norm": 1.0575764179229736,
+      "learning_rate": 9.382326034666164e-06,
+      "loss": 0.983,
+      "step": 47795
+    },
+    {
+      "epoch": 14.3,
+      "grad_norm": 1.7569162845611572,
+      "learning_rate": 9.377738753200066e-06,
+      "loss": 0.9052,
+      "step": 47800
+    },
+    {
+      "epoch": 14.3,
+      "grad_norm": 1.5973174571990967,
+      "learning_rate": 9.373152334535216e-06,
+      "loss": 1.0071,
+      "step": 47805
+    },
+    {
+      "epoch": 14.3,
+      "grad_norm": 2.8834640979766846,
+      "learning_rate": 9.36856677892492e-06,
+      "loss": 0.9311,
+      "step": 47810
+    },
+    {
+      "epoch": 14.31,
+      "grad_norm": 1.4674243927001953,
+      "learning_rate": 9.363982086622442e-06,
+      "loss": 0.9815,
+      "step": 47815
+    },
+    {
+      "epoch": 14.31,
+      "grad_norm": 1.9711565971374512,
+      "learning_rate": 9.359398257880963e-06,
+      "loss": 0.9914,
+      "step": 47820
+    },
+    {
+      "epoch": 14.31,
+      "grad_norm": 3.0938401222229004,
+      "learning_rate": 9.35481529295367e-06,
+      "loss": 1.0849,
+      "step": 47825
+    },
+    {
+      "epoch": 14.31,
+      "grad_norm": 3.689774513244629,
+      "learning_rate": 9.350233192093667e-06,
+      "loss": 1.2623,
+      "step": 47830
+    },
+    {
+      "epoch": 14.31,
+      "grad_norm": 11.061504364013672,
+      "learning_rate": 9.345651955554016e-06,
+      "loss": 0.9138,
+      "step": 47835
+    },
+    {
+      "epoch": 14.31,
+      "grad_norm": 2.238264560699463,
+      "learning_rate": 9.341071583587745e-06,
+      "loss": 1.0129,
+      "step": 47840
+    },
+    {
+      "epoch": 14.31,
+      "grad_norm": 3.2770984172821045,
+      "learning_rate": 9.336492076447794e-06,
+      "loss": 0.8907,
+      "step": 47845
+    },
+    {
+      "epoch": 14.32,
+      "grad_norm": 2.24216890335083,
+      "learning_rate": 9.331913434387127e-06,
+      "loss": 1.0423,
+      "step": 47850
+    },
+    {
+      "epoch": 14.32,
+      "grad_norm": 1.8534948825836182,
+      "learning_rate": 9.327335657658573e-06,
+      "loss": 1.002,
+      "step": 47855
+    },
+    {
+      "epoch": 14.32,
+      "grad_norm": 3.518404006958008,
+      "learning_rate": 9.322758746515e-06,
+      "loss": 1.0224,
+      "step": 47860
+    },
+    {
+      "epoch": 14.32,
+      "grad_norm": 2.3978283405303955,
+      "learning_rate": 9.318182701209147e-06,
+      "loss": 1.194,
+      "step": 47865
+    },
+    {
+      "epoch": 14.32,
+      "grad_norm": 2.3987321853637695,
+      "learning_rate": 9.31360752199378e-06,
+      "loss": 0.983,
+      "step": 47870
+    },
+    {
+      "epoch": 14.32,
+      "grad_norm": 4.193480491638184,
+      "learning_rate": 9.309033209121556e-06,
+      "loss": 0.8785,
+      "step": 47875
+    },
+    {
+      "epoch": 14.33,
+      "grad_norm": 2.2980704307556152,
+      "learning_rate": 9.30445976284512e-06,
+      "loss": 1.0265,
+      "step": 47880
+    },
+    {
+      "epoch": 14.33,
+      "grad_norm": 2.2900497913360596,
+      "learning_rate": 9.299887183417055e-06,
+      "loss": 0.9558,
+      "step": 47885
+    },
+    {
+      "epoch": 14.33,
+      "grad_norm": 1.8453694581985474,
+      "learning_rate": 9.295315471089903e-06,
+      "loss": 1.0211,
+      "step": 47890
+    },
+    {
+      "epoch": 14.33,
+      "grad_norm": 4.543096542358398,
+      "learning_rate": 9.290744626116152e-06,
+      "loss": 1.0024,
+      "step": 47895
+    },
+    {
+      "epoch": 14.33,
+      "grad_norm": 4.047824859619141,
+      "learning_rate": 9.286174648748247e-06,
+      "loss": 1.0203,
+      "step": 47900
+    },
+    {
+      "epoch": 14.33,
+      "grad_norm": 2.266964912414551,
+      "learning_rate": 9.28160553923858e-06,
+      "loss": 0.9312,
+      "step": 47905
+    },
+    {
+      "epoch": 14.33,
+      "grad_norm": 3.8564565181732178,
+      "learning_rate": 9.277037297839506e-06,
+      "loss": 0.998,
+      "step": 47910
+    },
+    {
+      "epoch": 14.34,
+      "grad_norm": 4.338145732879639,
+      "learning_rate": 9.272469924803315e-06,
+      "loss": 0.8935,
+      "step": 47915
+    },
+    {
+      "epoch": 14.34,
+      "grad_norm": 2.9112906455993652,
+      "learning_rate": 9.267903420382262e-06,
+      "loss": 0.9426,
+      "step": 47920
+    },
+    {
+      "epoch": 14.34,
+      "grad_norm": 2.4785072803497314,
+      "learning_rate": 9.263337784828551e-06,
+      "loss": 1.1364,
+      "step": 47925
+    },
+    {
+      "epoch": 14.34,
+      "grad_norm": 2.290468692779541,
+      "learning_rate": 9.258773018394337e-06,
+      "loss": 0.8601,
+      "step": 47930
+    },
+    {
+      "epoch": 14.34,
+      "grad_norm": 1.839494228363037,
+      "learning_rate": 9.254209121331728e-06,
+      "loss": 1.0367,
+      "step": 47935
+    },
+    {
+      "epoch": 14.34,
+      "grad_norm": 2.684028387069702,
+      "learning_rate": 9.249646093892783e-06,
+      "loss": 1.0927,
+      "step": 47940
+    },
+    {
+      "epoch": 14.34,
+      "grad_norm": 13.804609298706055,
+      "learning_rate": 9.245083936329521e-06,
+      "loss": 1.0277,
+      "step": 47945
+    },
+    {
+      "epoch": 14.35,
+      "grad_norm": 1.9250423908233643,
+      "learning_rate": 9.24052264889388e-06,
+      "loss": 0.9594,
+      "step": 47950
+    },
+    {
+      "epoch": 14.35,
+      "grad_norm": 1.3630380630493164,
+      "learning_rate": 9.235962231837805e-06,
+      "loss": 0.9599,
+      "step": 47955
+    },
+    {
+      "epoch": 14.35,
+      "grad_norm": 3.4717135429382324,
+      "learning_rate": 9.231402685413137e-06,
+      "loss": 1.0512,
+      "step": 47960
+    },
+    {
+      "epoch": 14.35,
+      "grad_norm": 2.14388370513916,
+      "learning_rate": 9.226844009871723e-06,
+      "loss": 0.8996,
+      "step": 47965
+    },
+    {
+      "epoch": 14.35,
+      "grad_norm": 4.260863304138184,
+      "learning_rate": 9.222286205465314e-06,
+      "loss": 0.9731,
+      "step": 47970
+    },
+    {
+      "epoch": 14.35,
+      "grad_norm": 1.8017550706863403,
+      "learning_rate": 9.217729272445636e-06,
+      "loss": 0.8765,
+      "step": 47975
+    },
+    {
+      "epoch": 14.36,
+      "grad_norm": 2.0987396240234375,
+      "learning_rate": 9.213173211064364e-06,
+      "loss": 1.0251,
+      "step": 47980
+    },
+    {
+      "epoch": 14.36,
+      "grad_norm": 2.1268539428710938,
+      "learning_rate": 9.208618021573126e-06,
+      "loss": 0.9544,
+      "step": 47985
+    },
+    {
+      "epoch": 14.36,
+      "grad_norm": 4.74144172668457,
+      "learning_rate": 9.2040637042235e-06,
+      "loss": 0.9862,
+      "step": 47990
+    },
+    {
+      "epoch": 14.36,
+      "grad_norm": 2.23129940032959,
+      "learning_rate": 9.199510259267018e-06,
+      "loss": 1.0009,
+      "step": 47995
+    },
+    {
+      "epoch": 14.36,
+      "grad_norm": 2.542621612548828,
+      "learning_rate": 9.194957686955164e-06,
+      "loss": 0.9754,
+      "step": 48000
+    },
+    {
+      "epoch": 14.36,
+      "grad_norm": 2.1914637088775635,
+      "learning_rate": 9.190405987539363e-06,
+      "loss": 1.049,
+      "step": 48005
+    },
+    {
+      "epoch": 14.36,
+      "grad_norm": 5.300290107727051,
+      "learning_rate": 9.185855161271009e-06,
+      "loss": 1.0649,
+      "step": 48010
+    },
+    {
+      "epoch": 14.37,
+      "grad_norm": 3.934770107269287,
+      "learning_rate": 9.181305208401437e-06,
+      "loss": 0.9544,
+      "step": 48015
+    },
+    {
+      "epoch": 14.37,
+      "grad_norm": 2.419064521789551,
+      "learning_rate": 9.176756129181935e-06,
+      "loss": 0.9372,
+      "step": 48020
+    },
+    {
+      "epoch": 14.37,
+      "grad_norm": 2.2804276943206787,
+      "learning_rate": 9.172207923863746e-06,
+      "loss": 0.92,
+      "step": 48025
+    },
+    {
+      "epoch": 14.37,
+      "grad_norm": 4.2205400466918945,
+      "learning_rate": 9.167660592698058e-06,
+      "loss": 1.0041,
+      "step": 48030
+    },
+    {
+      "epoch": 14.37,
+      "grad_norm": 5.472626209259033,
+      "learning_rate": 9.163114135936022e-06,
+      "loss": 1.0433,
+      "step": 48035
+    },
+    {
+      "epoch": 14.37,
+      "grad_norm": 3.6936001777648926,
+      "learning_rate": 9.158568553828737e-06,
+      "loss": 1.1153,
+      "step": 48040
+    },
+    {
+      "epoch": 14.37,
+      "grad_norm": 2.1582984924316406,
+      "learning_rate": 9.154023846627227e-06,
+      "loss": 0.9568,
+      "step": 48045
+    },
+    {
+      "epoch": 14.38,
+      "grad_norm": 3.0322723388671875,
+      "learning_rate": 9.149480014582529e-06,
+      "loss": 1.1763,
+      "step": 48050
+    },
+    {
+      "epoch": 14.38,
+      "grad_norm": 1.92725670337677,
+      "learning_rate": 9.144937057945555e-06,
+      "loss": 0.9095,
+      "step": 48055
+    },
+    {
+      "epoch": 14.38,
+      "grad_norm": 1.9671467542648315,
+      "learning_rate": 9.140394976967246e-06,
+      "loss": 1.1279,
+      "step": 48060
+    },
+    {
+      "epoch": 14.38,
+      "grad_norm": 1.5574039220809937,
+      "learning_rate": 9.135853771898418e-06,
+      "loss": 1.1226,
+      "step": 48065
+    },
+    {
+      "epoch": 14.38,
+      "grad_norm": 2.8291425704956055,
+      "learning_rate": 9.131313442989914e-06,
+      "loss": 0.9447,
+      "step": 48070
+    },
+    {
+      "epoch": 14.38,
+      "grad_norm": 3.1782844066619873,
+      "learning_rate": 9.126773990492466e-06,
+      "loss": 0.9002,
+      "step": 48075
+    },
+    {
+      "epoch": 14.38,
+      "grad_norm": 1.8452153205871582,
+      "learning_rate": 9.122235414656788e-06,
+      "loss": 1.1123,
+      "step": 48080
+    },
+    {
+      "epoch": 14.39,
+      "grad_norm": 6.0874834060668945,
+      "learning_rate": 9.117697715733548e-06,
+      "loss": 0.9248,
+      "step": 48085
+    },
+    {
+      "epoch": 14.39,
+      "grad_norm": 2.1895172595977783,
+      "learning_rate": 9.113160893973352e-06,
+      "loss": 1.0326,
+      "step": 48090
+    },
+    {
+      "epoch": 14.39,
+      "grad_norm": 1.827024221420288,
+      "learning_rate": 9.108624949626767e-06,
+      "loss": 0.9664,
+      "step": 48095
+    },
+    {
+      "epoch": 14.39,
+      "grad_norm": 1.7563681602478027,
+      "learning_rate": 9.104089882944308e-06,
+      "loss": 0.9242,
+      "step": 48100
+    },
+    {
+      "epoch": 14.39,
+      "grad_norm": 4.570804119110107,
+      "learning_rate": 9.099555694176443e-06,
+      "loss": 1.0313,
+      "step": 48105
+    },
+    {
+      "epoch": 14.39,
+      "grad_norm": 3.724151849746704,
+      "learning_rate": 9.095022383573587e-06,
+      "loss": 1.0457,
+      "step": 48110
+    },
+    {
+      "epoch": 14.4,
+      "grad_norm": 1.7167346477508545,
+      "learning_rate": 9.090489951386114e-06,
+      "loss": 1.0725,
+      "step": 48115
+    },
+    {
+      "epoch": 14.4,
+      "grad_norm": 3.227980136871338,
+      "learning_rate": 9.085958397864344e-06,
+      "loss": 1.0361,
+      "step": 48120
+    },
+    {
+      "epoch": 14.4,
+      "grad_norm": 3.609764337539673,
+      "learning_rate": 9.081427723258552e-06,
+      "loss": 1.0083,
+      "step": 48125
+    },
+    {
+      "epoch": 14.4,
+      "grad_norm": 2.5227441787719727,
+      "learning_rate": 9.076897927818956e-06,
+      "loss": 1.0384,
+      "step": 48130
+    },
+    {
+      "epoch": 14.4,
+      "grad_norm": 2.815845251083374,
+      "learning_rate": 9.072369011795747e-06,
+      "loss": 1.0781,
+      "step": 48135
+    },
+    {
+      "epoch": 14.4,
+      "grad_norm": 2.0394959449768066,
+      "learning_rate": 9.067840975439026e-06,
+      "loss": 1.0609,
+      "step": 48140
+    },
+    {
+      "epoch": 14.4,
+      "grad_norm": 2.39359712600708,
+      "learning_rate": 9.063313818998903e-06,
+      "loss": 0.9759,
+      "step": 48145
+    },
+    {
+      "epoch": 14.41,
+      "grad_norm": 2.754600763320923,
+      "learning_rate": 9.058787542725375e-06,
+      "loss": 0.7847,
+      "step": 48150
+    },
+    {
+      "epoch": 14.41,
+      "grad_norm": 2.4412641525268555,
+      "learning_rate": 9.054262146868459e-06,
+      "loss": 1.2569,
+      "step": 48155
+    },
+    {
+      "epoch": 14.41,
+      "grad_norm": 3.2604317665100098,
+      "learning_rate": 9.049737631678055e-06,
+      "loss": 1.1841,
+      "step": 48160
+    },
+    {
+      "epoch": 14.41,
+      "grad_norm": 1.3231927156448364,
+      "learning_rate": 9.045213997404075e-06,
+      "loss": 0.9842,
+      "step": 48165
+    },
+    {
+      "epoch": 14.41,
+      "grad_norm": 1.9147082567214966,
+      "learning_rate": 9.040691244296335e-06,
+      "loss": 0.9722,
+      "step": 48170
+    },
+    {
+      "epoch": 14.41,
+      "grad_norm": 2.9659924507141113,
+      "learning_rate": 9.036169372604627e-06,
+      "loss": 1.0222,
+      "step": 48175
+    },
+    {
+      "epoch": 14.41,
+      "grad_norm": 2.9592783451080322,
+      "learning_rate": 9.031648382578695e-06,
+      "loss": 0.9292,
+      "step": 48180
+    },
+    {
+      "epoch": 14.42,
+      "grad_norm": 1.969634771347046,
+      "learning_rate": 9.027128274468221e-06,
+      "loss": 1.1377,
+      "step": 48185
+    },
+    {
+      "epoch": 14.42,
+      "grad_norm": 1.3826353549957275,
+      "learning_rate": 9.022609048522854e-06,
+      "loss": 1.149,
+      "step": 48190
+    },
+    {
+      "epoch": 14.42,
+      "grad_norm": 3.4229960441589355,
+      "learning_rate": 9.018090704992177e-06,
+      "loss": 1.0666,
+      "step": 48195
+    },
+    {
+      "epoch": 14.42,
+      "grad_norm": 9.103164672851562,
+      "learning_rate": 9.013573244125742e-06,
+      "loss": 0.9819,
+      "step": 48200
+    },
+    {
+      "epoch": 14.42,
+      "grad_norm": 1.6387892961502075,
+      "learning_rate": 9.009056666173039e-06,
+      "loss": 1.0427,
+      "step": 48205
+    },
+    {
+      "epoch": 14.42,
+      "grad_norm": 5.4994425773620605,
+      "learning_rate": 9.00454097138351e-06,
+      "loss": 0.9706,
+      "step": 48210
+    },
+    {
+      "epoch": 14.43,
+      "grad_norm": 1.8383941650390625,
+      "learning_rate": 9.000026160006561e-06,
+      "loss": 1.0697,
+      "step": 48215
+    },
+    {
+      "epoch": 14.43,
+      "grad_norm": 4.727527618408203,
+      "learning_rate": 8.995512232291537e-06,
+      "loss": 1.0467,
+      "step": 48220
+    },
+    {
+      "epoch": 14.43,
+      "grad_norm": 3.8419857025146484,
+      "learning_rate": 8.990999188487734e-06,
+      "loss": 0.7724,
+      "step": 48225
+    },
+    {
+      "epoch": 14.43,
+      "grad_norm": 1.8518542051315308,
+      "learning_rate": 8.986487028844415e-06,
+      "loss": 1.0587,
+      "step": 48230
+    },
+    {
+      "epoch": 14.43,
+      "grad_norm": 2.0114905834198,
+      "learning_rate": 8.981975753610756e-06,
+      "loss": 1.127,
+      "step": 48235
+    },
+    {
+      "epoch": 14.43,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.977465363035942e-06,
+      "loss": 1.2177,
+      "step": 48240
+    },
+    {
+      "epoch": 14.43,
+      "grad_norm": 4.793522834777832,
+      "learning_rate": 8.972955857369045e-06,
+      "loss": 0.9406,
+      "step": 48245
+    },
+    {
+      "epoch": 14.44,
+      "grad_norm": 3.620276927947998,
+      "learning_rate": 8.968447236859154e-06,
+      "loss": 0.8755,
+      "step": 48250
+    },
+    {
+      "epoch": 14.44,
+      "grad_norm": 4.639894962310791,
+      "learning_rate": 8.963939501755242e-06,
+      "loss": 1.0752,
+      "step": 48255
+    },
+    {
+      "epoch": 14.44,
+      "grad_norm": 1.244881510734558,
+      "learning_rate": 8.959432652306299e-06,
+      "loss": 0.8917,
+      "step": 48260
+    },
+    {
+      "epoch": 14.44,
+      "grad_norm": 3.2123985290527344,
+      "learning_rate": 8.954926688761208e-06,
+      "loss": 1.0363,
+      "step": 48265
+    },
+    {
+      "epoch": 14.44,
+      "grad_norm": 2.0880956649780273,
+      "learning_rate": 8.950421611368836e-06,
+      "loss": 0.9985,
+      "step": 48270
+    },
+    {
+      "epoch": 14.44,
+      "grad_norm": 5.486839771270752,
+      "learning_rate": 8.945917420378e-06,
+      "loss": 1.0174,
+      "step": 48275
+    },
+    {
+      "epoch": 14.44,
+      "grad_norm": 2.638864040374756,
+      "learning_rate": 8.941414116037458e-06,
+      "loss": 0.9216,
+      "step": 48280
+    },
+    {
+      "epoch": 14.45,
+      "grad_norm": 3.484980344772339,
+      "learning_rate": 8.936911698595921e-06,
+      "loss": 0.9612,
+      "step": 48285
+    },
+    {
+      "epoch": 14.45,
+      "grad_norm": 4.663930416107178,
+      "learning_rate": 8.932410168302052e-06,
+      "loss": 1.0834,
+      "step": 48290
+    },
+    {
+      "epoch": 14.45,
+      "grad_norm": 3.5646653175354004,
+      "learning_rate": 8.92790952540447e-06,
+      "loss": 0.9851,
+      "step": 48295
+    },
+    {
+      "epoch": 14.45,
+      "grad_norm": 3.214585304260254,
+      "learning_rate": 8.923409770151739e-06,
+      "loss": 0.9698,
+      "step": 48300
+    },
+    {
+      "epoch": 14.45,
+      "grad_norm": 1.7530304193496704,
+      "learning_rate": 8.918910902792377e-06,
+      "loss": 0.9638,
+      "step": 48305
+    },
+    {
+      "epoch": 14.45,
+      "grad_norm": 3.360421895980835,
+      "learning_rate": 8.914412923574848e-06,
+      "loss": 0.979,
+      "step": 48310
+    },
+    {
+      "epoch": 14.46,
+      "grad_norm": 2.919114589691162,
+      "learning_rate": 8.909915832747573e-06,
+      "loss": 0.9093,
+      "step": 48315
+    },
+    {
+      "epoch": 14.46,
+      "grad_norm": 1.921119213104248,
+      "learning_rate": 8.905419630558922e-06,
+      "loss": 0.9799,
+      "step": 48320
+    },
+    {
+      "epoch": 14.46,
+      "grad_norm": 3.1176345348358154,
+      "learning_rate": 8.900924317257226e-06,
+      "loss": 1.0288,
+      "step": 48325
+    },
+    {
+      "epoch": 14.46,
+      "grad_norm": 2.6858246326446533,
+      "learning_rate": 8.896429893090729e-06,
+      "loss": 1.0427,
+      "step": 48330
+    },
+    {
+      "epoch": 14.46,
+      "grad_norm": 30.713518142700195,
+      "learning_rate": 8.891936358307685e-06,
+      "loss": 1.0085,
+      "step": 48335
+    },
+    {
+      "epoch": 14.46,
+      "grad_norm": 4.43396520614624,
+      "learning_rate": 8.887443713156237e-06,
+      "loss": 0.8242,
+      "step": 48340
+    },
+    {
+      "epoch": 14.46,
+      "grad_norm": 3.269070625305176,
+      "learning_rate": 8.882951957884541e-06,
+      "loss": 1.0081,
+      "step": 48345
+    },
+    {
+      "epoch": 14.47,
+      "grad_norm": 3.0732762813568115,
+      "learning_rate": 8.87846109274064e-06,
+      "loss": 1.0627,
+      "step": 48350
+    },
+    {
+      "epoch": 14.47,
+      "grad_norm": 1.8434253931045532,
+      "learning_rate": 8.87397111797259e-06,
+      "loss": 0.9998,
+      "step": 48355
+    },
+    {
+      "epoch": 14.47,
+      "grad_norm": 1.8734230995178223,
+      "learning_rate": 8.869482033828347e-06,
+      "loss": 1.0262,
+      "step": 48360
+    },
+    {
+      "epoch": 14.47,
+      "grad_norm": 2.2515382766723633,
+      "learning_rate": 8.864993840555844e-06,
+      "loss": 1.006,
+      "step": 48365
+    },
+    {
+      "epoch": 14.47,
+      "grad_norm": 1.439144253730774,
+      "learning_rate": 8.86050653840296e-06,
+      "loss": 0.9979,
+      "step": 48370
+    },
+    {
+      "epoch": 14.47,
+      "grad_norm": 3.4755709171295166,
+      "learning_rate": 8.856020127617524e-06,
+      "loss": 0.8835,
+      "step": 48375
+    },
+    {
+      "epoch": 14.47,
+      "grad_norm": 2.903697967529297,
+      "learning_rate": 8.851534608447311e-06,
+      "loss": 1.0311,
+      "step": 48380
+    },
+    {
+      "epoch": 14.48,
+      "grad_norm": 2.571676731109619,
+      "learning_rate": 8.847049981140063e-06,
+      "loss": 0.9793,
+      "step": 48385
+    },
+    {
+      "epoch": 14.48,
+      "grad_norm": 2.435349702835083,
+      "learning_rate": 8.84256624594345e-06,
+      "loss": 0.858,
+      "step": 48390
+    },
+    {
+      "epoch": 14.48,
+      "grad_norm": 1.679379940032959,
+      "learning_rate": 8.83808340310511e-06,
+      "loss": 1.1234,
+      "step": 48395
+    },
+    {
+      "epoch": 14.48,
+      "grad_norm": 4.435678005218506,
+      "learning_rate": 8.833601452872625e-06,
+      "loss": 0.9629,
+      "step": 48400
+    },
+    {
+      "epoch": 14.48,
+      "grad_norm": 3.5596141815185547,
+      "learning_rate": 8.829120395493527e-06,
+      "loss": 1.2037,
+      "step": 48405
+    },
+    {
+      "epoch": 14.48,
+      "grad_norm": 4.797013759613037,
+      "learning_rate": 8.8246402312153e-06,
+      "loss": 1.1643,
+      "step": 48410
+    },
+    {
+      "epoch": 14.49,
+      "grad_norm": 2.486677885055542,
+      "learning_rate": 8.820160960285382e-06,
+      "loss": 1.0333,
+      "step": 48415
+    },
+    {
+      "epoch": 14.49,
+      "grad_norm": 1.6059032678604126,
+      "learning_rate": 8.815682582951163e-06,
+      "loss": 1.1628,
+      "step": 48420
+    },
+    {
+      "epoch": 14.49,
+      "grad_norm": 2.2335846424102783,
+      "learning_rate": 8.811205099459954e-06,
+      "loss": 0.9198,
+      "step": 48425
+    },
+    {
+      "epoch": 14.49,
+      "grad_norm": 2.2356197834014893,
+      "learning_rate": 8.806728510059078e-06,
+      "loss": 1.0291,
+      "step": 48430
+    },
+    {
+      "epoch": 14.49,
+      "grad_norm": 4.154965400695801,
+      "learning_rate": 8.802252814995738e-06,
+      "loss": 1.0695,
+      "step": 48435
+    },
+    {
+      "epoch": 14.49,
+      "grad_norm": 1.8840675354003906,
+      "learning_rate": 8.797778014517156e-06,
+      "loss": 1.07,
+      "step": 48440
+    },
+    {
+      "epoch": 14.49,
+      "grad_norm": 7.801070213317871,
+      "learning_rate": 8.793304108870434e-06,
+      "loss": 0.9854,
+      "step": 48445
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 2.220947027206421,
+      "learning_rate": 8.7888310983027e-06,
+      "loss": 1.0613,
+      "step": 48450
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 2.6094961166381836,
+      "learning_rate": 8.784358983060964e-06,
+      "loss": 0.9946,
+      "step": 48455
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 2.7676682472229004,
+      "learning_rate": 8.779887763392225e-06,
+      "loss": 1.0854,
+      "step": 48460
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 4.184953212738037,
+      "learning_rate": 8.775417439543427e-06,
+      "loss": 0.7902,
+      "step": 48465
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 2.03581166267395,
+      "learning_rate": 8.770948011761456e-06,
+      "loss": 1.0086,
+      "step": 48470
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 3.659592866897583,
+      "learning_rate": 8.76647948029316e-06,
+      "loss": 0.8873,
+      "step": 48475
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 4.103082656860352,
+      "learning_rate": 8.762011845385321e-06,
+      "loss": 1.1404,
+      "step": 48480
+    },
+    {
+      "epoch": 14.51,
+      "grad_norm": 2.476278066635132,
+      "learning_rate": 8.757545107284704e-06,
+      "loss": 1.0754,
+      "step": 48485
+    },
+    {
+      "epoch": 14.51,
+      "grad_norm": 3.953524112701416,
+      "learning_rate": 8.753079266237978e-06,
+      "loss": 0.9528,
+      "step": 48490
+    },
+    {
+      "epoch": 14.51,
+      "grad_norm": 1.5628539323806763,
+      "learning_rate": 8.748614322491799e-06,
+      "loss": 0.9617,
+      "step": 48495
+    },
+    {
+      "epoch": 14.51,
+      "grad_norm": 2.0729362964630127,
+      "learning_rate": 8.744150276292759e-06,
+      "loss": 1.0597,
+      "step": 48500
+    },
+    {
+      "epoch": 14.51,
+      "grad_norm": 3.8265740871429443,
+      "learning_rate": 8.739687127887399e-06,
+      "loss": 1.1671,
+      "step": 48505
+    },
+    {
+      "epoch": 14.51,
+      "grad_norm": 2.81380558013916,
+      "learning_rate": 8.73522487752222e-06,
+      "loss": 0.9367,
+      "step": 48510
+    },
+    {
+      "epoch": 14.52,
+      "grad_norm": 3.2788760662078857,
+      "learning_rate": 8.730763525443663e-06,
+      "loss": 1.0131,
+      "step": 48515
+    },
+    {
+      "epoch": 14.52,
+      "grad_norm": 1.7596465349197388,
+      "learning_rate": 8.726303071898128e-06,
+      "loss": 0.9415,
+      "step": 48520
+    },
+    {
+      "epoch": 14.52,
+      "grad_norm": 3.689640760421753,
+      "learning_rate": 8.721843517131967e-06,
+      "loss": 1.0165,
+      "step": 48525
+    },
+    {
+      "epoch": 14.52,
+      "grad_norm": 3.1340339183807373,
+      "learning_rate": 8.717384861391448e-06,
+      "loss": 0.9638,
+      "step": 48530
+    },
+    {
+      "epoch": 14.52,
+      "grad_norm": 3.0869131088256836,
+      "learning_rate": 8.712927104922859e-06,
+      "loss": 1.1088,
+      "step": 48535
+    },
+    {
+      "epoch": 14.52,
+      "grad_norm": 2.496973991394043,
+      "learning_rate": 8.708470247972359e-06,
+      "loss": 1.0464,
+      "step": 48540
+    },
+    {
+      "epoch": 14.52,
+      "grad_norm": 3.2096173763275146,
+      "learning_rate": 8.704014290786128e-06,
+      "loss": 0.9417,
+      "step": 48545
+    },
+    {
+      "epoch": 14.53,
+      "grad_norm": 3.4847817420959473,
+      "learning_rate": 8.699559233610242e-06,
+      "loss": 0.8683,
+      "step": 48550
+    },
+    {
+      "epoch": 14.53,
+      "grad_norm": 2.737156867980957,
+      "learning_rate": 8.695105076690745e-06,
+      "loss": 0.8797,
+      "step": 48555
+    },
+    {
+      "epoch": 14.53,
+      "grad_norm": 1.7218374013900757,
+      "learning_rate": 8.690651820273668e-06,
+      "loss": 1.0199,
+      "step": 48560
+    },
+    {
+      "epoch": 14.53,
+      "grad_norm": 2.990959882736206,
+      "learning_rate": 8.68619946460492e-06,
+      "loss": 1.1195,
+      "step": 48565
+    },
+    {
+      "epoch": 14.53,
+      "grad_norm": 2.1475510597229004,
+      "learning_rate": 8.681748009930433e-06,
+      "loss": 1.0958,
+      "step": 48570
+    },
+    {
+      "epoch": 14.53,
+      "grad_norm": 2.3086507320404053,
+      "learning_rate": 8.677297456496026e-06,
+      "loss": 0.9051,
+      "step": 48575
+    },
+    {
+      "epoch": 14.53,
+      "grad_norm": 1.2065564393997192,
+      "learning_rate": 8.672847804547533e-06,
+      "loss": 1.0571,
+      "step": 48580
+    },
+    {
+      "epoch": 14.54,
+      "grad_norm": 1.4163763523101807,
+      "learning_rate": 8.668399054330672e-06,
+      "loss": 1.0684,
+      "step": 48585
+    },
+    {
+      "epoch": 14.54,
+      "grad_norm": 3.3399434089660645,
+      "learning_rate": 8.66395120609116e-06,
+      "loss": 1.1536,
+      "step": 48590
+    },
+    {
+      "epoch": 14.54,
+      "grad_norm": 1.9370921850204468,
+      "learning_rate": 8.659504260074638e-06,
+      "loss": 1.0895,
+      "step": 48595
+    },
+    {
+      "epoch": 14.54,
+      "grad_norm": 2.0222585201263428,
+      "learning_rate": 8.655058216526712e-06,
+      "loss": 0.9891,
+      "step": 48600
+    },
+    {
+      "epoch": 14.54,
+      "grad_norm": 2.3010804653167725,
+      "learning_rate": 8.650613075692931e-06,
+      "loss": 0.9449,
+      "step": 48605
+    },
+    {
+      "epoch": 14.54,
+      "grad_norm": 2.6222808361053467,
+      "learning_rate": 8.646168837818797e-06,
+      "loss": 1.08,
+      "step": 48610
+    },
+    {
+      "epoch": 14.55,
+      "grad_norm": 1.6273561716079712,
+      "learning_rate": 8.641725503149762e-06,
+      "loss": 1.0818,
+      "step": 48615
+    },
+    {
+      "epoch": 14.55,
+      "grad_norm": 2.0088725090026855,
+      "learning_rate": 8.637283071931227e-06,
+      "loss": 0.9697,
+      "step": 48620
+    },
+    {
+      "epoch": 14.55,
+      "grad_norm": 1.362270474433899,
+      "learning_rate": 8.632841544408526e-06,
+      "loss": 1.1185,
+      "step": 48625
+    },
+    {
+      "epoch": 14.55,
+      "grad_norm": 2.453993797302246,
+      "learning_rate": 8.62840092082699e-06,
+      "loss": 1.0598,
+      "step": 48630
+    },
+    {
+      "epoch": 14.55,
+      "grad_norm": 2.08078932762146,
+      "learning_rate": 8.623961201431835e-06,
+      "loss": 0.938,
+      "step": 48635
+    },
+    {
+      "epoch": 14.55,
+      "grad_norm": 2.452207565307617,
+      "learning_rate": 8.619522386468292e-06,
+      "loss": 0.896,
+      "step": 48640
+    },
+    {
+      "epoch": 14.55,
+      "grad_norm": 8.122790336608887,
+      "learning_rate": 8.6150844761815e-06,
+      "loss": 1.0177,
+      "step": 48645
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 2.5997066497802734,
+      "learning_rate": 8.61064747081656e-06,
+      "loss": 1.0295,
+      "step": 48650
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 3.2175190448760986,
+      "learning_rate": 8.606211370618537e-06,
+      "loss": 1.0199,
+      "step": 48655
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 2.0975887775421143,
+      "learning_rate": 8.601776175832399e-06,
+      "loss": 0.9438,
+      "step": 48660
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 2.3262147903442383,
+      "learning_rate": 8.597341886703134e-06,
+      "loss": 1.0345,
+      "step": 48665
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 1.7775872945785522,
+      "learning_rate": 8.592908503475614e-06,
+      "loss": 1.1137,
+      "step": 48670
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 3.6084024906158447,
+      "learning_rate": 8.588476026394716e-06,
+      "loss": 1.0825,
+      "step": 48675
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 3.329972982406616,
+      "learning_rate": 8.584044455705223e-06,
+      "loss": 1.1109,
+      "step": 48680
+    },
+    {
+      "epoch": 14.57,
+      "grad_norm": 1.461459994316101,
+      "learning_rate": 8.579613791651889e-06,
+      "loss": 0.7492,
+      "step": 48685
+    },
+    {
+      "epoch": 14.57,
+      "grad_norm": 3.0383195877075195,
+      "learning_rate": 8.575184034479416e-06,
+      "loss": 1.0271,
+      "step": 48690
+    },
+    {
+      "epoch": 14.57,
+      "grad_norm": 1.4644750356674194,
+      "learning_rate": 8.57075518443246e-06,
+      "loss": 1.0969,
+      "step": 48695
+    },
+    {
+      "epoch": 14.57,
+      "grad_norm": 2.943840503692627,
+      "learning_rate": 8.566327241755617e-06,
+      "loss": 1.1267,
+      "step": 48700
+    },
+    {
+      "epoch": 14.57,
+      "grad_norm": 4.192392349243164,
+      "learning_rate": 8.561900206693437e-06,
+      "loss": 0.8901,
+      "step": 48705
+    },
+    {
+      "epoch": 14.57,
+      "grad_norm": 1.5619021654129028,
+      "learning_rate": 8.557474079490421e-06,
+      "loss": 1.1158,
+      "step": 48710
+    },
+    {
+      "epoch": 14.57,
+      "grad_norm": 5.234612941741943,
+      "learning_rate": 8.553048860391025e-06,
+      "loss": 0.8096,
+      "step": 48715
+    },
+    {
+      "epoch": 14.58,
+      "grad_norm": 4.29495906829834,
+      "learning_rate": 8.548624549639641e-06,
+      "loss": 0.9903,
+      "step": 48720
+    },
+    {
+      "epoch": 14.58,
+      "grad_norm": 3.1692752838134766,
+      "learning_rate": 8.544201147480625e-06,
+      "loss": 1.1217,
+      "step": 48725
+    },
+    {
+      "epoch": 14.58,
+      "grad_norm": 1.6975233554840088,
+      "learning_rate": 8.539778654158274e-06,
+      "loss": 1.0455,
+      "step": 48730
+    },
+    {
+      "epoch": 14.58,
+      "grad_norm": 3.1467435359954834,
+      "learning_rate": 8.535357069916839e-06,
+      "loss": 0.7778,
+      "step": 48735
+    },
+    {
+      "epoch": 14.58,
+      "grad_norm": 2.9374170303344727,
+      "learning_rate": 8.530936395000517e-06,
+      "loss": 0.939,
+      "step": 48740
+    },
+    {
+      "epoch": 14.58,
+      "grad_norm": 3.6052603721618652,
+      "learning_rate": 8.52651662965346e-06,
+      "loss": 0.9546,
+      "step": 48745
+    },
+    {
+      "epoch": 14.59,
+      "grad_norm": 2.1928467750549316,
+      "learning_rate": 8.522097774119775e-06,
+      "loss": 1.1193,
+      "step": 48750
+    },
+    {
+      "epoch": 14.59,
+      "grad_norm": 3.1543407440185547,
+      "learning_rate": 8.517679828643485e-06,
+      "loss": 0.8595,
+      "step": 48755
+    },
+    {
+      "epoch": 14.59,
+      "grad_norm": 2.9056456089019775,
+      "learning_rate": 8.513262793468623e-06,
+      "loss": 0.9613,
+      "step": 48760
+    },
+    {
+      "epoch": 14.59,
+      "grad_norm": 2.9297358989715576,
+      "learning_rate": 8.508846668839104e-06,
+      "loss": 0.976,
+      "step": 48765
+    },
+    {
+      "epoch": 14.59,
+      "grad_norm": 2.0978548526763916,
+      "learning_rate": 8.504431454998856e-06,
+      "loss": 0.9262,
+      "step": 48770
+    },
+    {
+      "epoch": 14.59,
+      "grad_norm": 1.7425466775894165,
+      "learning_rate": 8.500017152191708e-06,
+      "loss": 1.0549,
+      "step": 48775
+    },
+    {
+      "epoch": 14.59,
+      "grad_norm": 3.82523250579834,
+      "learning_rate": 8.495603760661459e-06,
+      "loss": 1.0423,
+      "step": 48780
+    },
+    {
+      "epoch": 14.6,
+      "grad_norm": 3.274580717086792,
+      "learning_rate": 8.491191280651861e-06,
+      "loss": 0.8973,
+      "step": 48785
+    },
+    {
+      "epoch": 14.6,
+      "grad_norm": 8.596813201904297,
+      "learning_rate": 8.486779712406605e-06,
+      "loss": 0.9724,
+      "step": 48790
+    },
+    {
+      "epoch": 14.6,
+      "grad_norm": 2.4201295375823975,
+      "learning_rate": 8.482369056169345e-06,
+      "loss": 0.9512,
+      "step": 48795
+    },
+    {
+      "epoch": 14.6,
+      "grad_norm": 4.427967071533203,
+      "learning_rate": 8.47795931218367e-06,
+      "loss": 0.9629,
+      "step": 48800
+    },
+    {
+      "epoch": 14.6,
+      "grad_norm": 3.9436044692993164,
+      "learning_rate": 8.47355048069313e-06,
+      "loss": 0.9762,
+      "step": 48805
+    },
+    {
+      "epoch": 14.6,
+      "grad_norm": 4.806298732757568,
+      "learning_rate": 8.469142561941218e-06,
+      "loss": 1.0321,
+      "step": 48810
+    },
+    {
+      "epoch": 14.6,
+      "grad_norm": 1.625897765159607,
+      "learning_rate": 8.46473555617138e-06,
+      "loss": 1.068,
+      "step": 48815
+    },
+    {
+      "epoch": 14.61,
+      "grad_norm": 2.3483619689941406,
+      "learning_rate": 8.46032946362701e-06,
+      "loss": 1.077,
+      "step": 48820
+    },
+    {
+      "epoch": 14.61,
+      "grad_norm": 3.5805463790893555,
+      "learning_rate": 8.455924284551453e-06,
+      "loss": 1.0664,
+      "step": 48825
+    },
+    {
+      "epoch": 14.61,
+      "grad_norm": 1.450373649597168,
+      "learning_rate": 8.451520019187997e-06,
+      "loss": 1.1001,
+      "step": 48830
+    },
+    {
+      "epoch": 14.61,
+      "grad_norm": 2.2781779766082764,
+      "learning_rate": 8.447116667779892e-06,
+      "loss": 0.9847,
+      "step": 48835
+    },
+    {
+      "epoch": 14.61,
+      "grad_norm": 3.301300048828125,
+      "learning_rate": 8.442714230570328e-06,
+      "loss": 0.8831,
+      "step": 48840
+    },
+    {
+      "epoch": 14.61,
+      "grad_norm": 2.685093879699707,
+      "learning_rate": 8.438312707802453e-06,
+      "loss": 0.8498,
+      "step": 48845
+    },
+    {
+      "epoch": 14.62,
+      "grad_norm": 4.1189069747924805,
+      "learning_rate": 8.433912099719337e-06,
+      "loss": 0.9296,
+      "step": 48850
+    },
+    {
+      "epoch": 14.62,
+      "grad_norm": 2.040025234222412,
+      "learning_rate": 8.429512406564055e-06,
+      "loss": 1.0176,
+      "step": 48855
+    },
+    {
+      "epoch": 14.62,
+      "grad_norm": 2.4712417125701904,
+      "learning_rate": 8.42511362857956e-06,
+      "loss": 0.9175,
+      "step": 48860
+    },
+    {
+      "epoch": 14.62,
+      "grad_norm": 2.281320810317993,
+      "learning_rate": 8.420715766008826e-06,
+      "loss": 0.8373,
+      "step": 48865
+    },
+    {
+      "epoch": 14.62,
+      "grad_norm": 2.7027089595794678,
+      "learning_rate": 8.416318819094713e-06,
+      "loss": 0.9954,
+      "step": 48870
+    },
+    {
+      "epoch": 14.62,
+      "grad_norm": 2.729827880859375,
+      "learning_rate": 8.411922788080088e-06,
+      "loss": 0.8774,
+      "step": 48875
+    },
+    {
+      "epoch": 14.62,
+      "grad_norm": 1.878912329673767,
+      "learning_rate": 8.407527673207718e-06,
+      "loss": 0.9414,
+      "step": 48880
+    },
+    {
+      "epoch": 14.63,
+      "grad_norm": 2.9538416862487793,
+      "learning_rate": 8.403133474720348e-06,
+      "loss": 0.9427,
+      "step": 48885
+    },
+    {
+      "epoch": 14.63,
+      "grad_norm": 1.521445393562317,
+      "learning_rate": 8.398740192860663e-06,
+      "loss": 0.9906,
+      "step": 48890
+    },
+    {
+      "epoch": 14.63,
+      "grad_norm": 2.062879800796509,
+      "learning_rate": 8.394347827871302e-06,
+      "loss": 0.8437,
+      "step": 48895
+    },
+    {
+      "epoch": 14.63,
+      "grad_norm": 3.722719669342041,
+      "learning_rate": 8.38995637999485e-06,
+      "loss": 1.0068,
+      "step": 48900
+    },
+    {
+      "epoch": 14.63,
+      "grad_norm": 3.8463211059570312,
+      "learning_rate": 8.385565849473843e-06,
+      "loss": 0.9874,
+      "step": 48905
+    },
+    {
+      "epoch": 14.63,
+      "grad_norm": 2.462646484375,
+      "learning_rate": 8.381176236550764e-06,
+      "loss": 1.0429,
+      "step": 48910
+    },
+    {
+      "epoch": 14.63,
+      "grad_norm": 2.3265976905822754,
+      "learning_rate": 8.376787541468045e-06,
+      "loss": 1.1793,
+      "step": 48915
+    },
+    {
+      "epoch": 14.64,
+      "grad_norm": 4.777642726898193,
+      "learning_rate": 8.372399764468076e-06,
+      "loss": 0.9058,
+      "step": 48920
+    },
+    {
+      "epoch": 14.64,
+      "grad_norm": 2.799137592315674,
+      "learning_rate": 8.36801290579318e-06,
+      "loss": 0.9833,
+      "step": 48925
+    },
+    {
+      "epoch": 14.64,
+      "grad_norm": 1.032945156097412,
+      "learning_rate": 8.363626965685645e-06,
+      "loss": 1.1707,
+      "step": 48930
+    },
+    {
+      "epoch": 14.64,
+      "grad_norm": 2.207265853881836,
+      "learning_rate": 8.359241944387699e-06,
+      "loss": 0.9743,
+      "step": 48935
+    },
+    {
+      "epoch": 14.64,
+      "grad_norm": 2.7506961822509766,
+      "learning_rate": 8.354857842141533e-06,
+      "loss": 0.9348,
+      "step": 48940
+    },
+    {
+      "epoch": 14.64,
+      "grad_norm": 1.6580191850662231,
+      "learning_rate": 8.35047465918925e-06,
+      "loss": 1.0558,
+      "step": 48945
+    },
+    {
+      "epoch": 14.65,
+      "grad_norm": 4.469571590423584,
+      "learning_rate": 8.34609239577296e-06,
+      "loss": 1.0138,
+      "step": 48950
+    },
+    {
+      "epoch": 14.65,
+      "grad_norm": 2.21535325050354,
+      "learning_rate": 8.341711052134663e-06,
+      "loss": 1.0124,
+      "step": 48955
+    },
+    {
+      "epoch": 14.65,
+      "grad_norm": 5.841526031494141,
+      "learning_rate": 8.337330628516363e-06,
+      "loss": 1.0061,
+      "step": 48960
+    },
+    {
+      "epoch": 14.65,
+      "grad_norm": 2.309290647506714,
+      "learning_rate": 8.332951125159957e-06,
+      "loss": 0.9943,
+      "step": 48965
+    },
+    {
+      "epoch": 14.65,
+      "grad_norm": 5.338931560516357,
+      "learning_rate": 8.328572542307353e-06,
+      "loss": 0.9796,
+      "step": 48970
+    },
+    {
+      "epoch": 14.65,
+      "grad_norm": 4.652040958404541,
+      "learning_rate": 8.324194880200348e-06,
+      "loss": 0.7701,
+      "step": 48975
+    },
+    {
+      "epoch": 14.65,
+      "grad_norm": 4.739422798156738,
+      "learning_rate": 8.319818139080728e-06,
+      "loss": 1.0848,
+      "step": 48980
+    },
+    {
+      "epoch": 14.66,
+      "grad_norm": 1.5917638540267944,
+      "learning_rate": 8.315442319190214e-06,
+      "loss": 0.9494,
+      "step": 48985
+    },
+    {
+      "epoch": 14.66,
+      "grad_norm": 1.3649824857711792,
+      "learning_rate": 8.31106742077048e-06,
+      "loss": 1.1455,
+      "step": 48990
+    },
+    {
+      "epoch": 14.66,
+      "grad_norm": 1.3612852096557617,
+      "learning_rate": 8.30669344406314e-06,
+      "loss": 0.9826,
+      "step": 48995
+    },
+    {
+      "epoch": 14.66,
+      "grad_norm": 4.043242454528809,
+      "learning_rate": 8.302320389309776e-06,
+      "loss": 0.8944,
+      "step": 49000
+    },
+    {
+      "epoch": 14.66,
+      "grad_norm": 4.382472991943359,
+      "learning_rate": 8.297948256751897e-06,
+      "loss": 0.9066,
+      "step": 49005
+    },
+    {
+      "epoch": 14.66,
+      "grad_norm": 1.3883436918258667,
+      "learning_rate": 8.293577046630973e-06,
+      "loss": 1.0056,
+      "step": 49010
+    },
+    {
+      "epoch": 14.66,
+      "grad_norm": 2.2432875633239746,
+      "learning_rate": 8.289206759188426e-06,
+      "loss": 1.1799,
+      "step": 49015
+    },
+    {
+      "epoch": 14.67,
+      "grad_norm": 2.0689566135406494,
+      "learning_rate": 8.28483739466562e-06,
+      "loss": 0.9735,
+      "step": 49020
+    },
+    {
+      "epoch": 14.67,
+      "grad_norm": 2.059446096420288,
+      "learning_rate": 8.280468953303868e-06,
+      "loss": 0.9371,
+      "step": 49025
+    },
+    {
+      "epoch": 14.67,
+      "grad_norm": 2.9364001750946045,
+      "learning_rate": 8.276101435344441e-06,
+      "loss": 1.0071,
+      "step": 49030
+    },
+    {
+      "epoch": 14.67,
+      "grad_norm": 2.926053524017334,
+      "learning_rate": 8.271734841028553e-06,
+      "loss": 0.8205,
+      "step": 49035
+    },
+    {
+      "epoch": 14.67,
+      "grad_norm": 1.4417191743850708,
+      "learning_rate": 8.267369170597345e-06,
+      "loss": 0.9981,
+      "step": 49040
+    },
+    {
+      "epoch": 14.67,
+      "grad_norm": 6.5991973876953125,
+      "learning_rate": 8.263004424291962e-06,
+      "loss": 0.7934,
+      "step": 49045
+    },
+    {
+      "epoch": 14.68,
+      "grad_norm": 3.2373931407928467,
+      "learning_rate": 8.258640602353432e-06,
+      "loss": 0.9667,
+      "step": 49050
+    },
+    {
+      "epoch": 14.68,
+      "grad_norm": 2.341061592102051,
+      "learning_rate": 8.254277705022795e-06,
+      "loss": 1.0543,
+      "step": 49055
+    },
+    {
+      "epoch": 14.68,
+      "grad_norm": 2.1643905639648438,
+      "learning_rate": 8.24991573254098e-06,
+      "loss": 0.9737,
+      "step": 49060
+    },
+    {
+      "epoch": 14.68,
+      "grad_norm": 4.566475868225098,
+      "learning_rate": 8.245554685148924e-06,
+      "loss": 0.9937,
+      "step": 49065
+    },
+    {
+      "epoch": 14.68,
+      "grad_norm": 3.2137701511383057,
+      "learning_rate": 8.241194563087456e-06,
+      "loss": 1.0407,
+      "step": 49070
+    },
+    {
+      "epoch": 14.68,
+      "grad_norm": 1.9922786951065063,
+      "learning_rate": 8.236835366597397e-06,
+      "loss": 1.1503,
+      "step": 49075
+    },
+    {
+      "epoch": 14.68,
+      "grad_norm": 6.491326808929443,
+      "learning_rate": 8.232477095919495e-06,
+      "loss": 0.8255,
+      "step": 49080
+    },
+    {
+      "epoch": 14.69,
+      "grad_norm": 3.8574929237365723,
+      "learning_rate": 8.228119751294452e-06,
+      "loss": 0.8979,
+      "step": 49085
+    },
+    {
+      "epoch": 14.69,
+      "grad_norm": 1.933752417564392,
+      "learning_rate": 8.223763332962925e-06,
+      "loss": 0.992,
+      "step": 49090
+    },
+    {
+      "epoch": 14.69,
+      "grad_norm": 1.47145676612854,
+      "learning_rate": 8.219407841165508e-06,
+      "loss": 1.1588,
+      "step": 49095
+    },
+    {
+      "epoch": 14.69,
+      "grad_norm": 1.7857855558395386,
+      "learning_rate": 8.215053276142756e-06,
+      "loss": 0.9984,
+      "step": 49100
+    },
+    {
+      "epoch": 14.69,
+      "grad_norm": 3.7101316452026367,
+      "learning_rate": 8.210699638135164e-06,
+      "loss": 0.933,
+      "step": 49105
+    },
+    {
+      "epoch": 14.69,
+      "grad_norm": 3.045780658721924,
+      "learning_rate": 8.20634692738318e-06,
+      "loss": 0.8924,
+      "step": 49110
+    },
+    {
+      "epoch": 14.69,
+      "grad_norm": 4.745621204376221,
+      "learning_rate": 8.201995144127198e-06,
+      "loss": 0.9093,
+      "step": 49115
+    },
+    {
+      "epoch": 14.7,
+      "grad_norm": 1.8852388858795166,
+      "learning_rate": 8.197644288607562e-06,
+      "loss": 1.1749,
+      "step": 49120
+    },
+    {
+      "epoch": 14.7,
+      "grad_norm": 2.5533406734466553,
+      "learning_rate": 8.193294361064569e-06,
+      "loss": 1.1027,
+      "step": 49125
+    },
+    {
+      "epoch": 14.7,
+      "grad_norm": 1.9613593816757202,
+      "learning_rate": 8.188945361738468e-06,
+      "loss": 0.9726,
+      "step": 49130
+    },
+    {
+      "epoch": 14.7,
+      "grad_norm": 2.30924916267395,
+      "learning_rate": 8.184597290869423e-06,
+      "loss": 1.0676,
+      "step": 49135
+    },
+    {
+      "epoch": 14.7,
+      "grad_norm": 7.164217472076416,
+      "learning_rate": 8.180250148697605e-06,
+      "loss": 0.9149,
+      "step": 49140
+    },
+    {
+      "epoch": 14.7,
+      "grad_norm": 1.656010627746582,
+      "learning_rate": 8.175903935463076e-06,
+      "loss": 1.0562,
+      "step": 49145
+    },
+    {
+      "epoch": 14.71,
+      "grad_norm": 2.542595863342285,
+      "learning_rate": 8.171558651405897e-06,
+      "loss": 1.0277,
+      "step": 49150
+    },
+    {
+      "epoch": 14.71,
+      "grad_norm": 2.8253142833709717,
+      "learning_rate": 8.16721429676603e-06,
+      "loss": 1.0583,
+      "step": 49155
+    },
+    {
+      "epoch": 14.71,
+      "grad_norm": 4.2701945304870605,
+      "learning_rate": 8.162870871783435e-06,
+      "loss": 0.9444,
+      "step": 49160
+    },
+    {
+      "epoch": 14.71,
+      "grad_norm": 3.9655089378356934,
+      "learning_rate": 8.15852837669797e-06,
+      "loss": 0.9712,
+      "step": 49165
+    },
+    {
+      "epoch": 14.71,
+      "grad_norm": 3.8803188800811768,
+      "learning_rate": 8.154186811749479e-06,
+      "loss": 0.8939,
+      "step": 49170
+    },
+    {
+      "epoch": 14.71,
+      "grad_norm": 3.9758951663970947,
+      "learning_rate": 8.14984617717774e-06,
+      "loss": 0.9559,
+      "step": 49175
+    },
+    {
+      "epoch": 14.71,
+      "grad_norm": 1.9809141159057617,
+      "learning_rate": 8.145506473222481e-06,
+      "loss": 1.0336,
+      "step": 49180
+    },
+    {
+      "epoch": 14.72,
+      "grad_norm": 3.15104079246521,
+      "learning_rate": 8.14116770012338e-06,
+      "loss": 1.1105,
+      "step": 49185
+    },
+    {
+      "epoch": 14.72,
+      "grad_norm": 3.349479913711548,
+      "learning_rate": 8.136829858120066e-06,
+      "loss": 1.0803,
+      "step": 49190
+    },
+    {
+      "epoch": 14.72,
+      "grad_norm": 3.6429026126861572,
+      "learning_rate": 8.132492947452108e-06,
+      "loss": 0.9848,
+      "step": 49195
+    },
+    {
+      "epoch": 14.72,
+      "grad_norm": 2.910015106201172,
+      "learning_rate": 8.128156968359035e-06,
+      "loss": 0.9665,
+      "step": 49200
+    },
+    {
+      "epoch": 14.72,
+      "grad_norm": 6.490506172180176,
+      "learning_rate": 8.123821921080313e-06,
+      "loss": 0.8624,
+      "step": 49205
+    },
+    {
+      "epoch": 14.72,
+      "grad_norm": 1.679452896118164,
+      "learning_rate": 8.119487805855364e-06,
+      "loss": 0.9264,
+      "step": 49210
+    },
+    {
+      "epoch": 14.72,
+      "grad_norm": 5.132246971130371,
+      "learning_rate": 8.115154622923556e-06,
+      "loss": 0.8382,
+      "step": 49215
+    },
+    {
+      "epoch": 14.73,
+      "grad_norm": 1.5223355293273926,
+      "learning_rate": 8.11082237252421e-06,
+      "loss": 1.04,
+      "step": 49220
+    },
+    {
+      "epoch": 14.73,
+      "grad_norm": 2.475334644317627,
+      "learning_rate": 8.10649105489659e-06,
+      "loss": 1.0027,
+      "step": 49225
+    },
+    {
+      "epoch": 14.73,
+      "grad_norm": 1.9048575162887573,
+      "learning_rate": 8.102160670279906e-06,
+      "loss": 1.0384,
+      "step": 49230
+    },
+    {
+      "epoch": 14.73,
+      "grad_norm": 2.135935068130493,
+      "learning_rate": 8.097831218913333e-06,
+      "loss": 1.0354,
+      "step": 49235
+    },
+    {
+      "epoch": 14.73,
+      "grad_norm": 1.7497014999389648,
+      "learning_rate": 8.093502701035957e-06,
+      "loss": 0.8898,
+      "step": 49240
+    },
+    {
+      "epoch": 14.73,
+      "grad_norm": 2.735699415206909,
+      "learning_rate": 8.08917511688687e-06,
+      "loss": 0.9877,
+      "step": 49245
+    },
+    {
+      "epoch": 14.74,
+      "grad_norm": 1.427288293838501,
+      "learning_rate": 8.084848466705048e-06,
+      "loss": 0.8119,
+      "step": 49250
+    },
+    {
+      "epoch": 14.74,
+      "grad_norm": 3.9717910289764404,
+      "learning_rate": 8.080522750729477e-06,
+      "loss": 0.9547,
+      "step": 49255
+    },
+    {
+      "epoch": 14.74,
+      "grad_norm": 4.33355188369751,
+      "learning_rate": 8.07619796919904e-06,
+      "loss": 1.0372,
+      "step": 49260
+    },
+    {
+      "epoch": 14.74,
+      "grad_norm": 4.518667697906494,
+      "learning_rate": 8.071874122352598e-06,
+      "loss": 0.9114,
+      "step": 49265
+    },
+    {
+      "epoch": 14.74,
+      "grad_norm": 3.398543119430542,
+      "learning_rate": 8.067551210428952e-06,
+      "loss": 0.9405,
+      "step": 49270
+    },
+    {
+      "epoch": 14.74,
+      "grad_norm": 4.380087852478027,
+      "learning_rate": 8.063229233666854e-06,
+      "loss": 0.9915,
+      "step": 49275
+    },
+    {
+      "epoch": 14.74,
+      "grad_norm": 1.5817692279815674,
+      "learning_rate": 8.058908192305e-06,
+      "loss": 1.1618,
+      "step": 49280
+    },
+    {
+      "epoch": 14.75,
+      "grad_norm": 2.916731834411621,
+      "learning_rate": 8.054588086582029e-06,
+      "loss": 0.9583,
+      "step": 49285
+    },
+    {
+      "epoch": 14.75,
+      "grad_norm": 1.46005117893219,
+      "learning_rate": 8.050268916736561e-06,
+      "loss": 0.9385,
+      "step": 49290
+    },
+    {
+      "epoch": 14.75,
+      "grad_norm": 4.392984390258789,
+      "learning_rate": 8.045950683007115e-06,
+      "loss": 0.9728,
+      "step": 49295
+    },
+    {
+      "epoch": 14.75,
+      "grad_norm": 1.823644995689392,
+      "learning_rate": 8.041633385632186e-06,
+      "loss": 1.1487,
+      "step": 49300
+    },
+    {
+      "epoch": 14.75,
+      "grad_norm": 5.119616508483887,
+      "learning_rate": 8.037317024850222e-06,
+      "loss": 1.0758,
+      "step": 49305
+    },
+    {
+      "epoch": 14.75,
+      "grad_norm": 5.690764904022217,
+      "learning_rate": 8.033001600899606e-06,
+      "loss": 0.8963,
+      "step": 49310
+    },
+    {
+      "epoch": 14.75,
+      "grad_norm": 3.6248810291290283,
+      "learning_rate": 8.028687114018674e-06,
+      "loss": 0.8885,
+      "step": 49315
+    },
+    {
+      "epoch": 14.76,
+      "grad_norm": 2.2621960639953613,
+      "learning_rate": 8.024373564445714e-06,
+      "loss": 1.0696,
+      "step": 49320
+    },
+    {
+      "epoch": 14.76,
+      "grad_norm": 4.6661601066589355,
+      "learning_rate": 8.020060952418956e-06,
+      "loss": 1.1138,
+      "step": 49325
+    },
+    {
+      "epoch": 14.76,
+      "grad_norm": 2.248114824295044,
+      "learning_rate": 8.01574927817659e-06,
+      "loss": 1.1691,
+      "step": 49330
+    },
+    {
+      "epoch": 14.76,
+      "grad_norm": 1.4611231088638306,
+      "learning_rate": 8.01143854195672e-06,
+      "loss": 1.1069,
+      "step": 49335
+    },
+    {
+      "epoch": 14.76,
+      "grad_norm": 1.7469176054000854,
+      "learning_rate": 8.007128743997457e-06,
+      "loss": 1.0006,
+      "step": 49340
+    },
+    {
+      "epoch": 14.76,
+      "grad_norm": 1.546489953994751,
+      "learning_rate": 8.002819884536797e-06,
+      "loss": 0.9871,
+      "step": 49345
+    },
+    {
+      "epoch": 14.76,
+      "grad_norm": 3.91237473487854,
+      "learning_rate": 7.998511963812741e-06,
+      "loss": 0.9547,
+      "step": 49350
+    },
+    {
+      "epoch": 14.77,
+      "grad_norm": 2.6893093585968018,
+      "learning_rate": 7.99420498206318e-06,
+      "loss": 0.81,
+      "step": 49355
+    },
+    {
+      "epoch": 14.77,
+      "grad_norm": 3.268592119216919,
+      "learning_rate": 7.989898939526018e-06,
+      "loss": 1.152,
+      "step": 49360
+    },
+    {
+      "epoch": 14.77,
+      "grad_norm": 3.8002939224243164,
+      "learning_rate": 7.985593836439051e-06,
+      "loss": 1.1162,
+      "step": 49365
+    },
+    {
+      "epoch": 14.77,
+      "grad_norm": 2.8874077796936035,
+      "learning_rate": 7.981289673040041e-06,
+      "loss": 0.97,
+      "step": 49370
+    },
+    {
+      "epoch": 14.77,
+      "grad_norm": 3.9415621757507324,
+      "learning_rate": 7.976986449566728e-06,
+      "loss": 1.2578,
+      "step": 49375
+    },
+    {
+      "epoch": 14.77,
+      "grad_norm": 1.1208429336547852,
+      "learning_rate": 7.972684166256744e-06,
+      "loss": 0.9146,
+      "step": 49380
+    },
+    {
+      "epoch": 14.78,
+      "grad_norm": 4.455554962158203,
+      "learning_rate": 7.968382823347731e-06,
+      "loss": 0.9225,
+      "step": 49385
+    },
+    {
+      "epoch": 14.78,
+      "grad_norm": 3.3006980419158936,
+      "learning_rate": 7.964082421077223e-06,
+      "loss": 0.92,
+      "step": 49390
+    },
+    {
+      "epoch": 14.78,
+      "grad_norm": 1.741047739982605,
+      "learning_rate": 7.959782959682732e-06,
+      "loss": 0.919,
+      "step": 49395
+    },
+    {
+      "epoch": 14.78,
+      "grad_norm": 4.698997497558594,
+      "learning_rate": 7.95548443940172e-06,
+      "loss": 1.0252,
+      "step": 49400
+    },
+    {
+      "epoch": 14.78,
+      "grad_norm": 2.376089334487915,
+      "learning_rate": 7.951186860471582e-06,
+      "loss": 0.995,
+      "step": 49405
+    },
+    {
+      "epoch": 14.78,
+      "grad_norm": 3.1121981143951416,
+      "learning_rate": 7.946890223129677e-06,
+      "loss": 0.9083,
+      "step": 49410
+    },
+    {
+      "epoch": 14.78,
+      "grad_norm": 3.4917523860931396,
+      "learning_rate": 7.942594527613295e-06,
+      "loss": 0.8519,
+      "step": 49415
+    },
+    {
+      "epoch": 14.79,
+      "grad_norm": 6.6114420890808105,
+      "learning_rate": 7.938299774159691e-06,
+      "loss": 0.8482,
+      "step": 49420
+    },
+    {
+      "epoch": 14.79,
+      "grad_norm": 1.4963136911392212,
+      "learning_rate": 7.934005963006062e-06,
+      "loss": 0.9212,
+      "step": 49425
+    },
+    {
+      "epoch": 14.79,
+      "grad_norm": 3.31788969039917,
+      "learning_rate": 7.929713094389527e-06,
+      "loss": 1.1518,
+      "step": 49430
+    },
+    {
+      "epoch": 14.79,
+      "grad_norm": 2.473154067993164,
+      "learning_rate": 7.92542116854721e-06,
+      "loss": 0.7843,
+      "step": 49435
+    },
+    {
+      "epoch": 14.79,
+      "grad_norm": 2.462738513946533,
+      "learning_rate": 7.921130185716119e-06,
+      "loss": 0.8878,
+      "step": 49440
+    },
+    {
+      "epoch": 14.79,
+      "grad_norm": 3.9160032272338867,
+      "learning_rate": 7.916840146133264e-06,
+      "loss": 1.0054,
+      "step": 49445
+    },
+    {
+      "epoch": 14.79,
+      "grad_norm": 1.9631022214889526,
+      "learning_rate": 7.912551050035572e-06,
+      "loss": 1.0096,
+      "step": 49450
+    },
+    {
+      "epoch": 14.8,
+      "grad_norm": 3.000981330871582,
+      "learning_rate": 7.908262897659921e-06,
+      "loss": 1.051,
+      "step": 49455
+    },
+    {
+      "epoch": 14.8,
+      "grad_norm": 3.195512056350708,
+      "learning_rate": 7.903975689243155e-06,
+      "loss": 1.085,
+      "step": 49460
+    },
+    {
+      "epoch": 14.8,
+      "grad_norm": 1.471376657485962,
+      "learning_rate": 7.899689425022022e-06,
+      "loss": 1.0268,
+      "step": 49465
+    },
+    {
+      "epoch": 14.8,
+      "grad_norm": 2.2439353466033936,
+      "learning_rate": 7.895404105233286e-06,
+      "loss": 0.9531,
+      "step": 49470
+    },
+    {
+      "epoch": 14.8,
+      "grad_norm": 3.2079246044158936,
+      "learning_rate": 7.891119730113586e-06,
+      "loss": 0.9474,
+      "step": 49475
+    },
+    {
+      "epoch": 14.8,
+      "grad_norm": 2.1932594776153564,
+      "learning_rate": 7.886836299899574e-06,
+      "loss": 0.8452,
+      "step": 49480
+    },
+    {
+      "epoch": 14.81,
+      "grad_norm": 2.9864261150360107,
+      "learning_rate": 7.882553814827797e-06,
+      "loss": 1.0548,
+      "step": 49485
+    },
+    {
+      "epoch": 14.81,
+      "grad_norm": 1.2952245473861694,
+      "learning_rate": 7.878272275134782e-06,
+      "loss": 1.072,
+      "step": 49490
+    },
+    {
+      "epoch": 14.81,
+      "grad_norm": 2.7455410957336426,
+      "learning_rate": 7.873991681056992e-06,
+      "loss": 1.242,
+      "step": 49495
+    },
+    {
+      "epoch": 14.81,
+      "grad_norm": 2.4569225311279297,
+      "learning_rate": 7.86971203283084e-06,
+      "loss": 0.9538,
+      "step": 49500
+    },
+    {
+      "epoch": 14.81,
+      "grad_norm": 6.187790870666504,
+      "learning_rate": 7.865433330692684e-06,
+      "loss": 0.872,
+      "step": 49505
+    },
+    {
+      "epoch": 14.81,
+      "grad_norm": 2.7457752227783203,
+      "learning_rate": 7.861155574878838e-06,
+      "loss": 1.0486,
+      "step": 49510
+    },
+    {
+      "epoch": 14.81,
+      "grad_norm": 3.150146245956421,
+      "learning_rate": 7.856878765625553e-06,
+      "loss": 0.9142,
+      "step": 49515
+    },
+    {
+      "epoch": 14.82,
+      "grad_norm": 25.611284255981445,
+      "learning_rate": 7.852602903169043e-06,
+      "loss": 0.9913,
+      "step": 49520
+    },
+    {
+      "epoch": 14.82,
+      "grad_norm": 5.1431145668029785,
+      "learning_rate": 7.848327987745433e-06,
+      "loss": 0.7851,
+      "step": 49525
+    },
+    {
+      "epoch": 14.82,
+      "grad_norm": 5.0381388664245605,
+      "learning_rate": 7.844054019590851e-06,
+      "loss": 0.9707,
+      "step": 49530
+    },
+    {
+      "epoch": 14.82,
+      "grad_norm": 2.5118184089660645,
+      "learning_rate": 7.839780998941331e-06,
+      "loss": 0.9747,
+      "step": 49535
+    },
+    {
+      "epoch": 14.82,
+      "grad_norm": 3.1850192546844482,
+      "learning_rate": 7.83550892603287e-06,
+      "loss": 0.9729,
+      "step": 49540
+    },
+    {
+      "epoch": 14.82,
+      "grad_norm": 2.785214424133301,
+      "learning_rate": 7.83123780110141e-06,
+      "loss": 0.9847,
+      "step": 49545
+    },
+    {
+      "epoch": 14.82,
+      "grad_norm": 2.283158540725708,
+      "learning_rate": 7.826967624382839e-06,
+      "loss": 1.0142,
+      "step": 49550
+    },
+    {
+      "epoch": 14.83,
+      "grad_norm": 2.654142379760742,
+      "learning_rate": 7.822698396113005e-06,
+      "loss": 1.1584,
+      "step": 49555
+    },
+    {
+      "epoch": 14.83,
+      "grad_norm": 1.418227195739746,
+      "learning_rate": 7.818430116527668e-06,
+      "loss": 0.7142,
+      "step": 49560
+    },
+    {
+      "epoch": 14.83,
+      "grad_norm": 3.0038633346557617,
+      "learning_rate": 7.814162785862591e-06,
+      "loss": 1.0606,
+      "step": 49565
+    },
+    {
+      "epoch": 14.83,
+      "grad_norm": 3.01194429397583,
+      "learning_rate": 7.809896404353426e-06,
+      "loss": 0.9423,
+      "step": 49570
+    },
+    {
+      "epoch": 14.83,
+      "grad_norm": 3.334562063217163,
+      "learning_rate": 7.805630972235827e-06,
+      "loss": 0.9911,
+      "step": 49575
+    },
+    {
+      "epoch": 14.83,
+      "grad_norm": 4.44184684753418,
+      "learning_rate": 7.801366489745343e-06,
+      "loss": 0.8377,
+      "step": 49580
+    },
+    {
+      "epoch": 14.84,
+      "grad_norm": 2.8854176998138428,
+      "learning_rate": 7.797102957117527e-06,
+      "loss": 1.0517,
+      "step": 49585
+    },
+    {
+      "epoch": 14.84,
+      "grad_norm": 1.7378528118133545,
+      "learning_rate": 7.792840374587826e-06,
+      "loss": 0.9528,
+      "step": 49590
+    },
+    {
+      "epoch": 14.84,
+      "grad_norm": 4.007412910461426,
+      "learning_rate": 7.788578742391664e-06,
+      "loss": 0.9821,
+      "step": 49595
+    },
+    {
+      "epoch": 14.84,
+      "grad_norm": 5.08059549331665,
+      "learning_rate": 7.784318060764406e-06,
+      "loss": 0.9414,
+      "step": 49600
+    },
+    {
+      "epoch": 14.84,
+      "grad_norm": 1.6905008554458618,
+      "learning_rate": 7.78005832994137e-06,
+      "loss": 0.9635,
+      "step": 49605
+    },
+    {
+      "epoch": 14.84,
+      "grad_norm": 1.3702781200408936,
+      "learning_rate": 7.775799550157811e-06,
+      "loss": 1.0004,
+      "step": 49610
+    },
+    {
+      "epoch": 14.84,
+      "grad_norm": 2.688145875930786,
+      "learning_rate": 7.771541721648943e-06,
+      "loss": 1.0286,
+      "step": 49615
+    },
+    {
+      "epoch": 14.85,
+      "grad_norm": 2.3341612815856934,
+      "learning_rate": 7.767284844649914e-06,
+      "loss": 1.0534,
+      "step": 49620
+    },
+    {
+      "epoch": 14.85,
+      "grad_norm": 3.92333984375,
+      "learning_rate": 7.763028919395832e-06,
+      "loss": 1.0172,
+      "step": 49625
+    },
+    {
+      "epoch": 14.85,
+      "grad_norm": 3.7512214183807373,
+      "learning_rate": 7.758773946121745e-06,
+      "loss": 1.1164,
+      "step": 49630
+    },
+    {
+      "epoch": 14.85,
+      "grad_norm": 2.6205503940582275,
+      "learning_rate": 7.754519925062651e-06,
+      "loss": 0.9167,
+      "step": 49635
+    },
+    {
+      "epoch": 14.85,
+      "grad_norm": 4.808284282684326,
+      "learning_rate": 7.750266856453498e-06,
+      "loss": 1.1432,
+      "step": 49640
+    },
+    {
+      "epoch": 14.85,
+      "grad_norm": 3.2909045219421387,
+      "learning_rate": 7.746014740529176e-06,
+      "loss": 1.1521,
+      "step": 49645
+    },
+    {
+      "epoch": 14.85,
+      "grad_norm": 4.219728946685791,
+      "learning_rate": 7.741763577524532e-06,
+      "loss": 0.9124,
+      "step": 49650
+    },
+    {
+      "epoch": 14.86,
+      "grad_norm": 3.6862242221832275,
+      "learning_rate": 7.737513367674331e-06,
+      "loss": 1.0414,
+      "step": 49655
+    },
+    {
+      "epoch": 14.86,
+      "grad_norm": 2.3704283237457275,
+      "learning_rate": 7.73326411121334e-06,
+      "loss": 1.0735,
+      "step": 49660
+    },
+    {
+      "epoch": 14.86,
+      "grad_norm": 2.672393321990967,
+      "learning_rate": 7.729015808376208e-06,
+      "loss": 0.9972,
+      "step": 49665
+    },
+    {
+      "epoch": 14.86,
+      "grad_norm": 3.06726336479187,
+      "learning_rate": 7.724768459397597e-06,
+      "loss": 0.9701,
+      "step": 49670
+    },
+    {
+      "epoch": 14.86,
+      "grad_norm": 1.9261653423309326,
+      "learning_rate": 7.720522064512054e-06,
+      "loss": 1.0339,
+      "step": 49675
+    },
+    {
+      "epoch": 14.86,
+      "grad_norm": 4.520108222961426,
+      "learning_rate": 7.716276623954127e-06,
+      "loss": 0.8752,
+      "step": 49680
+    },
+    {
+      "epoch": 14.87,
+      "grad_norm": 6.763810157775879,
+      "learning_rate": 7.712032137958273e-06,
+      "loss": 1.0836,
+      "step": 49685
+    },
+    {
+      "epoch": 14.87,
+      "grad_norm": 2.1053125858306885,
+      "learning_rate": 7.707788606758912e-06,
+      "loss": 0.8817,
+      "step": 49690
+    },
+    {
+      "epoch": 14.87,
+      "grad_norm": 1.749535083770752,
+      "learning_rate": 7.703546030590414e-06,
+      "loss": 0.8303,
+      "step": 49695
+    },
+    {
+      "epoch": 14.87,
+      "grad_norm": 2.739657163619995,
+      "learning_rate": 7.699304409687089e-06,
+      "loss": 0.9729,
+      "step": 49700
+    },
+    {
+      "epoch": 14.87,
+      "grad_norm": 3.52274489402771,
+      "learning_rate": 7.695063744283196e-06,
+      "loss": 0.9307,
+      "step": 49705
+    },
+    {
+      "epoch": 14.87,
+      "grad_norm": 2.7961769104003906,
+      "learning_rate": 7.690824034612948e-06,
+      "loss": 0.991,
+      "step": 49710
+    },
+    {
+      "epoch": 14.87,
+      "grad_norm": 2.9371488094329834,
+      "learning_rate": 7.686585280910497e-06,
+      "loss": 0.999,
+      "step": 49715
+    },
+    {
+      "epoch": 14.88,
+      "grad_norm": 2.072927236557007,
+      "learning_rate": 7.68234748340994e-06,
+      "loss": 0.9193,
+      "step": 49720
+    },
+    {
+      "epoch": 14.88,
+      "grad_norm": 1.750504732131958,
+      "learning_rate": 7.678110642345334e-06,
+      "loss": 0.8694,
+      "step": 49725
+    },
+    {
+      "epoch": 14.88,
+      "grad_norm": 8.674440383911133,
+      "learning_rate": 7.673874757950675e-06,
+      "loss": 0.8633,
+      "step": 49730
+    },
+    {
+      "epoch": 14.88,
+      "grad_norm": 2.6912498474121094,
+      "learning_rate": 7.6696398304599e-06,
+      "loss": 1.0709,
+      "step": 49735
+    },
+    {
+      "epoch": 14.88,
+      "grad_norm": 4.30122184753418,
+      "learning_rate": 7.665405860106902e-06,
+      "loss": 0.8949,
+      "step": 49740
+    },
+    {
+      "epoch": 14.88,
+      "grad_norm": 2.0600149631500244,
+      "learning_rate": 7.66117284712553e-06,
+      "loss": 0.9048,
+      "step": 49745
+    },
+    {
+      "epoch": 14.88,
+      "grad_norm": 3.0497703552246094,
+      "learning_rate": 7.656940791749545e-06,
+      "loss": 0.933,
+      "step": 49750
+    },
+    {
+      "epoch": 14.89,
+      "grad_norm": 4.7076311111450195,
+      "learning_rate": 7.652709694212706e-06,
+      "loss": 0.8802,
+      "step": 49755
+    },
+    {
+      "epoch": 14.89,
+      "grad_norm": 2.6163299083709717,
+      "learning_rate": 7.648479554748666e-06,
+      "loss": 1.0973,
+      "step": 49760
+    },
+    {
+      "epoch": 14.89,
+      "grad_norm": 6.505290508270264,
+      "learning_rate": 7.644250373591078e-06,
+      "loss": 0.9932,
+      "step": 49765
+    },
+    {
+      "epoch": 14.89,
+      "grad_norm": 3.616623878479004,
+      "learning_rate": 7.640022150973485e-06,
+      "loss": 0.8964,
+      "step": 49770
+    },
+    {
+      "epoch": 14.89,
+      "grad_norm": 3.170760154724121,
+      "learning_rate": 7.635794887129441e-06,
+      "loss": 1.1569,
+      "step": 49775
+    },
+    {
+      "epoch": 14.89,
+      "grad_norm": 5.005249500274658,
+      "learning_rate": 7.631568582292389e-06,
+      "loss": 0.9821,
+      "step": 49780
+    },
+    {
+      "epoch": 14.9,
+      "grad_norm": 3.093559503555298,
+      "learning_rate": 7.62734323669575e-06,
+      "loss": 1.0313,
+      "step": 49785
+    },
+    {
+      "epoch": 14.9,
+      "grad_norm": 5.463796138763428,
+      "learning_rate": 7.623118850572886e-06,
+      "loss": 1.0058,
+      "step": 49790
+    },
+    {
+      "epoch": 14.9,
+      "grad_norm": 2.4942896366119385,
+      "learning_rate": 7.618895424157105e-06,
+      "loss": 0.9905,
+      "step": 49795
+    },
+    {
+      "epoch": 14.9,
+      "grad_norm": 3.5765671730041504,
+      "learning_rate": 7.614672957681665e-06,
+      "loss": 0.9716,
+      "step": 49800
+    },
+    {
+      "epoch": 14.9,
+      "grad_norm": 3.6634316444396973,
+      "learning_rate": 7.610451451379763e-06,
+      "loss": 1.0055,
+      "step": 49805
+    },
+    {
+      "epoch": 14.9,
+      "grad_norm": 1.4103604555130005,
+      "learning_rate": 7.606230905484557e-06,
+      "loss": 1.0418,
+      "step": 49810
+    },
+    {
+      "epoch": 14.9,
+      "grad_norm": 5.684398651123047,
+      "learning_rate": 7.602011320229132e-06,
+      "loss": 1.0942,
+      "step": 49815
+    },
+    {
+      "epoch": 14.91,
+      "grad_norm": 1.4336367845535278,
+      "learning_rate": 7.597792695846542e-06,
+      "loss": 1.0598,
+      "step": 49820
+    },
+    {
+      "epoch": 14.91,
+      "grad_norm": 3.313039779663086,
+      "learning_rate": 7.593575032569772e-06,
+      "loss": 1.0155,
+      "step": 49825
+    },
+    {
+      "epoch": 14.91,
+      "grad_norm": 4.043692111968994,
+      "learning_rate": 7.58935833063176e-06,
+      "loss": 0.8216,
+      "step": 49830
+    },
+    {
+      "epoch": 14.91,
+      "grad_norm": 1.758624792098999,
+      "learning_rate": 7.58514259026539e-06,
+      "loss": 1.0355,
+      "step": 49835
+    },
+    {
+      "epoch": 14.91,
+      "grad_norm": 2.282357931137085,
+      "learning_rate": 7.5809278117035006e-06,
+      "loss": 1.1196,
+      "step": 49840
+    },
+    {
+      "epoch": 14.91,
+      "grad_norm": 2.790658950805664,
+      "learning_rate": 7.576713995178847e-06,
+      "loss": 1.0328,
+      "step": 49845
+    },
+    {
+      "epoch": 14.91,
+      "grad_norm": 7.928509712219238,
+      "learning_rate": 7.572501140924184e-06,
+      "loss": 0.8045,
+      "step": 49850
+    },
+    {
+      "epoch": 14.92,
+      "grad_norm": 1.6898220777511597,
+      "learning_rate": 7.568289249172153e-06,
+      "loss": 0.9176,
+      "step": 49855
+    },
+    {
+      "epoch": 14.92,
+      "grad_norm": 1.0324536561965942,
+      "learning_rate": 7.5640783201554046e-06,
+      "loss": 1.0659,
+      "step": 49860
+    },
+    {
+      "epoch": 14.92,
+      "grad_norm": 2.4962313175201416,
+      "learning_rate": 7.5598683541064665e-06,
+      "loss": 1.0566,
+      "step": 49865
+    },
+    {
+      "epoch": 14.92,
+      "grad_norm": 4.477339744567871,
+      "learning_rate": 7.55565935125789e-06,
+      "loss": 0.9315,
+      "step": 49870
+    },
+    {
+      "epoch": 14.92,
+      "grad_norm": 1.8302847146987915,
+      "learning_rate": 7.551451311842109e-06,
+      "loss": 0.9124,
+      "step": 49875
+    },
+    {
+      "epoch": 14.92,
+      "grad_norm": 4.733458995819092,
+      "learning_rate": 7.547244236091533e-06,
+      "loss": 1.0675,
+      "step": 49880
+    },
+    {
+      "epoch": 14.93,
+      "grad_norm": 3.320866584777832,
+      "learning_rate": 7.543038124238516e-06,
+      "loss": 1.038,
+      "step": 49885
+    },
+    {
+      "epoch": 14.93,
+      "grad_norm": 1.8176350593566895,
+      "learning_rate": 7.5388329765153585e-06,
+      "loss": 1.1704,
+      "step": 49890
+    },
+    {
+      "epoch": 14.93,
+      "grad_norm": 3.144420623779297,
+      "learning_rate": 7.534628793154308e-06,
+      "loss": 0.948,
+      "step": 49895
+    },
+    {
+      "epoch": 14.93,
+      "grad_norm": 2.1764090061187744,
+      "learning_rate": 7.530425574387554e-06,
+      "loss": 0.9405,
+      "step": 49900
+    },
+    {
+      "epoch": 14.93,
+      "grad_norm": 2.065609931945801,
+      "learning_rate": 7.526223320447234e-06,
+      "loss": 1.005,
+      "step": 49905
+    },
+    {
+      "epoch": 14.93,
+      "grad_norm": 2.02253794670105,
+      "learning_rate": 7.52202203156544e-06,
+      "loss": 0.9092,
+      "step": 49910
+    },
+    {
+      "epoch": 14.93,
+      "grad_norm": 2.1637492179870605,
+      "learning_rate": 7.517821707974202e-06,
+      "loss": 0.8773,
+      "step": 49915
+    },
+    {
+      "epoch": 14.94,
+      "grad_norm": 10.103838920593262,
+      "learning_rate": 7.513622349905497e-06,
+      "loss": 0.9454,
+      "step": 49920
+    },
+    {
+      "epoch": 14.94,
+      "grad_norm": 2.852343797683716,
+      "learning_rate": 7.509423957591255e-06,
+      "loss": 1.1144,
+      "step": 49925
+    },
+    {
+      "epoch": 14.94,
+      "grad_norm": 3.75535249710083,
+      "learning_rate": 7.505226531263349e-06,
+      "loss": 1.0971,
+      "step": 49930
+    },
+    {
+      "epoch": 14.94,
+      "grad_norm": 2.4124093055725098,
+      "learning_rate": 7.501030071153594e-06,
+      "loss": 0.9609,
+      "step": 49935
+    },
+    {
+      "epoch": 14.94,
+      "grad_norm": 3.048701286315918,
+      "learning_rate": 7.496834577493761e-06,
+      "loss": 0.9755,
+      "step": 49940
+    },
+    {
+      "epoch": 14.94,
+      "grad_norm": 2.259594202041626,
+      "learning_rate": 7.492640050515567e-06,
+      "loss": 1.0239,
+      "step": 49945
+    },
+    {
+      "epoch": 14.94,
+      "grad_norm": 2.8681416511535645,
+      "learning_rate": 7.488446490450651e-06,
+      "loss": 0.9469,
+      "step": 49950
+    },
+    {
+      "epoch": 14.95,
+      "grad_norm": 3.7870285511016846,
+      "learning_rate": 7.484253897530649e-06,
+      "loss": 1.0347,
+      "step": 49955
+    },
+    {
+      "epoch": 14.95,
+      "grad_norm": 1.4063019752502441,
+      "learning_rate": 7.480900519694567e-06,
+      "loss": 0.9334,
+      "step": 49960
+    },
+    {
+      "epoch": 14.95,
+      "grad_norm": 2.873671293258667,
+      "learning_rate": 7.476709668218851e-06,
+      "loss": 1.0212,
+      "step": 49965
+    },
+    {
+      "epoch": 14.95,
+      "grad_norm": 4.571763515472412,
+      "learning_rate": 7.472519784536242e-06,
+      "loss": 0.9162,
+      "step": 49970
+    },
+    {
+      "epoch": 14.95,
+      "grad_norm": 2.2711031436920166,
+      "learning_rate": 7.468330868878149e-06,
+      "loss": 0.9683,
+      "step": 49975
+    },
+    {
+      "epoch": 14.95,
+      "grad_norm": 2.531252384185791,
+      "learning_rate": 7.464142921475919e-06,
+      "loss": 1.0777,
+      "step": 49980
+    },
+    {
+      "epoch": 14.95,
+      "grad_norm": 2.323441982269287,
+      "learning_rate": 7.459955942560848e-06,
+      "loss": 1.1148,
+      "step": 49985
+    },
+    {
+      "epoch": 14.96,
+      "grad_norm": 3.1378073692321777,
+      "learning_rate": 7.455769932364185e-06,
+      "loss": 1.0343,
+      "step": 49990
+    },
+    {
+      "epoch": 14.96,
+      "grad_norm": 3.8024823665618896,
+      "learning_rate": 7.4515848911170975e-06,
+      "loss": 0.9482,
+      "step": 49995
+    },
+    {
+      "epoch": 14.96,
+      "grad_norm": 1.5496140718460083,
+      "learning_rate": 7.447400819050751e-06,
+      "loss": 0.9271,
+      "step": 50000
+    },
+    {
+      "epoch": 14.96,
+      "grad_norm": 3.12982177734375,
+      "learning_rate": 7.443217716396198e-06,
+      "loss": 0.9457,
+      "step": 50005
+    },
+    {
+      "epoch": 14.96,
+      "grad_norm": 1.8738044500350952,
+      "learning_rate": 7.439035583384496e-06,
+      "loss": 1.1598,
+      "step": 50010
+    },
+    {
+      "epoch": 14.96,
+      "grad_norm": 5.965039253234863,
+      "learning_rate": 7.43485442024659e-06,
+      "loss": 0.8773,
+      "step": 50015
+    },
+    {
+      "epoch": 14.97,
+      "grad_norm": 4.203342914581299,
+      "learning_rate": 7.43067422721343e-06,
+      "loss": 0.7135,
+      "step": 50020
+    },
+    {
+      "epoch": 14.97,
+      "grad_norm": 3.4606893062591553,
+      "learning_rate": 7.426495004515865e-06,
+      "loss": 0.9092,
+      "step": 50025
+    },
+    {
+      "epoch": 14.97,
+      "grad_norm": 6.534817218780518,
+      "learning_rate": 7.422316752384711e-06,
+      "loss": 1.0587,
+      "step": 50030
+    },
+    {
+      "epoch": 14.97,
+      "grad_norm": 2.5901870727539062,
+      "learning_rate": 7.418139471050736e-06,
+      "loss": 1.0401,
+      "step": 50035
+    },
+    {
+      "epoch": 14.97,
+      "grad_norm": 3.344170331954956,
+      "learning_rate": 7.413963160744642e-06,
+      "loss": 1.0651,
+      "step": 50040
+    },
+    {
+      "epoch": 14.97,
+      "grad_norm": 4.951648712158203,
+      "learning_rate": 7.409787821697078e-06,
+      "loss": 1.0275,
+      "step": 50045
+    },
+    {
+      "epoch": 14.97,
+      "grad_norm": 2.5820136070251465,
+      "learning_rate": 7.405613454138655e-06,
+      "loss": 1.0403,
+      "step": 50050
+    },
+    {
+      "epoch": 14.98,
+      "grad_norm": 4.226273059844971,
+      "learning_rate": 7.4014400582999075e-06,
+      "loss": 0.7765,
+      "step": 50055
+    },
+    {
+      "epoch": 14.98,
+      "grad_norm": 2.259936571121216,
+      "learning_rate": 7.3972676344113366e-06,
+      "loss": 0.9865,
+      "step": 50060
+    },
+    {
+      "epoch": 14.98,
+      "grad_norm": 5.1020917892456055,
+      "learning_rate": 7.3930961827033765e-06,
+      "loss": 0.9311,
+      "step": 50065
+    },
+    {
+      "epoch": 14.98,
+      "grad_norm": 3.1059951782226562,
+      "learning_rate": 7.388925703406413e-06,
+      "loss": 1.0568,
+      "step": 50070
+    },
+    {
+      "epoch": 14.98,
+      "grad_norm": 2.5398685932159424,
+      "learning_rate": 7.384756196750775e-06,
+      "loss": 0.8391,
+      "step": 50075
+    },
+    {
+      "epoch": 14.98,
+      "grad_norm": 6.540744304656982,
+      "learning_rate": 7.380587662966743e-06,
+      "loss": 0.8211,
+      "step": 50080
+    },
+    {
+      "epoch": 14.98,
+      "grad_norm": 6.105128765106201,
+      "learning_rate": 7.376420102284543e-06,
+      "loss": 0.8739,
+      "step": 50085
+    },
+    {
+      "epoch": 14.99,
+      "grad_norm": 4.216371059417725,
+      "learning_rate": 7.372253514934338e-06,
+      "loss": 1.1114,
+      "step": 50090
+    },
+    {
+      "epoch": 14.99,
+      "grad_norm": 3.220215082168579,
+      "learning_rate": 7.368087901146259e-06,
+      "loss": 1.1085,
+      "step": 50095
+    },
+    {
+      "epoch": 14.99,
+      "grad_norm": 2.48602557182312,
+      "learning_rate": 7.36392326115034e-06,
+      "loss": 0.9271,
+      "step": 50100
+    },
+    {
+      "epoch": 14.99,
+      "grad_norm": 2.3491077423095703,
+      "learning_rate": 7.3597595951766245e-06,
+      "loss": 0.994,
+      "step": 50105
+    },
+    {
+      "epoch": 14.99,
+      "grad_norm": 2.6628293991088867,
+      "learning_rate": 7.3555969034550335e-06,
+      "loss": 1.0288,
+      "step": 50110
+    },
+    {
+      "epoch": 14.99,
+      "grad_norm": 3.2842299938201904,
+      "learning_rate": 7.351435186215502e-06,
+      "loss": 0.766,
+      "step": 50115
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 8.026897430419922,
+      "learning_rate": 7.347274443687854e-06,
+      "loss": 0.8578,
+      "step": 50120
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 1.960700273513794,
+      "learning_rate": 7.3431146761018866e-06,
+      "loss": 1.0008,
+      "step": 50125
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 1.605670690536499,
+      "learning_rate": 7.338955883687346e-06,
+      "loss": 0.9886,
+      "step": 50130
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 4.049495697021484,
+      "learning_rate": 7.334798066673912e-06,
+      "loss": 1.1398,
+      "step": 50135
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 1.0559757947921753,
+      "learning_rate": 7.330641225291218e-06,
+      "loss": 1.0428,
+      "step": 50140
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 3.241338014602661,
+      "learning_rate": 7.326485359768845e-06,
+      "loss": 0.7697,
+      "step": 50145
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 3.934131145477295,
+      "learning_rate": 7.3223304703363135e-06,
+      "loss": 0.8568,
+      "step": 50150
+    },
+    {
+      "epoch": 15.01,
+      "grad_norm": 1.8612024784088135,
+      "learning_rate": 7.318176557223097e-06,
+      "loss": 1.0974,
+      "step": 50155
+    },
+    {
+      "epoch": 15.01,
+      "grad_norm": 4.100109100341797,
+      "learning_rate": 7.314023620658608e-06,
+      "loss": 1.0554,
+      "step": 50160
+    },
+    {
+      "epoch": 15.01,
+      "grad_norm": 5.153068542480469,
+      "learning_rate": 7.309871660872211e-06,
+      "loss": 1.0952,
+      "step": 50165
+    },
+    {
+      "epoch": 15.01,
+      "grad_norm": 3.603163003921509,
+      "learning_rate": 7.305720678093214e-06,
+      "loss": 0.8898,
+      "step": 50170
+    },
+    {
+      "epoch": 15.01,
+      "grad_norm": 2.5232720375061035,
+      "learning_rate": 7.301570672550873e-06,
+      "loss": 0.909,
+      "step": 50175
+    },
+    {
+      "epoch": 15.01,
+      "grad_norm": 1.8282480239868164,
+      "learning_rate": 7.297421644474389e-06,
+      "loss": 0.8844,
+      "step": 50180
+    },
+    {
+      "epoch": 15.01,
+      "grad_norm": 3.7669854164123535,
+      "learning_rate": 7.293273594092903e-06,
+      "loss": 0.8167,
+      "step": 50185
+    },
+    {
+      "epoch": 15.02,
+      "grad_norm": 3.2310571670532227,
+      "learning_rate": 7.289126521635522e-06,
+      "loss": 0.8983,
+      "step": 50190
+    },
+    {
+      "epoch": 15.02,
+      "grad_norm": 3.063481569290161,
+      "learning_rate": 7.284980427331256e-06,
+      "loss": 0.9985,
+      "step": 50195
+    },
+    {
+      "epoch": 15.02,
+      "grad_norm": 4.190944671630859,
+      "learning_rate": 7.280835311409123e-06,
+      "loss": 1.027,
+      "step": 50200
+    },
+    {
+      "epoch": 15.02,
+      "grad_norm": 3.723928213119507,
+      "learning_rate": 7.276691174098024e-06,
+      "loss": 1.0427,
+      "step": 50205
+    },
+    {
+      "epoch": 15.02,
+      "grad_norm": 1.9666212797164917,
+      "learning_rate": 7.272548015626865e-06,
+      "loss": 1.149,
+      "step": 50210
+    },
+    {
+      "epoch": 15.02,
+      "grad_norm": 2.085639238357544,
+      "learning_rate": 7.268405836224443e-06,
+      "loss": 0.9174,
+      "step": 50215
+    },
+    {
+      "epoch": 15.03,
+      "grad_norm": 2.700467586517334,
+      "learning_rate": 7.264264636119536e-06,
+      "loss": 1.0293,
+      "step": 50220
+    },
+    {
+      "epoch": 15.03,
+      "grad_norm": 2.4150545597076416,
+      "learning_rate": 7.260124415540859e-06,
+      "loss": 0.9693,
+      "step": 50225
+    },
+    {
+      "epoch": 15.03,
+      "grad_norm": 1.4733829498291016,
+      "learning_rate": 7.255985174717067e-06,
+      "loss": 0.9674,
+      "step": 50230
+    },
+    {
+      "epoch": 15.03,
+      "grad_norm": 3.151942253112793,
+      "learning_rate": 7.251846913876772e-06,
+      "loss": 0.8926,
+      "step": 50235
+    },
+    {
+      "epoch": 15.03,
+      "grad_norm": 3.5010616779327393,
+      "learning_rate": 7.247709633248526e-06,
+      "loss": 0.8898,
+      "step": 50240
+    },
+    {
+      "epoch": 15.03,
+      "grad_norm": 3.5921566486358643,
+      "learning_rate": 7.243573333060824e-06,
+      "loss": 0.9497,
+      "step": 50245
+    },
+    {
+      "epoch": 15.03,
+      "grad_norm": 2.436879873275757,
+      "learning_rate": 7.239438013542107e-06,
+      "loss": 1.0677,
+      "step": 50250
+    },
+    {
+      "epoch": 15.04,
+      "grad_norm": 2.137514352798462,
+      "learning_rate": 7.235303674920771e-06,
+      "loss": 1.1743,
+      "step": 50255
+    },
+    {
+      "epoch": 15.04,
+      "grad_norm": 2.3740313053131104,
+      "learning_rate": 7.2311703174251454e-06,
+      "loss": 0.9629,
+      "step": 50260
+    },
+    {
+      "epoch": 15.04,
+      "grad_norm": 3.162233591079712,
+      "learning_rate": 7.227037941283515e-06,
+      "loss": 1.0236,
+      "step": 50265
+    },
+    {
+      "epoch": 15.04,
+      "grad_norm": 3.699193000793457,
+      "learning_rate": 7.222906546724104e-06,
+      "loss": 0.955,
+      "step": 50270
+    },
+    {
+      "epoch": 15.04,
+      "grad_norm": 2.988489866256714,
+      "learning_rate": 7.218776133975086e-06,
+      "loss": 1.1703,
+      "step": 50275
+    },
+    {
+      "epoch": 15.04,
+      "grad_norm": 3.407639980316162,
+      "learning_rate": 7.21464670326458e-06,
+      "loss": 0.8461,
+      "step": 50280
+    },
+    {
+      "epoch": 15.04,
+      "grad_norm": 1.7939776182174683,
+      "learning_rate": 7.210518254820658e-06,
+      "loss": 1.1161,
+      "step": 50285
+    },
+    {
+      "epoch": 15.05,
+      "grad_norm": 4.75887393951416,
+      "learning_rate": 7.206390788871306e-06,
+      "loss": 0.9254,
+      "step": 50290
+    },
+    {
+      "epoch": 15.05,
+      "grad_norm": 2.7789371013641357,
+      "learning_rate": 7.20226430564451e-06,
+      "loss": 0.9432,
+      "step": 50295
+    },
+    {
+      "epoch": 15.05,
+      "grad_norm": 5.088046073913574,
+      "learning_rate": 7.198138805368143e-06,
+      "loss": 1.1504,
+      "step": 50300
+    },
+    {
+      "epoch": 15.05,
+      "grad_norm": 3.722242593765259,
+      "learning_rate": 7.194014288270079e-06,
+      "loss": 1.0487,
+      "step": 50305
+    },
+    {
+      "epoch": 15.05,
+      "grad_norm": 1.9623161554336548,
+      "learning_rate": 7.189890754578083e-06,
+      "loss": 0.9153,
+      "step": 50310
+    },
+    {
+      "epoch": 15.05,
+      "grad_norm": 1.74498450756073,
+      "learning_rate": 7.185768204519924e-06,
+      "loss": 0.976,
+      "step": 50315
+    },
+    {
+      "epoch": 15.06,
+      "grad_norm": 1.8223540782928467,
+      "learning_rate": 7.181646638323261e-06,
+      "loss": 0.9415,
+      "step": 50320
+    },
+    {
+      "epoch": 15.06,
+      "grad_norm": 2.9013192653656006,
+      "learning_rate": 7.177526056215733e-06,
+      "loss": 0.9538,
+      "step": 50325
+    },
+    {
+      "epoch": 15.06,
+      "grad_norm": 2.3087453842163086,
+      "learning_rate": 7.173406458424917e-06,
+      "loss": 1.1462,
+      "step": 50330
+    },
+    {
+      "epoch": 15.06,
+      "grad_norm": 1.529740333557129,
+      "learning_rate": 7.169287845178335e-06,
+      "loss": 0.9129,
+      "step": 50335
+    },
+    {
+      "epoch": 15.06,
+      "grad_norm": 3.4274179935455322,
+      "learning_rate": 7.165170216703446e-06,
+      "loss": 1.035,
+      "step": 50340
+    },
+    {
+      "epoch": 15.06,
+      "grad_norm": 2.6179728507995605,
+      "learning_rate": 7.161053573227671e-06,
+      "loss": 1.0247,
+      "step": 50345
+    },
+    {
+      "epoch": 15.06,
+      "grad_norm": 3.914541721343994,
+      "learning_rate": 7.156937914978365e-06,
+      "loss": 0.888,
+      "step": 50350
+    },
+    {
+      "epoch": 15.07,
+      "grad_norm": 2.483247995376587,
+      "learning_rate": 7.152823242182829e-06,
+      "loss": 1.1135,
+      "step": 50355
+    },
+    {
+      "epoch": 15.07,
+      "grad_norm": 2.098078727722168,
+      "learning_rate": 7.148709555068314e-06,
+      "loss": 1.1288,
+      "step": 50360
+    },
+    {
+      "epoch": 15.07,
+      "grad_norm": 1.533263921737671,
+      "learning_rate": 7.144596853862015e-06,
+      "loss": 1.0257,
+      "step": 50365
+    },
+    {
+      "epoch": 15.07,
+      "grad_norm": 2.030672550201416,
+      "learning_rate": 7.140485138791075e-06,
+      "loss": 0.9171,
+      "step": 50370
+    },
+    {
+      "epoch": 15.07,
+      "grad_norm": 2.625800371170044,
+      "learning_rate": 7.136374410082575e-06,
+      "loss": 1.0061,
+      "step": 50375
+    },
+    {
+      "epoch": 15.07,
+      "grad_norm": 2.7783334255218506,
+      "learning_rate": 7.132264667963556e-06,
+      "loss": 0.9497,
+      "step": 50380
+    },
+    {
+      "epoch": 15.07,
+      "grad_norm": 1.379747748374939,
+      "learning_rate": 7.128155912660972e-06,
+      "loss": 0.8702,
+      "step": 50385
+    },
+    {
+      "epoch": 15.08,
+      "grad_norm": 2.8253347873687744,
+      "learning_rate": 7.124048144401774e-06,
+      "loss": 0.8353,
+      "step": 50390
+    },
+    {
+      "epoch": 15.08,
+      "grad_norm": 3.046046733856201,
+      "learning_rate": 7.119941363412802e-06,
+      "loss": 0.8835,
+      "step": 50395
+    },
+    {
+      "epoch": 15.08,
+      "grad_norm": 2.3068737983703613,
+      "learning_rate": 7.115835569920898e-06,
+      "loss": 1.0548,
+      "step": 50400
+    },
+    {
+      "epoch": 15.08,
+      "grad_norm": 2.7872204780578613,
+      "learning_rate": 7.111730764152791e-06,
+      "loss": 1.0162,
+      "step": 50405
+    },
+    {
+      "epoch": 15.08,
+      "grad_norm": 1.9851144552230835,
+      "learning_rate": 7.107626946335214e-06,
+      "loss": 0.8412,
+      "step": 50410
+    },
+    {
+      "epoch": 15.08,
+      "grad_norm": 2.1134822368621826,
+      "learning_rate": 7.103524116694795e-06,
+      "loss": 1.049,
+      "step": 50415
+    },
+    {
+      "epoch": 15.09,
+      "grad_norm": 1.651341199874878,
+      "learning_rate": 7.0994222754581395e-06,
+      "loss": 1.052,
+      "step": 50420
+    },
+    {
+      "epoch": 15.09,
+      "grad_norm": 1.9419851303100586,
+      "learning_rate": 7.095321422851784e-06,
+      "loss": 1.0539,
+      "step": 50425
+    },
+    {
+      "epoch": 15.09,
+      "grad_norm": 2.2368361949920654,
+      "learning_rate": 7.091221559102212e-06,
+      "loss": 0.9218,
+      "step": 50430
+    },
+    {
+      "epoch": 15.09,
+      "grad_norm": 1.586853265762329,
+      "learning_rate": 7.087122684435862e-06,
+      "loss": 0.9266,
+      "step": 50435
+    },
+    {
+      "epoch": 15.09,
+      "grad_norm": 2.1675949096679688,
+      "learning_rate": 7.083024799079099e-06,
+      "loss": 0.9255,
+      "step": 50440
+    },
+    {
+      "epoch": 15.09,
+      "grad_norm": 2.4687063694000244,
+      "learning_rate": 7.078927903258267e-06,
+      "loss": 0.9003,
+      "step": 50445
+    },
+    {
+      "epoch": 15.09,
+      "grad_norm": 3.6023154258728027,
+      "learning_rate": 7.0748319971996104e-06,
+      "loss": 1.0,
+      "step": 50450
+    },
+    {
+      "epoch": 15.1,
+      "grad_norm": 3.7124223709106445,
+      "learning_rate": 7.070737081129353e-06,
+      "loss": 0.8707,
+      "step": 50455
+    },
+    {
+      "epoch": 15.1,
+      "grad_norm": 2.767544746398926,
+      "learning_rate": 7.066643155273647e-06,
+      "loss": 1.0032,
+      "step": 50460
+    },
+    {
+      "epoch": 15.1,
+      "grad_norm": 3.66680645942688,
+      "learning_rate": 7.062550219858602e-06,
+      "loss": 1.023,
+      "step": 50465
+    },
+    {
+      "epoch": 15.1,
+      "grad_norm": 4.994945049285889,
+      "learning_rate": 7.058458275110261e-06,
+      "loss": 1.0194,
+      "step": 50470
+    },
+    {
+      "epoch": 15.1,
+      "grad_norm": 4.854514122009277,
+      "learning_rate": 7.054367321254629e-06,
+      "loss": 0.9221,
+      "step": 50475
+    },
+    {
+      "epoch": 15.1,
+      "grad_norm": 3.2869067192077637,
+      "learning_rate": 7.050277358517618e-06,
+      "loss": 0.8274,
+      "step": 50480
+    },
+    {
+      "epoch": 15.1,
+      "grad_norm": 3.371748685836792,
+      "learning_rate": 7.046188387125149e-06,
+      "loss": 1.0187,
+      "step": 50485
+    },
+    {
+      "epoch": 15.11,
+      "grad_norm": 2.176785945892334,
+      "learning_rate": 7.042100407303018e-06,
+      "loss": 0.9843,
+      "step": 50490
+    },
+    {
+      "epoch": 15.11,
+      "grad_norm": 2.4034008979797363,
+      "learning_rate": 7.038013419277034e-06,
+      "loss": 1.0416,
+      "step": 50495
+    },
+    {
+      "epoch": 15.11,
+      "grad_norm": 3.674947500228882,
+      "learning_rate": 7.033927423272879e-06,
+      "loss": 0.9973,
+      "step": 50500
+    },
+    {
+      "epoch": 15.11,
+      "grad_norm": 2.3554413318634033,
+      "learning_rate": 7.029842419516253e-06,
+      "loss": 1.0056,
+      "step": 50505
+    },
+    {
+      "epoch": 15.11,
+      "grad_norm": 4.33696985244751,
+      "learning_rate": 7.025758408232744e-06,
+      "loss": 0.8275,
+      "step": 50510
+    },
+    {
+      "epoch": 15.11,
+      "grad_norm": 2.8192827701568604,
+      "learning_rate": 7.021675389647916e-06,
+      "loss": 0.9531,
+      "step": 50515
+    },
+    {
+      "epoch": 15.12,
+      "grad_norm": 1.3056793212890625,
+      "learning_rate": 7.017593363987268e-06,
+      "loss": 0.9896,
+      "step": 50520
+    },
+    {
+      "epoch": 15.12,
+      "grad_norm": 4.758703708648682,
+      "learning_rate": 7.013512331476238e-06,
+      "loss": 1.0297,
+      "step": 50525
+    },
+    {
+      "epoch": 15.12,
+      "grad_norm": 2.7597901821136475,
+      "learning_rate": 7.009432292340243e-06,
+      "loss": 0.9807,
+      "step": 50530
+    },
+    {
+      "epoch": 15.12,
+      "grad_norm": 2.2206053733825684,
+      "learning_rate": 7.0053532468045855e-06,
+      "loss": 0.8794,
+      "step": 50535
+    },
+    {
+      "epoch": 15.12,
+      "grad_norm": 1.7539916038513184,
+      "learning_rate": 7.001275195094581e-06,
+      "loss": 1.0552,
+      "step": 50540
+    },
+    {
+      "epoch": 15.12,
+      "grad_norm": 1.2006642818450928,
+      "learning_rate": 6.997198137435432e-06,
+      "loss": 0.8482,
+      "step": 50545
+    },
+    {
+      "epoch": 15.12,
+      "grad_norm": 2.1931300163269043,
+      "learning_rate": 6.993122074052314e-06,
+      "loss": 0.8871,
+      "step": 50550
+    },
+    {
+      "epoch": 15.13,
+      "grad_norm": 1.5462733507156372,
+      "learning_rate": 6.989047005170349e-06,
+      "loss": 1.0243,
+      "step": 50555
+    },
+    {
+      "epoch": 15.13,
+      "grad_norm": 2.5775654315948486,
+      "learning_rate": 6.984972931014597e-06,
+      "loss": 1.0924,
+      "step": 50560
+    },
+    {
+      "epoch": 15.13,
+      "grad_norm": 1.4666215181350708,
+      "learning_rate": 6.980899851810061e-06,
+      "loss": 1.0157,
+      "step": 50565
+    },
+    {
+      "epoch": 15.13,
+      "grad_norm": 1.5045851469039917,
+      "learning_rate": 6.9768277677817075e-06,
+      "loss": 1.0205,
+      "step": 50570
+    },
+    {
+      "epoch": 15.13,
+      "grad_norm": 3.089066505432129,
+      "learning_rate": 6.972756679154404e-06,
+      "loss": 1.0415,
+      "step": 50575
+    },
+    {
+      "epoch": 15.13,
+      "grad_norm": 1.6471816301345825,
+      "learning_rate": 6.9686865861530285e-06,
+      "loss": 1.0791,
+      "step": 50580
+    },
+    {
+      "epoch": 15.13,
+      "grad_norm": 1.9261685609817505,
+      "learning_rate": 6.964617489002334e-06,
+      "loss": 0.8684,
+      "step": 50585
+    },
+    {
+      "epoch": 15.14,
+      "grad_norm": 14.480093002319336,
+      "learning_rate": 6.960549387927087e-06,
+      "loss": 0.8348,
+      "step": 50590
+    },
+    {
+      "epoch": 15.14,
+      "grad_norm": 3.1349940299987793,
+      "learning_rate": 6.956482283151927e-06,
+      "loss": 1.0708,
+      "step": 50595
+    },
+    {
+      "epoch": 15.14,
+      "grad_norm": 3.3612797260284424,
+      "learning_rate": 6.952416174901505e-06,
+      "loss": 1.0016,
+      "step": 50600
+    },
+    {
+      "epoch": 15.14,
+      "grad_norm": 1.3571832180023193,
+      "learning_rate": 6.948351063400385e-06,
+      "loss": 1.0697,
+      "step": 50605
+    },
+    {
+      "epoch": 15.14,
+      "grad_norm": 1.7172069549560547,
+      "learning_rate": 6.9442869488730574e-06,
+      "loss": 0.9281,
+      "step": 50610
+    },
+    {
+      "epoch": 15.14,
+      "grad_norm": 1.9647376537322998,
+      "learning_rate": 6.94022383154401e-06,
+      "loss": 1.0363,
+      "step": 50615
+    },
+    {
+      "epoch": 15.14,
+      "grad_norm": 2.1266095638275146,
+      "learning_rate": 6.936161711637612e-06,
+      "loss": 0.8694,
+      "step": 50620
+    },
+    {
+      "epoch": 15.15,
+      "grad_norm": 2.7892374992370605,
+      "learning_rate": 6.9321005893782404e-06,
+      "loss": 1.0771,
+      "step": 50625
+    },
+    {
+      "epoch": 15.15,
+      "grad_norm": 3.818957805633545,
+      "learning_rate": 6.92804046499016e-06,
+      "loss": 1.0741,
+      "step": 50630
+    },
+    {
+      "epoch": 15.15,
+      "grad_norm": 3.792024850845337,
+      "learning_rate": 6.923981338697632e-06,
+      "loss": 1.1138,
+      "step": 50635
+    },
+    {
+      "epoch": 15.15,
+      "grad_norm": 9.00833511352539,
+      "learning_rate": 6.919923210724821e-06,
+      "loss": 0.8402,
+      "step": 50640
+    },
+    {
+      "epoch": 15.15,
+      "grad_norm": 3.8722808361053467,
+      "learning_rate": 6.915866081295858e-06,
+      "loss": 0.9153,
+      "step": 50645
+    },
+    {
+      "epoch": 15.15,
+      "grad_norm": 10.343985557556152,
+      "learning_rate": 6.911809950634813e-06,
+      "loss": 0.9226,
+      "step": 50650
+    },
+    {
+      "epoch": 15.16,
+      "grad_norm": 4.360511779785156,
+      "learning_rate": 6.9077548189657e-06,
+      "loss": 0.8344,
+      "step": 50655
+    },
+    {
+      "epoch": 15.16,
+      "grad_norm": 3.0080530643463135,
+      "learning_rate": 6.903700686512488e-06,
+      "loss": 0.9623,
+      "step": 50660
+    },
+    {
+      "epoch": 15.16,
+      "grad_norm": 3.3220059871673584,
+      "learning_rate": 6.899647553499073e-06,
+      "loss": 0.8129,
+      "step": 50665
+    },
+    {
+      "epoch": 15.16,
+      "grad_norm": 1.465591549873352,
+      "learning_rate": 6.895595420149309e-06,
+      "loss": 0.9545,
+      "step": 50670
+    },
+    {
+      "epoch": 15.16,
+      "grad_norm": 2.015324115753174,
+      "learning_rate": 6.891544286687002e-06,
+      "loss": 1.1375,
+      "step": 50675
+    },
+    {
+      "epoch": 15.16,
+      "grad_norm": 2.517958164215088,
+      "learning_rate": 6.887494153335863e-06,
+      "loss": 1.1182,
+      "step": 50680
+    },
+    {
+      "epoch": 15.16,
+      "grad_norm": 4.011493682861328,
+      "learning_rate": 6.883445020319604e-06,
+      "loss": 1.1003,
+      "step": 50685
+    },
+    {
+      "epoch": 15.17,
+      "grad_norm": 3.0748703479766846,
+      "learning_rate": 6.879396887861847e-06,
+      "loss": 1.0012,
+      "step": 50690
+    },
+    {
+      "epoch": 15.17,
+      "grad_norm": 2.5038511753082275,
+      "learning_rate": 6.8753497561861615e-06,
+      "loss": 0.8941,
+      "step": 50695
+    },
+    {
+      "epoch": 15.17,
+      "grad_norm": 3.3688976764678955,
+      "learning_rate": 6.871303625516079e-06,
+      "loss": 0.9082,
+      "step": 50700
+    },
+    {
+      "epoch": 15.17,
+      "grad_norm": 1.502925992012024,
+      "learning_rate": 6.867258496075038e-06,
+      "loss": 1.0297,
+      "step": 50705
+    },
+    {
+      "epoch": 15.17,
+      "grad_norm": 2.218376636505127,
+      "learning_rate": 6.863214368086479e-06,
+      "loss": 0.7532,
+      "step": 50710
+    },
+    {
+      "epoch": 15.17,
+      "grad_norm": 3.5436737537384033,
+      "learning_rate": 6.859171241773721e-06,
+      "loss": 1.0066,
+      "step": 50715
+    },
+    {
+      "epoch": 15.17,
+      "grad_norm": 3.2597169876098633,
+      "learning_rate": 6.855129117360096e-06,
+      "loss": 0.9328,
+      "step": 50720
+    },
+    {
+      "epoch": 15.18,
+      "grad_norm": 3.4346742630004883,
+      "learning_rate": 6.851087995068811e-06,
+      "loss": 0.7659,
+      "step": 50725
+    },
+    {
+      "epoch": 15.18,
+      "grad_norm": 1.6817134618759155,
+      "learning_rate": 6.847047875123089e-06,
+      "loss": 0.9972,
+      "step": 50730
+    },
+    {
+      "epoch": 15.18,
+      "grad_norm": 3.8097851276397705,
+      "learning_rate": 6.843008757746036e-06,
+      "loss": 1.0971,
+      "step": 50735
+    },
+    {
+      "epoch": 15.18,
+      "grad_norm": 4.3598551750183105,
+      "learning_rate": 6.838970643160736e-06,
+      "loss": 0.9112,
+      "step": 50740
+    },
+    {
+      "epoch": 15.18,
+      "grad_norm": 7.945572853088379,
+      "learning_rate": 6.834933531590209e-06,
+      "loss": 1.0161,
+      "step": 50745
+    },
+    {
+      "epoch": 15.18,
+      "grad_norm": 2.397972345352173,
+      "learning_rate": 6.830897423257423e-06,
+      "loss": 1.1107,
+      "step": 50750
+    },
+    {
+      "epoch": 15.19,
+      "grad_norm": 3.211143732070923,
+      "learning_rate": 6.826862318385285e-06,
+      "loss": 1.0359,
+      "step": 50755
+    },
+    {
+      "epoch": 15.19,
+      "grad_norm": 1.8211517333984375,
+      "learning_rate": 6.822828217196656e-06,
+      "loss": 0.9641,
+      "step": 50760
+    },
+    {
+      "epoch": 15.19,
+      "grad_norm": 3.2200350761413574,
+      "learning_rate": 6.818795119914326e-06,
+      "loss": 1.0015,
+      "step": 50765
+    },
+    {
+      "epoch": 15.19,
+      "grad_norm": 1.9324058294296265,
+      "learning_rate": 6.8147630267610454e-06,
+      "loss": 1.0077,
+      "step": 50770
+    },
+    {
+      "epoch": 15.19,
+      "grad_norm": 2.95210337638855,
+      "learning_rate": 6.810731937959497e-06,
+      "loss": 1.1641,
+      "step": 50775
+    },
+    {
+      "epoch": 15.19,
+      "grad_norm": 1.5571585893630981,
+      "learning_rate": 6.8067018537323195e-06,
+      "loss": 1.0368,
+      "step": 50780
+    },
+    {
+      "epoch": 15.19,
+      "grad_norm": 12.76740550994873,
+      "learning_rate": 6.802672774302088e-06,
+      "loss": 0.9348,
+      "step": 50785
+    },
+    {
+      "epoch": 15.2,
+      "grad_norm": 1.9661290645599365,
+      "learning_rate": 6.798644699891321e-06,
+      "loss": 1.102,
+      "step": 50790
+    },
+    {
+      "epoch": 15.2,
+      "grad_norm": 1.8433258533477783,
+      "learning_rate": 6.794617630722491e-06,
+      "loss": 1.1175,
+      "step": 50795
+    },
+    {
+      "epoch": 15.2,
+      "grad_norm": 5.543290138244629,
+      "learning_rate": 6.7905915670180065e-06,
+      "loss": 0.9451,
+      "step": 50800
+    },
+    {
+      "epoch": 15.2,
+      "grad_norm": 4.0879998207092285,
+      "learning_rate": 6.786566509000228e-06,
+      "loss": 1.0948,
+      "step": 50805
+    },
+    {
+      "epoch": 15.2,
+      "grad_norm": 3.9433319568634033,
+      "learning_rate": 6.782542456891433e-06,
+      "loss": 1.0932,
+      "step": 50810
+    },
+    {
+      "epoch": 15.2,
+      "grad_norm": 1.188195824623108,
+      "learning_rate": 6.7785194109139e-06,
+      "loss": 1.0968,
+      "step": 50815
+    },
+    {
+      "epoch": 15.2,
+      "grad_norm": 1.7227709293365479,
+      "learning_rate": 6.774497371289784e-06,
+      "loss": 0.9657,
+      "step": 50820
+    },
+    {
+      "epoch": 15.21,
+      "grad_norm": 0.8999642133712769,
+      "learning_rate": 6.770476338241247e-06,
+      "loss": 0.9252,
+      "step": 50825
+    },
+    {
+      "epoch": 15.21,
+      "grad_norm": 2.9426021575927734,
+      "learning_rate": 6.766456311990347e-06,
+      "loss": 0.9599,
+      "step": 50830
+    },
+    {
+      "epoch": 15.21,
+      "grad_norm": 1.2726308107376099,
+      "learning_rate": 6.7624372927591135e-06,
+      "loss": 0.953,
+      "step": 50835
+    },
+    {
+      "epoch": 15.21,
+      "grad_norm": 1.873698353767395,
+      "learning_rate": 6.75841928076951e-06,
+      "loss": 0.8177,
+      "step": 50840
+    },
+    {
+      "epoch": 15.21,
+      "grad_norm": 3.4419801235198975,
+      "learning_rate": 6.7544022762434486e-06,
+      "loss": 0.9171,
+      "step": 50845
+    },
+    {
+      "epoch": 15.21,
+      "grad_norm": 2.00152325630188,
+      "learning_rate": 6.750386279402784e-06,
+      "loss": 0.981,
+      "step": 50850
+    },
+    {
+      "epoch": 15.22,
+      "grad_norm": 1.6282256841659546,
+      "learning_rate": 6.7463712904693165e-06,
+      "loss": 0.9112,
+      "step": 50855
+    },
+    {
+      "epoch": 15.22,
+      "grad_norm": 3.2624893188476562,
+      "learning_rate": 6.742357309664787e-06,
+      "loss": 0.9756,
+      "step": 50860
+    },
+    {
+      "epoch": 15.22,
+      "grad_norm": 3.1753547191619873,
+      "learning_rate": 6.738344337210889e-06,
+      "loss": 1.0622,
+      "step": 50865
+    },
+    {
+      "epoch": 15.22,
+      "grad_norm": 4.24165678024292,
+      "learning_rate": 6.73433237332925e-06,
+      "loss": 0.9718,
+      "step": 50870
+    },
+    {
+      "epoch": 15.22,
+      "grad_norm": 3.261056661605835,
+      "learning_rate": 6.730321418241448e-06,
+      "loss": 0.9434,
+      "step": 50875
+    },
+    {
+      "epoch": 15.22,
+      "grad_norm": 3.543851852416992,
+      "learning_rate": 6.726311472169003e-06,
+      "loss": 1.0265,
+      "step": 50880
+    },
+    {
+      "epoch": 15.22,
+      "grad_norm": 2.614902973175049,
+      "learning_rate": 6.72230253533338e-06,
+      "loss": 0.8983,
+      "step": 50885
+    },
+    {
+      "epoch": 15.23,
+      "grad_norm": 2.52410626411438,
+      "learning_rate": 6.718294607955991e-06,
+      "loss": 0.9063,
+      "step": 50890
+    },
+    {
+      "epoch": 15.23,
+      "grad_norm": 1.27847421169281,
+      "learning_rate": 6.7142876902581895e-06,
+      "loss": 0.9997,
+      "step": 50895
+    },
+    {
+      "epoch": 15.23,
+      "grad_norm": 1.5926611423492432,
+      "learning_rate": 6.710281782461275e-06,
+      "loss": 0.9342,
+      "step": 50900
+    },
+    {
+      "epoch": 15.23,
+      "grad_norm": 2.9667863845825195,
+      "learning_rate": 6.706276884786475e-06,
+      "loss": 1.1293,
+      "step": 50905
+    },
+    {
+      "epoch": 15.23,
+      "grad_norm": 2.5114152431488037,
+      "learning_rate": 6.702272997455e-06,
+      "loss": 1.1009,
+      "step": 50910
+    },
+    {
+      "epoch": 15.23,
+      "grad_norm": 3.5501632690429688,
+      "learning_rate": 6.698270120687955e-06,
+      "loss": 0.9759,
+      "step": 50915
+    },
+    {
+      "epoch": 15.23,
+      "grad_norm": 5.344282627105713,
+      "learning_rate": 6.694268254706443e-06,
+      "loss": 0.8718,
+      "step": 50920
+    },
+    {
+      "epoch": 15.24,
+      "grad_norm": 4.164098262786865,
+      "learning_rate": 6.690267399731459e-06,
+      "loss": 0.976,
+      "step": 50925
+    },
+    {
+      "epoch": 15.24,
+      "grad_norm": 2.1116974353790283,
+      "learning_rate": 6.6862675559839744e-06,
+      "loss": 0.8628,
+      "step": 50930
+    },
+    {
+      "epoch": 15.24,
+      "grad_norm": 1.2934943437576294,
+      "learning_rate": 6.682268723684898e-06,
+      "loss": 0.9742,
+      "step": 50935
+    },
+    {
+      "epoch": 15.24,
+      "grad_norm": 1.8546779155731201,
+      "learning_rate": 6.678270903055078e-06,
+      "loss": 0.9964,
+      "step": 50940
+    },
+    {
+      "epoch": 15.24,
+      "grad_norm": 1.931983470916748,
+      "learning_rate": 6.674274094315311e-06,
+      "loss": 0.8238,
+      "step": 50945
+    },
+    {
+      "epoch": 15.24,
+      "grad_norm": 1.8833105564117432,
+      "learning_rate": 6.670278297686341e-06,
+      "loss": 0.9469,
+      "step": 50950
+    },
+    {
+      "epoch": 15.25,
+      "grad_norm": 2.2532496452331543,
+      "learning_rate": 6.666283513388844e-06,
+      "loss": 0.9581,
+      "step": 50955
+    },
+    {
+      "epoch": 15.25,
+      "grad_norm": 3.5239064693450928,
+      "learning_rate": 6.662289741643454e-06,
+      "loss": 1.1331,
+      "step": 50960
+    },
+    {
+      "epoch": 15.25,
+      "grad_norm": 4.513704299926758,
+      "learning_rate": 6.658296982670739e-06,
+      "loss": 0.9937,
+      "step": 50965
+    },
+    {
+      "epoch": 15.25,
+      "grad_norm": 3.6698179244995117,
+      "learning_rate": 6.654305236691219e-06,
+      "loss": 0.9723,
+      "step": 50970
+    },
+    {
+      "epoch": 15.25,
+      "grad_norm": 2.5509257316589355,
+      "learning_rate": 6.650314503925348e-06,
+      "loss": 1.0257,
+      "step": 50975
+    },
+    {
+      "epoch": 15.25,
+      "grad_norm": 4.316878795623779,
+      "learning_rate": 6.646324784593536e-06,
+      "loss": 0.8953,
+      "step": 50980
+    },
+    {
+      "epoch": 15.25,
+      "grad_norm": 2.966710090637207,
+      "learning_rate": 6.6423360789161285e-06,
+      "loss": 1.0324,
+      "step": 50985
+    },
+    {
+      "epoch": 15.26,
+      "grad_norm": 3.1171488761901855,
+      "learning_rate": 6.638348387113416e-06,
+      "loss": 0.92,
+      "step": 50990
+    },
+    {
+      "epoch": 15.26,
+      "grad_norm": 4.053133487701416,
+      "learning_rate": 6.634361709405645e-06,
+      "loss": 0.9734,
+      "step": 50995
+    },
+    {
+      "epoch": 15.26,
+      "grad_norm": 2.0579705238342285,
+      "learning_rate": 6.630376046012973e-06,
+      "loss": 0.9514,
+      "step": 51000
+    },
+    {
+      "epoch": 15.26,
+      "grad_norm": 1.2330838441848755,
+      "learning_rate": 6.6263913971555515e-06,
+      "loss": 1.0869,
+      "step": 51005
+    },
+    {
+      "epoch": 15.26,
+      "grad_norm": 4.011294841766357,
+      "learning_rate": 6.6224077630534205e-06,
+      "loss": 1.148,
+      "step": 51010
+    },
+    {
+      "epoch": 15.26,
+      "grad_norm": 1.641052007675171,
+      "learning_rate": 6.618425143926618e-06,
+      "loss": 1.0063,
+      "step": 51015
+    },
+    {
+      "epoch": 15.26,
+      "grad_norm": 1.6458262205123901,
+      "learning_rate": 6.614443539995074e-06,
+      "loss": 0.9671,
+      "step": 51020
+    },
+    {
+      "epoch": 15.27,
+      "grad_norm": 4.2398529052734375,
+      "learning_rate": 6.6104629514787215e-06,
+      "loss": 0.902,
+      "step": 51025
+    },
+    {
+      "epoch": 15.27,
+      "grad_norm": 2.740884780883789,
+      "learning_rate": 6.606483378597375e-06,
+      "loss": 0.9838,
+      "step": 51030
+    },
+    {
+      "epoch": 15.27,
+      "grad_norm": 1.063081979751587,
+      "learning_rate": 6.6025048215708345e-06,
+      "loss": 0.7767,
+      "step": 51035
+    },
+    {
+      "epoch": 15.27,
+      "grad_norm": 1.965700387954712,
+      "learning_rate": 6.598527280618827e-06,
+      "loss": 1.1383,
+      "step": 51040
+    },
+    {
+      "epoch": 15.27,
+      "grad_norm": 5.353956699371338,
+      "learning_rate": 6.594550755961032e-06,
+      "loss": 0.9201,
+      "step": 51045
+    },
+    {
+      "epoch": 15.27,
+      "grad_norm": 3.3293633460998535,
+      "learning_rate": 6.590575247817069e-06,
+      "loss": 1.0393,
+      "step": 51050
+    },
+    {
+      "epoch": 15.28,
+      "grad_norm": 3.897473096847534,
+      "learning_rate": 6.586600756406494e-06,
+      "loss": 0.8148,
+      "step": 51055
+    },
+    {
+      "epoch": 15.28,
+      "grad_norm": 2.9738924503326416,
+      "learning_rate": 6.582627281948825e-06,
+      "loss": 1.107,
+      "step": 51060
+    },
+    {
+      "epoch": 15.28,
+      "grad_norm": 1.3063501119613647,
+      "learning_rate": 6.578654824663505e-06,
+      "loss": 0.9212,
+      "step": 51065
+    },
+    {
+      "epoch": 15.28,
+      "grad_norm": 2.1776340007781982,
+      "learning_rate": 6.574683384769933e-06,
+      "loss": 1.1221,
+      "step": 51070
+    },
+    {
+      "epoch": 15.28,
+      "grad_norm": 2.1266050338745117,
+      "learning_rate": 6.570712962487444e-06,
+      "loss": 0.9498,
+      "step": 51075
+    },
+    {
+      "epoch": 15.28,
+      "grad_norm": 1.2182793617248535,
+      "learning_rate": 6.566743558035324e-06,
+      "loss": 0.9176,
+      "step": 51080
+    },
+    {
+      "epoch": 15.28,
+      "grad_norm": 11.197833061218262,
+      "learning_rate": 6.562775171632793e-06,
+      "loss": 0.8579,
+      "step": 51085
+    },
+    {
+      "epoch": 15.29,
+      "grad_norm": 1.7195172309875488,
+      "learning_rate": 6.558807803499034e-06,
+      "loss": 1.1437,
+      "step": 51090
+    },
+    {
+      "epoch": 15.29,
+      "grad_norm": 2.056217670440674,
+      "learning_rate": 6.554841453853139e-06,
+      "loss": 1.1385,
+      "step": 51095
+    },
+    {
+      "epoch": 15.29,
+      "grad_norm": 5.98038387298584,
+      "learning_rate": 6.55087612291419e-06,
+      "loss": 0.7964,
+      "step": 51100
+    },
+    {
+      "epoch": 15.29,
+      "grad_norm": 5.708714962005615,
+      "learning_rate": 6.54691181090116e-06,
+      "loss": 0.9003,
+      "step": 51105
+    },
+    {
+      "epoch": 15.29,
+      "grad_norm": 1.2579269409179688,
+      "learning_rate": 6.542948518033029e-06,
+      "loss": 1.0455,
+      "step": 51110
+    },
+    {
+      "epoch": 15.29,
+      "grad_norm": 1.0137869119644165,
+      "learning_rate": 6.538986244528647e-06,
+      "loss": 0.8792,
+      "step": 51115
+    },
+    {
+      "epoch": 15.29,
+      "grad_norm": 1.5064419507980347,
+      "learning_rate": 6.535024990606883e-06,
+      "loss": 0.8194,
+      "step": 51120
+    },
+    {
+      "epoch": 15.3,
+      "grad_norm": 2.8707802295684814,
+      "learning_rate": 6.531064756486488e-06,
+      "loss": 1.0377,
+      "step": 51125
+    },
+    {
+      "epoch": 15.3,
+      "grad_norm": 2.328263521194458,
+      "learning_rate": 6.527105542386189e-06,
+      "loss": 0.9897,
+      "step": 51130
+    },
+    {
+      "epoch": 15.3,
+      "grad_norm": 5.99672269821167,
+      "learning_rate": 6.523147348524649e-06,
+      "loss": 1.0021,
+      "step": 51135
+    },
+    {
+      "epoch": 15.3,
+      "grad_norm": 1.5641214847564697,
+      "learning_rate": 6.519190175120473e-06,
+      "loss": 1.1588,
+      "step": 51140
+    },
+    {
+      "epoch": 15.3,
+      "grad_norm": 6.234478950500488,
+      "learning_rate": 6.515234022392217e-06,
+      "loss": 0.9205,
+      "step": 51145
+    },
+    {
+      "epoch": 15.3,
+      "grad_norm": 1.717020869255066,
+      "learning_rate": 6.511278890558373e-06,
+      "loss": 1.0511,
+      "step": 51150
+    },
+    {
+      "epoch": 15.3,
+      "grad_norm": 3.1675846576690674,
+      "learning_rate": 6.507324779837376e-06,
+      "loss": 0.8302,
+      "step": 51155
+    },
+    {
+      "epoch": 15.31,
+      "grad_norm": 2.2740492820739746,
+      "learning_rate": 6.503371690447607e-06,
+      "loss": 0.8826,
+      "step": 51160
+    },
+    {
+      "epoch": 15.31,
+      "grad_norm": 3.2151083946228027,
+      "learning_rate": 6.499419622607397e-06,
+      "loss": 0.9839,
+      "step": 51165
+    },
+    {
+      "epoch": 15.31,
+      "grad_norm": 5.2867560386657715,
+      "learning_rate": 6.49546857653501e-06,
+      "loss": 0.9079,
+      "step": 51170
+    },
+    {
+      "epoch": 15.31,
+      "grad_norm": 2.4737038612365723,
+      "learning_rate": 6.491518552448658e-06,
+      "loss": 0.989,
+      "step": 51175
+    },
+    {
+      "epoch": 15.31,
+      "grad_norm": 2.2063138484954834,
+      "learning_rate": 6.487569550566499e-06,
+      "loss": 0.8596,
+      "step": 51180
+    },
+    {
+      "epoch": 15.31,
+      "grad_norm": 3.6992015838623047,
+      "learning_rate": 6.48362157110664e-06,
+      "loss": 1.1441,
+      "step": 51185
+    },
+    {
+      "epoch": 15.32,
+      "grad_norm": 4.015367031097412,
+      "learning_rate": 6.479674614287098e-06,
+      "loss": 1.054,
+      "step": 51190
+    },
+    {
+      "epoch": 15.32,
+      "grad_norm": 2.584434747695923,
+      "learning_rate": 6.47572868032589e-06,
+      "loss": 1.0146,
+      "step": 51195
+    },
+    {
+      "epoch": 15.32,
+      "grad_norm": 2.181087017059326,
+      "learning_rate": 6.47178376944092e-06,
+      "loss": 0.9372,
+      "step": 51200
+    },
+    {
+      "epoch": 15.32,
+      "grad_norm": 4.231384754180908,
+      "learning_rate": 6.467839881850085e-06,
+      "loss": 1.0406,
+      "step": 51205
+    },
+    {
+      "epoch": 15.32,
+      "grad_norm": 3.6681032180786133,
+      "learning_rate": 6.463897017771178e-06,
+      "loss": 0.9692,
+      "step": 51210
+    },
+    {
+      "epoch": 15.32,
+      "grad_norm": 1.9487011432647705,
+      "learning_rate": 6.459955177421986e-06,
+      "loss": 1.0082,
+      "step": 51215
+    },
+    {
+      "epoch": 15.32,
+      "grad_norm": 2.813549518585205,
+      "learning_rate": 6.45601436102019e-06,
+      "loss": 1.1155,
+      "step": 51220
+    },
+    {
+      "epoch": 15.33,
+      "grad_norm": 1.6334689855575562,
+      "learning_rate": 6.4520745687834476e-06,
+      "loss": 0.9634,
+      "step": 51225
+    },
+    {
+      "epoch": 15.33,
+      "grad_norm": 3.823737382888794,
+      "learning_rate": 6.448135800929347e-06,
+      "loss": 1.1307,
+      "step": 51230
+    },
+    {
+      "epoch": 15.33,
+      "grad_norm": 1.4004770517349243,
+      "learning_rate": 6.444198057675418e-06,
+      "loss": 1.0298,
+      "step": 51235
+    },
+    {
+      "epoch": 15.33,
+      "grad_norm": 6.257365703582764,
+      "learning_rate": 6.440261339239148e-06,
+      "loss": 0.97,
+      "step": 51240
+    },
+    {
+      "epoch": 15.33,
+      "grad_norm": 6.089005947113037,
+      "learning_rate": 6.436325645837951e-06,
+      "loss": 0.9796,
+      "step": 51245
+    },
+    {
+      "epoch": 15.33,
+      "grad_norm": 4.073474884033203,
+      "learning_rate": 6.432390977689193e-06,
+      "loss": 0.9491,
+      "step": 51250
+    },
+    {
+      "epoch": 15.33,
+      "grad_norm": 1.1611909866333008,
+      "learning_rate": 6.4284573350101765e-06,
+      "loss": 1.0043,
+      "step": 51255
+    },
+    {
+      "epoch": 15.34,
+      "grad_norm": 1.883181095123291,
+      "learning_rate": 6.424524718018163e-06,
+      "loss": 0.8922,
+      "step": 51260
+    },
+    {
+      "epoch": 15.34,
+      "grad_norm": 12.2318754196167,
+      "learning_rate": 6.4205931269303385e-06,
+      "loss": 1.024,
+      "step": 51265
+    },
+    {
+      "epoch": 15.34,
+      "grad_norm": 2.295652151107788,
+      "learning_rate": 6.416662561963846e-06,
+      "loss": 0.8952,
+      "step": 51270
+    },
+    {
+      "epoch": 15.34,
+      "grad_norm": 6.346905708312988,
+      "learning_rate": 6.412733023335763e-06,
+      "loss": 1.072,
+      "step": 51275
+    },
+    {
+      "epoch": 15.34,
+      "grad_norm": 3.8380446434020996,
+      "learning_rate": 6.408804511263119e-06,
+      "loss": 0.9072,
+      "step": 51280
+    },
+    {
+      "epoch": 15.34,
+      "grad_norm": 2.404541492462158,
+      "learning_rate": 6.404877025962866e-06,
+      "loss": 1.0707,
+      "step": 51285
+    },
+    {
+      "epoch": 15.35,
+      "grad_norm": 2.2299270629882812,
+      "learning_rate": 6.400950567651939e-06,
+      "loss": 0.9094,
+      "step": 51290
+    },
+    {
+      "epoch": 15.35,
+      "grad_norm": 4.42881965637207,
+      "learning_rate": 6.397025136547166e-06,
+      "loss": 1.04,
+      "step": 51295
+    },
+    {
+      "epoch": 15.35,
+      "grad_norm": 1.782752513885498,
+      "learning_rate": 6.393100732865373e-06,
+      "loss": 0.9773,
+      "step": 51300
+    },
+    {
+      "epoch": 15.35,
+      "grad_norm": 3.4827709197998047,
+      "learning_rate": 6.38917735682327e-06,
+      "loss": 0.8695,
+      "step": 51305
+    },
+    {
+      "epoch": 15.35,
+      "grad_norm": 2.491976022720337,
+      "learning_rate": 6.385255008637573e-06,
+      "loss": 0.8696,
+      "step": 51310
+    },
+    {
+      "epoch": 15.35,
+      "grad_norm": 3.1498947143554688,
+      "learning_rate": 6.381333688524885e-06,
+      "loss": 0.849,
+      "step": 51315
+    },
+    {
+      "epoch": 15.35,
+      "grad_norm": 2.436060667037964,
+      "learning_rate": 6.377413396701781e-06,
+      "loss": 0.9044,
+      "step": 51320
+    },
+    {
+      "epoch": 15.36,
+      "grad_norm": 2.632699966430664,
+      "learning_rate": 6.373494133384783e-06,
+      "loss": 1.0249,
+      "step": 51325
+    },
+    {
+      "epoch": 15.36,
+      "grad_norm": 10.201811790466309,
+      "learning_rate": 6.369575898790334e-06,
+      "loss": 0.9744,
+      "step": 51330
+    },
+    {
+      "epoch": 15.36,
+      "grad_norm": 2.500356435775757,
+      "learning_rate": 6.365658693134857e-06,
+      "loss": 1.0102,
+      "step": 51335
+    },
+    {
+      "epoch": 15.36,
+      "grad_norm": 2.9083356857299805,
+      "learning_rate": 6.361742516634664e-06,
+      "loss": 1.012,
+      "step": 51340
+    },
+    {
+      "epoch": 15.36,
+      "grad_norm": 4.8380022048950195,
+      "learning_rate": 6.357827369506075e-06,
+      "loss": 0.9587,
+      "step": 51345
+    },
+    {
+      "epoch": 15.36,
+      "grad_norm": 4.04415225982666,
+      "learning_rate": 6.353913251965296e-06,
+      "loss": 0.7099,
+      "step": 51350
+    },
+    {
+      "epoch": 15.36,
+      "grad_norm": 3.18339204788208,
+      "learning_rate": 6.350000164228509e-06,
+      "loss": 1.0013,
+      "step": 51355
+    },
+    {
+      "epoch": 15.37,
+      "grad_norm": 1.7995153665542603,
+      "learning_rate": 6.346088106511821e-06,
+      "loss": 0.8609,
+      "step": 51360
+    },
+    {
+      "epoch": 15.37,
+      "grad_norm": 5.062925338745117,
+      "learning_rate": 6.3421770790313025e-06,
+      "loss": 0.9407,
+      "step": 51365
+    },
+    {
+      "epoch": 15.37,
+      "grad_norm": 2.892843723297119,
+      "learning_rate": 6.338267082002949e-06,
+      "loss": 1.0163,
+      "step": 51370
+    },
+    {
+      "epoch": 15.37,
+      "grad_norm": 2.197019577026367,
+      "learning_rate": 6.334358115642708e-06,
+      "loss": 0.9733,
+      "step": 51375
+    },
+    {
+      "epoch": 15.37,
+      "grad_norm": 3.871853828430176,
+      "learning_rate": 6.330450180166464e-06,
+      "loss": 1.108,
+      "step": 51380
+    },
+    {
+      "epoch": 15.37,
+      "grad_norm": 1.472909927368164,
+      "learning_rate": 6.326543275790059e-06,
+      "loss": 1.0104,
+      "step": 51385
+    },
+    {
+      "epoch": 15.38,
+      "grad_norm": 3.942326068878174,
+      "learning_rate": 6.322637402729242e-06,
+      "loss": 1.0015,
+      "step": 51390
+    },
+    {
+      "epoch": 15.38,
+      "grad_norm": 7.206786155700684,
+      "learning_rate": 6.318732561199767e-06,
+      "loss": 1.0726,
+      "step": 51395
+    },
+    {
+      "epoch": 15.38,
+      "grad_norm": 1.5072174072265625,
+      "learning_rate": 6.314828751417257e-06,
+      "loss": 1.0457,
+      "step": 51400
+    },
+    {
+      "epoch": 15.38,
+      "grad_norm": 2.0470387935638428,
+      "learning_rate": 6.3109259735973455e-06,
+      "loss": 1.1581,
+      "step": 51405
+    },
+    {
+      "epoch": 15.38,
+      "grad_norm": 5.198572635650635,
+      "learning_rate": 6.3070242279555606e-06,
+      "loss": 1.0023,
+      "step": 51410
+    },
+    {
+      "epoch": 15.38,
+      "grad_norm": 3.871081590652466,
+      "learning_rate": 6.303123514707393e-06,
+      "loss": 0.9422,
+      "step": 51415
+    },
+    {
+      "epoch": 15.38,
+      "grad_norm": 4.593938827514648,
+      "learning_rate": 6.299223834068288e-06,
+      "loss": 1.1043,
+      "step": 51420
+    },
+    {
+      "epoch": 15.39,
+      "grad_norm": 3.9437222480773926,
+      "learning_rate": 6.2953251862536036e-06,
+      "loss": 0.77,
+      "step": 51425
+    },
+    {
+      "epoch": 15.39,
+      "grad_norm": 1.9378674030303955,
+      "learning_rate": 6.291427571478678e-06,
+      "loss": 1.0061,
+      "step": 51430
+    },
+    {
+      "epoch": 15.39,
+      "grad_norm": 1.423578143119812,
+      "learning_rate": 6.287530989958748e-06,
+      "loss": 0.8501,
+      "step": 51435
+    },
+    {
+      "epoch": 15.39,
+      "grad_norm": 7.725799560546875,
+      "learning_rate": 6.283635441909045e-06,
+      "loss": 0.8871,
+      "step": 51440
+    },
+    {
+      "epoch": 15.39,
+      "grad_norm": 1.6603666543960571,
+      "learning_rate": 6.2797409275446914e-06,
+      "loss": 0.7305,
+      "step": 51445
+    },
+    {
+      "epoch": 15.39,
+      "grad_norm": 1.9601699113845825,
+      "learning_rate": 6.275847447080793e-06,
+      "loss": 1.0657,
+      "step": 51450
+    },
+    {
+      "epoch": 15.39,
+      "grad_norm": 3.121793746948242,
+      "learning_rate": 6.271955000732374e-06,
+      "loss": 1.1533,
+      "step": 51455
+    },
+    {
+      "epoch": 15.4,
+      "grad_norm": 1.4244256019592285,
+      "learning_rate": 6.268063588714415e-06,
+      "loss": 0.9971,
+      "step": 51460
+    },
+    {
+      "epoch": 15.4,
+      "grad_norm": 2.398967981338501,
+      "learning_rate": 6.264173211241833e-06,
+      "loss": 0.9865,
+      "step": 51465
+    },
+    {
+      "epoch": 15.4,
+      "grad_norm": 1.3114820718765259,
+      "learning_rate": 6.26028386852949e-06,
+      "loss": 0.9019,
+      "step": 51470
+    },
+    {
+      "epoch": 15.4,
+      "grad_norm": 5.757796764373779,
+      "learning_rate": 6.256395560792188e-06,
+      "loss": 1.0239,
+      "step": 51475
+    },
+    {
+      "epoch": 15.4,
+      "grad_norm": 1.7129333019256592,
+      "learning_rate": 6.252508288244688e-06,
+      "loss": 1.1315,
+      "step": 51480
+    },
+    {
+      "epoch": 15.4,
+      "grad_norm": 3.2258822917938232,
+      "learning_rate": 6.248622051101649e-06,
+      "loss": 1.0204,
+      "step": 51485
+    },
+    {
+      "epoch": 15.41,
+      "grad_norm": 2.3875536918640137,
+      "learning_rate": 6.244736849577734e-06,
+      "loss": 0.9069,
+      "step": 51490
+    },
+    {
+      "epoch": 15.41,
+      "grad_norm": 2.4159421920776367,
+      "learning_rate": 6.240852683887508e-06,
+      "loss": 0.9534,
+      "step": 51495
+    },
+    {
+      "epoch": 15.41,
+      "grad_norm": 3.4338598251342773,
+      "learning_rate": 6.236969554245486e-06,
+      "loss": 1.0511,
+      "step": 51500
+    },
+    {
+      "epoch": 15.41,
+      "grad_norm": 1.4087140560150146,
+      "learning_rate": 6.233087460866141e-06,
+      "loss": 0.987,
+      "step": 51505
+    },
+    {
+      "epoch": 15.41,
+      "grad_norm": 2.022752285003662,
+      "learning_rate": 6.229206403963852e-06,
+      "loss": 0.9237,
+      "step": 51510
+    },
+    {
+      "epoch": 15.41,
+      "grad_norm": 2.3156323432922363,
+      "learning_rate": 6.225326383753e-06,
+      "loss": 0.8778,
+      "step": 51515
+    },
+    {
+      "epoch": 15.41,
+      "grad_norm": 2.3037240505218506,
+      "learning_rate": 6.221447400447841e-06,
+      "loss": 1.0857,
+      "step": 51520
+    },
+    {
+      "epoch": 15.42,
+      "grad_norm": 3.583475112915039,
+      "learning_rate": 6.2175694542626394e-06,
+      "loss": 1.0723,
+      "step": 51525
+    },
+    {
+      "epoch": 15.42,
+      "grad_norm": 4.166855335235596,
+      "learning_rate": 6.213692545411537e-06,
+      "loss": 1.1171,
+      "step": 51530
+    },
+    {
+      "epoch": 15.42,
+      "grad_norm": 1.957580327987671,
+      "learning_rate": 6.2098166741086815e-06,
+      "loss": 0.9268,
+      "step": 51535
+    },
+    {
+      "epoch": 15.42,
+      "grad_norm": 2.228228807449341,
+      "learning_rate": 6.205941840568114e-06,
+      "loss": 0.9613,
+      "step": 51540
+    },
+    {
+      "epoch": 15.42,
+      "grad_norm": 3.6298060417175293,
+      "learning_rate": 6.2020680450038425e-06,
+      "loss": 1.1636,
+      "step": 51545
+    },
+    {
+      "epoch": 15.42,
+      "grad_norm": 3.282169818878174,
+      "learning_rate": 6.198195287629813e-06,
+      "loss": 0.898,
+      "step": 51550
+    },
+    {
+      "epoch": 15.42,
+      "grad_norm": 15.333877563476562,
+      "learning_rate": 6.194323568659916e-06,
+      "loss": 0.9714,
+      "step": 51555
+    },
+    {
+      "epoch": 15.43,
+      "grad_norm": 4.079663276672363,
+      "learning_rate": 6.190452888307979e-06,
+      "loss": 1.148,
+      "step": 51560
+    },
+    {
+      "epoch": 15.43,
+      "grad_norm": 3.9293205738067627,
+      "learning_rate": 6.18658324678778e-06,
+      "loss": 0.8797,
+      "step": 51565
+    },
+    {
+      "epoch": 15.43,
+      "grad_norm": 3.552079916000366,
+      "learning_rate": 6.1827146443130315e-06,
+      "loss": 1.0032,
+      "step": 51570
+    },
+    {
+      "epoch": 15.43,
+      "grad_norm": 1.169809341430664,
+      "learning_rate": 6.1788470810973944e-06,
+      "loss": 0.9592,
+      "step": 51575
+    },
+    {
+      "epoch": 15.43,
+      "grad_norm": 1.6891237497329712,
+      "learning_rate": 6.174980557354468e-06,
+      "loss": 1.0002,
+      "step": 51580
+    },
+    {
+      "epoch": 15.43,
+      "grad_norm": 1.8492525815963745,
+      "learning_rate": 6.171115073297801e-06,
+      "loss": 0.9641,
+      "step": 51585
+    },
+    {
+      "epoch": 15.44,
+      "grad_norm": 5.5648579597473145,
+      "learning_rate": 6.167250629140872e-06,
+      "loss": 0.9558,
+      "step": 51590
+    },
+    {
+      "epoch": 15.44,
+      "grad_norm": 2.914482831954956,
+      "learning_rate": 6.16338722509712e-06,
+      "loss": 1.0153,
+      "step": 51595
+    },
+    {
+      "epoch": 15.44,
+      "grad_norm": 2.55684757232666,
+      "learning_rate": 6.159524861379909e-06,
+      "loss": 1.0406,
+      "step": 51600
+    },
+    {
+      "epoch": 15.44,
+      "grad_norm": 2.4347622394561768,
+      "learning_rate": 6.155663538202558e-06,
+      "loss": 0.9547,
+      "step": 51605
+    },
+    {
+      "epoch": 15.44,
+      "grad_norm": 5.680244445800781,
+      "learning_rate": 6.1518032557783265e-06,
+      "loss": 1.0837,
+      "step": 51610
+    },
+    {
+      "epoch": 15.44,
+      "grad_norm": 4.704986572265625,
+      "learning_rate": 6.147944014320398e-06,
+      "loss": 1.177,
+      "step": 51615
+    },
+    {
+      "epoch": 15.44,
+      "grad_norm": 4.459755897521973,
+      "learning_rate": 6.1440858140419414e-06,
+      "loss": 1.2561,
+      "step": 51620
+    },
+    {
+      "epoch": 15.45,
+      "grad_norm": 2.4121649265289307,
+      "learning_rate": 6.140228655156008e-06,
+      "loss": 1.0119,
+      "step": 51625
+    },
+    {
+      "epoch": 15.45,
+      "grad_norm": 1.6860464811325073,
+      "learning_rate": 6.136372537875659e-06,
+      "loss": 0.9792,
+      "step": 51630
+    },
+    {
+      "epoch": 15.45,
+      "grad_norm": 2.846508741378784,
+      "learning_rate": 6.132517462413839e-06,
+      "loss": 0.8266,
+      "step": 51635
+    },
+    {
+      "epoch": 15.45,
+      "grad_norm": 3.4774718284606934,
+      "learning_rate": 6.128663428983469e-06,
+      "loss": 0.9554,
+      "step": 51640
+    },
+    {
+      "epoch": 15.45,
+      "grad_norm": 2.086501121520996,
+      "learning_rate": 6.124810437797402e-06,
+      "loss": 0.9001,
+      "step": 51645
+    },
+    {
+      "epoch": 15.45,
+      "grad_norm": 2.1175014972686768,
+      "learning_rate": 6.120958489068434e-06,
+      "loss": 0.8511,
+      "step": 51650
+    },
+    {
+      "epoch": 15.45,
+      "grad_norm": 2.3501827716827393,
+      "learning_rate": 6.117107583009304e-06,
+      "loss": 1.0259,
+      "step": 51655
+    },
+    {
+      "epoch": 15.46,
+      "grad_norm": 2.106353998184204,
+      "learning_rate": 6.113257719832696e-06,
+      "loss": 0.799,
+      "step": 51660
+    },
+    {
+      "epoch": 15.46,
+      "grad_norm": 1.536186695098877,
+      "learning_rate": 6.1094088997512315e-06,
+      "loss": 0.9458,
+      "step": 51665
+    },
+    {
+      "epoch": 15.46,
+      "grad_norm": 3.473801612854004,
+      "learning_rate": 6.105561122977479e-06,
+      "loss": 0.9084,
+      "step": 51670
+    },
+    {
+      "epoch": 15.46,
+      "grad_norm": 2.5647025108337402,
+      "learning_rate": 6.101714389723945e-06,
+      "loss": 0.9893,
+      "step": 51675
+    },
+    {
+      "epoch": 15.46,
+      "grad_norm": 5.690295696258545,
+      "learning_rate": 6.097868700203083e-06,
+      "loss": 1.0979,
+      "step": 51680
+    },
+    {
+      "epoch": 15.46,
+      "grad_norm": 3.3370258808135986,
+      "learning_rate": 6.094024054627281e-06,
+      "loss": 0.9205,
+      "step": 51685
+    },
+    {
+      "epoch": 15.47,
+      "grad_norm": 5.14525032043457,
+      "learning_rate": 6.0901804532088824e-06,
+      "loss": 0.9944,
+      "step": 51690
+    },
+    {
+      "epoch": 15.47,
+      "grad_norm": 12.52432632446289,
+      "learning_rate": 6.086337896160163e-06,
+      "loss": 0.9203,
+      "step": 51695
+    },
+    {
+      "epoch": 15.47,
+      "grad_norm": 3.637500047683716,
+      "learning_rate": 6.0824963836933396e-06,
+      "loss": 0.9591,
+      "step": 51700
+    },
+    {
+      "epoch": 15.47,
+      "grad_norm": 2.3651740550994873,
+      "learning_rate": 6.078655916020584e-06,
+      "loss": 0.8642,
+      "step": 51705
+    },
+    {
+      "epoch": 15.47,
+      "grad_norm": 4.2891645431518555,
+      "learning_rate": 6.0748164933539796e-06,
+      "loss": 1.094,
+      "step": 51710
+    },
+    {
+      "epoch": 15.47,
+      "grad_norm": 1.5036563873291016,
+      "learning_rate": 6.0709781159056064e-06,
+      "loss": 1.0111,
+      "step": 51715
+    },
+    {
+      "epoch": 15.47,
+      "grad_norm": 4.821185111999512,
+      "learning_rate": 6.0671407838874205e-06,
+      "loss": 0.9488,
+      "step": 51720
+    },
+    {
+      "epoch": 15.48,
+      "grad_norm": 3.7103047370910645,
+      "learning_rate": 6.063304497511382e-06,
+      "loss": 0.9017,
+      "step": 51725
+    },
+    {
+      "epoch": 15.48,
+      "grad_norm": 3.0226051807403564,
+      "learning_rate": 6.059469256989339e-06,
+      "loss": 0.8582,
+      "step": 51730
+    },
+    {
+      "epoch": 15.48,
+      "grad_norm": 2.9177422523498535,
+      "learning_rate": 6.055635062533138e-06,
+      "loss": 0.9762,
+      "step": 51735
+    },
+    {
+      "epoch": 15.48,
+      "grad_norm": 3.6491503715515137,
+      "learning_rate": 6.051801914354511e-06,
+      "loss": 0.9779,
+      "step": 51740
+    },
+    {
+      "epoch": 15.48,
+      "grad_norm": 2.9927515983581543,
+      "learning_rate": 6.047969812665169e-06,
+      "loss": 0.9971,
+      "step": 51745
+    },
+    {
+      "epoch": 15.48,
+      "grad_norm": 2.7734007835388184,
+      "learning_rate": 6.044138757676757e-06,
+      "loss": 1.0923,
+      "step": 51750
+    },
+    {
+      "epoch": 15.48,
+      "grad_norm": 1.9728281497955322,
+      "learning_rate": 6.040308749600854e-06,
+      "loss": 1.0897,
+      "step": 51755
+    },
+    {
+      "epoch": 15.49,
+      "grad_norm": 3.1228339672088623,
+      "learning_rate": 6.036479788648994e-06,
+      "loss": 0.9912,
+      "step": 51760
+    },
+    {
+      "epoch": 15.49,
+      "grad_norm": 2.159838914871216,
+      "learning_rate": 6.032651875032644e-06,
+      "loss": 0.957,
+      "step": 51765
+    },
+    {
+      "epoch": 15.49,
+      "grad_norm": 4.584501266479492,
+      "learning_rate": 6.028825008963215e-06,
+      "loss": 0.8692,
+      "step": 51770
+    },
+    {
+      "epoch": 15.49,
+      "grad_norm": 2.946192979812622,
+      "learning_rate": 6.024999190652058e-06,
+      "loss": 0.9615,
+      "step": 51775
+    },
+    {
+      "epoch": 15.49,
+      "grad_norm": 3.8030338287353516,
+      "learning_rate": 6.0211744203104765e-06,
+      "loss": 0.8554,
+      "step": 51780
+    },
+    {
+      "epoch": 15.49,
+      "grad_norm": 2.243009090423584,
+      "learning_rate": 6.0173506981497045e-06,
+      "loss": 0.9453,
+      "step": 51785
+    },
+    {
+      "epoch": 15.49,
+      "grad_norm": 4.18550968170166,
+      "learning_rate": 6.013528024380921e-06,
+      "loss": 1.0779,
+      "step": 51790
+    },
+    {
+      "epoch": 15.5,
+      "grad_norm": 1.270992636680603,
+      "learning_rate": 6.009706399215248e-06,
+      "loss": 0.882,
+      "step": 51795
+    },
+    {
+      "epoch": 15.5,
+      "grad_norm": 5.413167953491211,
+      "learning_rate": 6.00588582286376e-06,
+      "loss": 0.9145,
+      "step": 51800
+    },
+    {
+      "epoch": 15.5,
+      "grad_norm": 4.414466381072998,
+      "learning_rate": 6.00206629553744e-06,
+      "loss": 0.9951,
+      "step": 51805
+    },
+    {
+      "epoch": 15.5,
+      "grad_norm": 1.6211533546447754,
+      "learning_rate": 5.998247817447264e-06,
+      "loss": 0.8533,
+      "step": 51810
+    },
+    {
+      "epoch": 15.5,
+      "grad_norm": 2.9947216510772705,
+      "learning_rate": 5.994430388804098e-06,
+      "loss": 1.0242,
+      "step": 51815
+    },
+    {
+      "epoch": 15.5,
+      "grad_norm": 1.102436900138855,
+      "learning_rate": 5.9906140098188015e-06,
+      "loss": 0.892,
+      "step": 51820
+    },
+    {
+      "epoch": 15.51,
+      "grad_norm": 2.9922842979431152,
+      "learning_rate": 5.986798680702116e-06,
+      "loss": 0.941,
+      "step": 51825
+    },
+    {
+      "epoch": 15.51,
+      "grad_norm": 7.5577569007873535,
+      "learning_rate": 5.982984401664793e-06,
+      "loss": 1.0939,
+      "step": 51830
+    },
+    {
+      "epoch": 15.51,
+      "grad_norm": 2.929849863052368,
+      "learning_rate": 5.979171172917467e-06,
+      "loss": 0.9213,
+      "step": 51835
+    },
+    {
+      "epoch": 15.51,
+      "grad_norm": 3.348611831665039,
+      "learning_rate": 5.9753589946707464e-06,
+      "loss": 1.044,
+      "step": 51840
+    },
+    {
+      "epoch": 15.51,
+      "grad_norm": 1.479557752609253,
+      "learning_rate": 5.971547867135174e-06,
+      "loss": 0.8942,
+      "step": 51845
+    },
+    {
+      "epoch": 15.51,
+      "grad_norm": 3.3817899227142334,
+      "learning_rate": 5.9677377905212324e-06,
+      "loss": 1.2193,
+      "step": 51850
+    },
+    {
+      "epoch": 15.51,
+      "grad_norm": 2.0309934616088867,
+      "learning_rate": 5.9639287650393504e-06,
+      "loss": 1.1913,
+      "step": 51855
+    },
+    {
+      "epoch": 15.52,
+      "grad_norm": 1.8293267488479614,
+      "learning_rate": 5.9601207908998956e-06,
+      "loss": 0.7932,
+      "step": 51860
+    },
+    {
+      "epoch": 15.52,
+      "grad_norm": 3.6836533546447754,
+      "learning_rate": 5.956313868313179e-06,
+      "loss": 0.978,
+      "step": 51865
+    },
+    {
+      "epoch": 15.52,
+      "grad_norm": 4.970479488372803,
+      "learning_rate": 5.952507997489451e-06,
+      "loss": 1.0691,
+      "step": 51870
+    },
+    {
+      "epoch": 15.52,
+      "grad_norm": 5.052027702331543,
+      "learning_rate": 5.948703178638909e-06,
+      "loss": 1.1021,
+      "step": 51875
+    },
+    {
+      "epoch": 15.52,
+      "grad_norm": 2.7056381702423096,
+      "learning_rate": 5.944899411971688e-06,
+      "loss": 0.8781,
+      "step": 51880
+    },
+    {
+      "epoch": 15.52,
+      "grad_norm": 3.194413900375366,
+      "learning_rate": 5.941096697697865e-06,
+      "loss": 0.9483,
+      "step": 51885
+    },
+    {
+      "epoch": 15.52,
+      "grad_norm": 3.6588056087493896,
+      "learning_rate": 5.937295036027463e-06,
+      "loss": 0.8384,
+      "step": 51890
+    },
+    {
+      "epoch": 15.53,
+      "grad_norm": 1.5710773468017578,
+      "learning_rate": 5.933494427170447e-06,
+      "loss": 0.9579,
+      "step": 51895
+    },
+    {
+      "epoch": 15.53,
+      "grad_norm": 1.679604411125183,
+      "learning_rate": 5.9296948713367e-06,
+      "loss": 1.0501,
+      "step": 51900
+    },
+    {
+      "epoch": 15.53,
+      "grad_norm": 5.178948879241943,
+      "learning_rate": 5.925896368736098e-06,
+      "loss": 0.8546,
+      "step": 51905
+    },
+    {
+      "epoch": 15.53,
+      "grad_norm": 2.5556845664978027,
+      "learning_rate": 5.922098919578398e-06,
+      "loss": 1.0649,
+      "step": 51910
+    },
+    {
+      "epoch": 15.53,
+      "grad_norm": 1.5957000255584717,
+      "learning_rate": 5.918302524073361e-06,
+      "loss": 0.964,
+      "step": 51915
+    },
+    {
+      "epoch": 15.53,
+      "grad_norm": 2.147024631500244,
+      "learning_rate": 5.9145071824306255e-06,
+      "loss": 1.1415,
+      "step": 51920
+    },
+    {
+      "epoch": 15.54,
+      "grad_norm": 2.6177618503570557,
+      "learning_rate": 5.910712894859832e-06,
+      "loss": 0.9363,
+      "step": 51925
+    },
+    {
+      "epoch": 15.54,
+      "grad_norm": 2.6957054138183594,
+      "learning_rate": 5.90691966157052e-06,
+      "loss": 0.8722,
+      "step": 51930
+    },
+    {
+      "epoch": 15.54,
+      "grad_norm": 3.155423402786255,
+      "learning_rate": 5.903127482772186e-06,
+      "loss": 0.974,
+      "step": 51935
+    },
+    {
+      "epoch": 15.54,
+      "grad_norm": 1.4433481693267822,
+      "learning_rate": 5.899336358674273e-06,
+      "loss": 1.0228,
+      "step": 51940
+    },
+    {
+      "epoch": 15.54,
+      "grad_norm": 2.683156967163086,
+      "learning_rate": 5.895546289486159e-06,
+      "loss": 1.0334,
+      "step": 51945
+    },
+    {
+      "epoch": 15.54,
+      "grad_norm": 1.733872890472412,
+      "learning_rate": 5.891757275417165e-06,
+      "loss": 0.8581,
+      "step": 51950
+    },
+    {
+      "epoch": 15.54,
+      "grad_norm": 1.4323784112930298,
+      "learning_rate": 5.8879693166765535e-06,
+      "loss": 0.9999,
+      "step": 51955
+    },
+    {
+      "epoch": 15.55,
+      "grad_norm": 3.402912139892578,
+      "learning_rate": 5.8841824134735295e-06,
+      "loss": 0.8503,
+      "step": 51960
+    },
+    {
+      "epoch": 15.55,
+      "grad_norm": 2.6373584270477295,
+      "learning_rate": 5.880396566017246e-06,
+      "loss": 0.8351,
+      "step": 51965
+    },
+    {
+      "epoch": 15.55,
+      "grad_norm": 1.3156791925430298,
+      "learning_rate": 5.876611774516783e-06,
+      "loss": 0.9677,
+      "step": 51970
+    },
+    {
+      "epoch": 15.55,
+      "grad_norm": 2.961254835128784,
+      "learning_rate": 5.872828039181175e-06,
+      "loss": 0.8483,
+      "step": 51975
+    },
+    {
+      "epoch": 15.55,
+      "grad_norm": 1.9350943565368652,
+      "learning_rate": 5.869045360219391e-06,
+      "loss": 0.9348,
+      "step": 51980
+    },
+    {
+      "epoch": 15.55,
+      "grad_norm": 3.028542995452881,
+      "learning_rate": 5.865263737840349e-06,
+      "loss": 0.9053,
+      "step": 51985
+    },
+    {
+      "epoch": 15.55,
+      "grad_norm": 2.8941564559936523,
+      "learning_rate": 5.861483172252907e-06,
+      "loss": 1.0792,
+      "step": 51990
+    },
+    {
+      "epoch": 15.56,
+      "grad_norm": 2.4049479961395264,
+      "learning_rate": 5.857703663665839e-06,
+      "loss": 0.8029,
+      "step": 51995
+    },
+    {
+      "epoch": 15.56,
+      "grad_norm": 2.2416865825653076,
+      "learning_rate": 5.85392521228792e-06,
+      "loss": 0.8356,
+      "step": 52000
+    },
+    {
+      "epoch": 15.56,
+      "grad_norm": 1.7893521785736084,
+      "learning_rate": 5.850147818327792e-06,
+      "loss": 1.0072,
+      "step": 52005
+    },
+    {
+      "epoch": 15.56,
+      "grad_norm": 4.932260513305664,
+      "learning_rate": 5.8463714819941116e-06,
+      "loss": 0.962,
+      "step": 52010
+    },
+    {
+      "epoch": 15.56,
+      "grad_norm": 1.802653193473816,
+      "learning_rate": 5.84259620349541e-06,
+      "loss": 1.0114,
+      "step": 52015
+    },
+    {
+      "epoch": 15.56,
+      "grad_norm": 4.934607982635498,
+      "learning_rate": 5.838821983040221e-06,
+      "loss": 0.9888,
+      "step": 52020
+    },
+    {
+      "epoch": 15.57,
+      "grad_norm": 3.02289080619812,
+      "learning_rate": 5.835048820836969e-06,
+      "loss": 0.9526,
+      "step": 52025
+    },
+    {
+      "epoch": 15.57,
+      "grad_norm": 2.7780637741088867,
+      "learning_rate": 5.83127671709405e-06,
+      "loss": 0.9735,
+      "step": 52030
+    },
+    {
+      "epoch": 15.57,
+      "grad_norm": 3.7103610038757324,
+      "learning_rate": 5.827505672019795e-06,
+      "loss": 0.8817,
+      "step": 52035
+    },
+    {
+      "epoch": 15.57,
+      "grad_norm": 1.6637479066848755,
+      "learning_rate": 5.82373568582247e-06,
+      "loss": 0.8395,
+      "step": 52040
+    },
+    {
+      "epoch": 15.57,
+      "grad_norm": 4.275225639343262,
+      "learning_rate": 5.819966758710293e-06,
+      "loss": 1.0206,
+      "step": 52045
+    },
+    {
+      "epoch": 15.57,
+      "grad_norm": 3.6473262310028076,
+      "learning_rate": 5.816198890891417e-06,
+      "loss": 0.9276,
+      "step": 52050
+    },
+    {
+      "epoch": 15.57,
+      "grad_norm": 3.159449815750122,
+      "learning_rate": 5.812432082573932e-06,
+      "loss": 1.243,
+      "step": 52055
+    },
+    {
+      "epoch": 15.58,
+      "grad_norm": 3.919372320175171,
+      "learning_rate": 5.8086663339658815e-06,
+      "loss": 1.1114,
+      "step": 52060
+    },
+    {
+      "epoch": 15.58,
+      "grad_norm": 3.567432403564453,
+      "learning_rate": 5.80490164527524e-06,
+      "loss": 0.8347,
+      "step": 52065
+    },
+    {
+      "epoch": 15.58,
+      "grad_norm": 4.436063289642334,
+      "learning_rate": 5.801138016709928e-06,
+      "loss": 1.0756,
+      "step": 52070
+    },
+    {
+      "epoch": 15.58,
+      "grad_norm": 3.9391560554504395,
+      "learning_rate": 5.797375448477807e-06,
+      "loss": 0.925,
+      "step": 52075
+    },
+    {
+      "epoch": 15.58,
+      "grad_norm": 4.531538963317871,
+      "learning_rate": 5.793613940786679e-06,
+      "loss": 0.9541,
+      "step": 52080
+    },
+    {
+      "epoch": 15.58,
+      "grad_norm": 2.420389413833618,
+      "learning_rate": 5.789853493844294e-06,
+      "loss": 0.8691,
+      "step": 52085
+    },
+    {
+      "epoch": 15.58,
+      "grad_norm": 2.241720676422119,
+      "learning_rate": 5.7860941078583284e-06,
+      "loss": 1.0427,
+      "step": 52090
+    },
+    {
+      "epoch": 15.59,
+      "grad_norm": 1.725215196609497,
+      "learning_rate": 5.782335783036422e-06,
+      "loss": 1.0402,
+      "step": 52095
+    },
+    {
+      "epoch": 15.59,
+      "grad_norm": 2.595808744430542,
+      "learning_rate": 5.778578519586119e-06,
+      "loss": 1.0218,
+      "step": 52100
+    },
+    {
+      "epoch": 15.59,
+      "grad_norm": 2.5729830265045166,
+      "learning_rate": 5.774822317714959e-06,
+      "loss": 0.8591,
+      "step": 52105
+    },
+    {
+      "epoch": 15.59,
+      "grad_norm": 3.0964512825012207,
+      "learning_rate": 5.771067177630368e-06,
+      "loss": 0.9244,
+      "step": 52110
+    },
+    {
+      "epoch": 15.59,
+      "grad_norm": 2.0227348804473877,
+      "learning_rate": 5.767313099539762e-06,
+      "loss": 0.7325,
+      "step": 52115
+    },
+    {
+      "epoch": 15.59,
+      "grad_norm": 1.4698666334152222,
+      "learning_rate": 5.763560083650451e-06,
+      "loss": 1.0007,
+      "step": 52120
+    },
+    {
+      "epoch": 15.6,
+      "grad_norm": 4.7408857345581055,
+      "learning_rate": 5.7598081301697276e-06,
+      "loss": 1.1846,
+      "step": 52125
+    },
+    {
+      "epoch": 15.6,
+      "grad_norm": 1.241419792175293,
+      "learning_rate": 5.756057239304799e-06,
+      "loss": 0.9077,
+      "step": 52130
+    },
+    {
+      "epoch": 15.6,
+      "grad_norm": 6.334906578063965,
+      "learning_rate": 5.752307411262825e-06,
+      "loss": 0.9065,
+      "step": 52135
+    },
+    {
+      "epoch": 15.6,
+      "grad_norm": 1.5211411714553833,
+      "learning_rate": 5.748558646250909e-06,
+      "loss": 0.9217,
+      "step": 52140
+    },
+    {
+      "epoch": 15.6,
+      "grad_norm": 2.1848652362823486,
+      "learning_rate": 5.744810944476079e-06,
+      "loss": 0.9956,
+      "step": 52145
+    },
+    {
+      "epoch": 15.6,
+      "grad_norm": 1.960239291191101,
+      "learning_rate": 5.741064306145341e-06,
+      "loss": 1.0003,
+      "step": 52150
+    },
+    {
+      "epoch": 15.6,
+      "grad_norm": 2.8279387950897217,
+      "learning_rate": 5.737318731465593e-06,
+      "loss": 1.0655,
+      "step": 52155
+    },
+    {
+      "epoch": 15.61,
+      "grad_norm": 1.6879123449325562,
+      "learning_rate": 5.733574220643712e-06,
+      "loss": 0.8975,
+      "step": 52160
+    },
+    {
+      "epoch": 15.61,
+      "grad_norm": 1.8145525455474854,
+      "learning_rate": 5.729830773886502e-06,
+      "loss": 0.9761,
+      "step": 52165
+    },
+    {
+      "epoch": 15.61,
+      "grad_norm": 2.2954049110412598,
+      "learning_rate": 5.726088391400705e-06,
+      "loss": 1.0937,
+      "step": 52170
+    },
+    {
+      "epoch": 15.61,
+      "grad_norm": 3.0802268981933594,
+      "learning_rate": 5.722347073393012e-06,
+      "loss": 1.1665,
+      "step": 52175
+    },
+    {
+      "epoch": 15.61,
+      "grad_norm": 1.6516430377960205,
+      "learning_rate": 5.718606820070055e-06,
+      "loss": 0.9886,
+      "step": 52180
+    },
+    {
+      "epoch": 15.61,
+      "grad_norm": 2.8192756175994873,
+      "learning_rate": 5.714867631638399e-06,
+      "loss": 0.9826,
+      "step": 52185
+    },
+    {
+      "epoch": 15.61,
+      "grad_norm": 4.979455471038818,
+      "learning_rate": 5.7111295083045636e-06,
+      "loss": 0.9313,
+      "step": 52190
+    },
+    {
+      "epoch": 15.62,
+      "grad_norm": 2.0236334800720215,
+      "learning_rate": 5.707392450274984e-06,
+      "loss": 0.9639,
+      "step": 52195
+    },
+    {
+      "epoch": 15.62,
+      "grad_norm": 1.1548880338668823,
+      "learning_rate": 5.7036564577560794e-06,
+      "loss": 1.1111,
+      "step": 52200
+    },
+    {
+      "epoch": 15.62,
+      "grad_norm": 3.311762571334839,
+      "learning_rate": 5.699921530954158e-06,
+      "loss": 1.0053,
+      "step": 52205
+    },
+    {
+      "epoch": 15.62,
+      "grad_norm": 2.1778554916381836,
+      "learning_rate": 5.696187670075523e-06,
+      "loss": 1.0448,
+      "step": 52210
+    },
+    {
+      "epoch": 15.62,
+      "grad_norm": 6.210904121398926,
+      "learning_rate": 5.6924548753263694e-06,
+      "loss": 0.9885,
+      "step": 52215
+    },
+    {
+      "epoch": 15.62,
+      "grad_norm": 3.4226996898651123,
+      "learning_rate": 5.688723146912859e-06,
+      "loss": 1.1471,
+      "step": 52220
+    },
+    {
+      "epoch": 15.63,
+      "grad_norm": 3.190202474594116,
+      "learning_rate": 5.684992485041108e-06,
+      "loss": 0.8414,
+      "step": 52225
+    },
+    {
+      "epoch": 15.63,
+      "grad_norm": 3.1030070781707764,
+      "learning_rate": 5.681262889917133e-06,
+      "loss": 1.0518,
+      "step": 52230
+    },
+    {
+      "epoch": 15.63,
+      "grad_norm": 1.8778613805770874,
+      "learning_rate": 5.677534361746939e-06,
+      "loss": 1.0345,
+      "step": 52235
+    },
+    {
+      "epoch": 15.63,
+      "grad_norm": 2.847775459289551,
+      "learning_rate": 5.6738069007364255e-06,
+      "loss": 1.0059,
+      "step": 52240
+    },
+    {
+      "epoch": 15.63,
+      "grad_norm": 1.932309627532959,
+      "learning_rate": 5.670080507091483e-06,
+      "loss": 0.9203,
+      "step": 52245
+    },
+    {
+      "epoch": 15.63,
+      "grad_norm": 13.216365814208984,
+      "learning_rate": 5.666355181017893e-06,
+      "loss": 0.9181,
+      "step": 52250
+    },
+    {
+      "epoch": 15.63,
+      "grad_norm": 1.8943175077438354,
+      "learning_rate": 5.662630922721413e-06,
+      "loss": 0.9876,
+      "step": 52255
+    },
+    {
+      "epoch": 15.64,
+      "grad_norm": 1.8065849542617798,
+      "learning_rate": 5.658907732407726e-06,
+      "loss": 0.9735,
+      "step": 52260
+    },
+    {
+      "epoch": 15.64,
+      "grad_norm": 1.6062184572219849,
+      "learning_rate": 5.65518561028246e-06,
+      "loss": 0.8603,
+      "step": 52265
+    },
+    {
+      "epoch": 15.64,
+      "grad_norm": 1.6171048879623413,
+      "learning_rate": 5.651464556551186e-06,
+      "loss": 1.1217,
+      "step": 52270
+    },
+    {
+      "epoch": 15.64,
+      "grad_norm": 2.9715936183929443,
+      "learning_rate": 5.647744571419411e-06,
+      "loss": 1.195,
+      "step": 52275
+    },
+    {
+      "epoch": 15.64,
+      "grad_norm": 3.835232734680176,
+      "learning_rate": 5.644025655092591e-06,
+      "loss": 1.1408,
+      "step": 52280
+    },
+    {
+      "epoch": 15.64,
+      "grad_norm": 3.526592969894409,
+      "learning_rate": 5.6403078077761186e-06,
+      "loss": 0.8728,
+      "step": 52285
+    },
+    {
+      "epoch": 15.64,
+      "grad_norm": 4.8948187828063965,
+      "learning_rate": 5.636591029675311e-06,
+      "loss": 0.8828,
+      "step": 52290
+    },
+    {
+      "epoch": 15.65,
+      "grad_norm": 1.3252886533737183,
+      "learning_rate": 5.632875320995468e-06,
+      "loss": 0.9628,
+      "step": 52295
+    },
+    {
+      "epoch": 15.65,
+      "grad_norm": 3.719871759414673,
+      "learning_rate": 5.629160681941772e-06,
+      "loss": 0.8964,
+      "step": 52300
+    },
+    {
+      "epoch": 15.65,
+      "grad_norm": 2.8799922466278076,
+      "learning_rate": 5.6254471127194064e-06,
+      "loss": 1.073,
+      "step": 52305
+    },
+    {
+      "epoch": 15.65,
+      "grad_norm": 3.387589454650879,
+      "learning_rate": 5.621734613533458e-06,
+      "loss": 0.7555,
+      "step": 52310
+    },
+    {
+      "epoch": 15.65,
+      "grad_norm": 3.153660535812378,
+      "learning_rate": 5.618023184588964e-06,
+      "loss": 1.031,
+      "step": 52315
+    },
+    {
+      "epoch": 15.65,
+      "grad_norm": 2.53116774559021,
+      "learning_rate": 5.6143128260909115e-06,
+      "loss": 0.8774,
+      "step": 52320
+    },
+    {
+      "epoch": 15.66,
+      "grad_norm": 2.402709722518921,
+      "learning_rate": 5.610603538244197e-06,
+      "loss": 0.9746,
+      "step": 52325
+    },
+    {
+      "epoch": 15.66,
+      "grad_norm": 3.300175666809082,
+      "learning_rate": 5.606895321253705e-06,
+      "loss": 0.9147,
+      "step": 52330
+    },
+    {
+      "epoch": 15.66,
+      "grad_norm": 2.9742469787597656,
+      "learning_rate": 5.603188175324217e-06,
+      "loss": 1.1199,
+      "step": 52335
+    },
+    {
+      "epoch": 15.66,
+      "grad_norm": 2.81020450592041,
+      "learning_rate": 5.599482100660497e-06,
+      "loss": 0.961,
+      "step": 52340
+    },
+    {
+      "epoch": 15.66,
+      "grad_norm": 2.7677392959594727,
+      "learning_rate": 5.595777097467206e-06,
+      "loss": 1.1416,
+      "step": 52345
+    },
+    {
+      "epoch": 15.66,
+      "grad_norm": 2.42520809173584,
+      "learning_rate": 5.59207316594898e-06,
+      "loss": 1.1443,
+      "step": 52350
+    },
+    {
+      "epoch": 15.66,
+      "grad_norm": 2.994718313217163,
+      "learning_rate": 5.588370306310375e-06,
+      "loss": 1.0766,
+      "step": 52355
+    },
+    {
+      "epoch": 15.67,
+      "grad_norm": 2.0844037532806396,
+      "learning_rate": 5.584668518755904e-06,
+      "loss": 0.9827,
+      "step": 52360
+    },
+    {
+      "epoch": 15.67,
+      "grad_norm": 2.0058093070983887,
+      "learning_rate": 5.580967803490008e-06,
+      "loss": 0.925,
+      "step": 52365
+    },
+    {
+      "epoch": 15.67,
+      "grad_norm": 2.180823564529419,
+      "learning_rate": 5.577268160717078e-06,
+      "loss": 0.9218,
+      "step": 52370
+    },
+    {
+      "epoch": 15.67,
+      "grad_norm": 4.118743419647217,
+      "learning_rate": 5.573569590641436e-06,
+      "loss": 0.9062,
+      "step": 52375
+    },
+    {
+      "epoch": 15.67,
+      "grad_norm": 2.850475788116455,
+      "learning_rate": 5.569872093467354e-06,
+      "loss": 1.0542,
+      "step": 52380
+    },
+    {
+      "epoch": 15.67,
+      "grad_norm": 4.366810321807861,
+      "learning_rate": 5.566175669399037e-06,
+      "loss": 0.9499,
+      "step": 52385
+    },
+    {
+      "epoch": 15.67,
+      "grad_norm": 1.8252012729644775,
+      "learning_rate": 5.562480318640642e-06,
+      "loss": 1.0941,
+      "step": 52390
+    },
+    {
+      "epoch": 15.68,
+      "grad_norm": 6.694037437438965,
+      "learning_rate": 5.558786041396252e-06,
+      "loss": 1.0343,
+      "step": 52395
+    },
+    {
+      "epoch": 15.68,
+      "grad_norm": 3.3669612407684326,
+      "learning_rate": 5.555092837869902e-06,
+      "loss": 1.0908,
+      "step": 52400
+    },
+    {
+      "epoch": 15.68,
+      "grad_norm": 2.897421360015869,
+      "learning_rate": 5.551400708265561e-06,
+      "loss": 1.0238,
+      "step": 52405
+    },
+    {
+      "epoch": 15.68,
+      "grad_norm": 2.594632387161255,
+      "learning_rate": 5.5477096527871445e-06,
+      "loss": 0.8819,
+      "step": 52410
+    },
+    {
+      "epoch": 15.68,
+      "grad_norm": 2.2491109371185303,
+      "learning_rate": 5.544019671638512e-06,
+      "loss": 0.9708,
+      "step": 52415
+    },
+    {
+      "epoch": 15.68,
+      "grad_norm": 2.392650604248047,
+      "learning_rate": 5.5403307650234355e-06,
+      "loss": 0.9749,
+      "step": 52420
+    },
+    {
+      "epoch": 15.68,
+      "grad_norm": 2.6764276027679443,
+      "learning_rate": 5.536642933145677e-06,
+      "loss": 1.1067,
+      "step": 52425
+    },
+    {
+      "epoch": 15.69,
+      "grad_norm": 1.6986552476882935,
+      "learning_rate": 5.532956176208884e-06,
+      "loss": 0.8994,
+      "step": 52430
+    },
+    {
+      "epoch": 15.69,
+      "grad_norm": 3.7357470989227295,
+      "learning_rate": 5.5292704944167026e-06,
+      "loss": 0.9956,
+      "step": 52435
+    },
+    {
+      "epoch": 15.69,
+      "grad_norm": 2.2799248695373535,
+      "learning_rate": 5.525585887972657e-06,
+      "loss": 1.203,
+      "step": 52440
+    },
+    {
+      "epoch": 15.69,
+      "grad_norm": 3.3140718936920166,
+      "learning_rate": 5.521902357080275e-06,
+      "loss": 0.8632,
+      "step": 52445
+    },
+    {
+      "epoch": 15.69,
+      "grad_norm": 2.275482416152954,
+      "learning_rate": 5.518219901942972e-06,
+      "loss": 0.9642,
+      "step": 52450
+    },
+    {
+      "epoch": 15.69,
+      "grad_norm": 2.353581428527832,
+      "learning_rate": 5.514538522764135e-06,
+      "loss": 1.0428,
+      "step": 52455
+    },
+    {
+      "epoch": 15.7,
+      "grad_norm": 3.2124972343444824,
+      "learning_rate": 5.510858219747078e-06,
+      "loss": 0.9647,
+      "step": 52460
+    },
+    {
+      "epoch": 15.7,
+      "grad_norm": 2.8147921562194824,
+      "learning_rate": 5.507178993095067e-06,
+      "loss": 1.0069,
+      "step": 52465
+    },
+    {
+      "epoch": 15.7,
+      "grad_norm": 2.2177577018737793,
+      "learning_rate": 5.503500843011297e-06,
+      "loss": 1.0085,
+      "step": 52470
+    },
+    {
+      "epoch": 15.7,
+      "grad_norm": 2.825120210647583,
+      "learning_rate": 5.4998237696989085e-06,
+      "loss": 1.0708,
+      "step": 52475
+    },
+    {
+      "epoch": 15.7,
+      "grad_norm": 3.487179756164551,
+      "learning_rate": 5.496147773360988e-06,
+      "loss": 1.0576,
+      "step": 52480
+    },
+    {
+      "epoch": 15.7,
+      "grad_norm": 1.7475980520248413,
+      "learning_rate": 5.492472854200551e-06,
+      "loss": 1.0345,
+      "step": 52485
+    },
+    {
+      "epoch": 15.7,
+      "grad_norm": 1.6811771392822266,
+      "learning_rate": 5.488799012420559e-06,
+      "loss": 1.0522,
+      "step": 52490
+    },
+    {
+      "epoch": 15.71,
+      "grad_norm": 3.03035044670105,
+      "learning_rate": 5.4851262482239205e-06,
+      "loss": 1.0216,
+      "step": 52495
+    },
+    {
+      "epoch": 15.71,
+      "grad_norm": 1.8432042598724365,
+      "learning_rate": 5.481454561813473e-06,
+      "loss": 1.1813,
+      "step": 52500
+    },
+    {
+      "epoch": 15.71,
+      "grad_norm": 3.2820653915405273,
+      "learning_rate": 5.477783953392002e-06,
+      "loss": 0.9292,
+      "step": 52505
+    },
+    {
+      "epoch": 15.71,
+      "grad_norm": 1.9322882890701294,
+      "learning_rate": 5.474114423162236e-06,
+      "loss": 0.9244,
+      "step": 52510
+    },
+    {
+      "epoch": 15.71,
+      "grad_norm": 4.189065933227539,
+      "learning_rate": 5.470445971326821e-06,
+      "loss": 1.1351,
+      "step": 52515
+    },
+    {
+      "epoch": 15.71,
+      "grad_norm": 1.8628734350204468,
+      "learning_rate": 5.4667785980883925e-06,
+      "loss": 0.9412,
+      "step": 52520
+    },
+    {
+      "epoch": 15.71,
+      "grad_norm": 3.525163412094116,
+      "learning_rate": 5.463112303649462e-06,
+      "loss": 0.8848,
+      "step": 52525
+    },
+    {
+      "epoch": 15.72,
+      "grad_norm": 3.689204692840576,
+      "learning_rate": 5.459447088212544e-06,
+      "loss": 0.938,
+      "step": 52530
+    },
+    {
+      "epoch": 15.72,
+      "grad_norm": 7.576735019683838,
+      "learning_rate": 5.45578295198004e-06,
+      "loss": 0.6826,
+      "step": 52535
+    },
+    {
+      "epoch": 15.72,
+      "grad_norm": 3.6514053344726562,
+      "learning_rate": 5.452119895154343e-06,
+      "loss": 0.8254,
+      "step": 52540
+    },
+    {
+      "epoch": 15.72,
+      "grad_norm": 2.51778507232666,
+      "learning_rate": 5.4484579179377385e-06,
+      "loss": 0.9759,
+      "step": 52545
+    },
+    {
+      "epoch": 15.72,
+      "grad_norm": 3.0608720779418945,
+      "learning_rate": 5.444797020532477e-06,
+      "loss": 1.0327,
+      "step": 52550
+    },
+    {
+      "epoch": 15.72,
+      "grad_norm": 3.0184848308563232,
+      "learning_rate": 5.441137203140753e-06,
+      "loss": 0.9639,
+      "step": 52555
+    },
+    {
+      "epoch": 15.73,
+      "grad_norm": 3.508617401123047,
+      "learning_rate": 5.43747846596469e-06,
+      "loss": 0.8989,
+      "step": 52560
+    },
+    {
+      "epoch": 15.73,
+      "grad_norm": 2.4190101623535156,
+      "learning_rate": 5.433820809206358e-06,
+      "loss": 1.0506,
+      "step": 52565
+    },
+    {
+      "epoch": 15.73,
+      "grad_norm": 3.336304187774658,
+      "learning_rate": 5.430164233067764e-06,
+      "loss": 1.0166,
+      "step": 52570
+    },
+    {
+      "epoch": 15.73,
+      "grad_norm": 3.363173246383667,
+      "learning_rate": 5.426508737750857e-06,
+      "loss": 1.0955,
+      "step": 52575
+    },
+    {
+      "epoch": 15.73,
+      "grad_norm": 2.3108577728271484,
+      "learning_rate": 5.422854323457527e-06,
+      "loss": 1.0257,
+      "step": 52580
+    },
+    {
+      "epoch": 15.73,
+      "grad_norm": 2.772678852081299,
+      "learning_rate": 5.419200990389603e-06,
+      "loss": 0.9683,
+      "step": 52585
+    },
+    {
+      "epoch": 15.73,
+      "grad_norm": 2.349119186401367,
+      "learning_rate": 5.415548738748857e-06,
+      "loss": 0.9251,
+      "step": 52590
+    },
+    {
+      "epoch": 15.74,
+      "grad_norm": 2.551915407180786,
+      "learning_rate": 5.411897568736995e-06,
+      "loss": 1.0174,
+      "step": 52595
+    },
+    {
+      "epoch": 15.74,
+      "grad_norm": 2.7730255126953125,
+      "learning_rate": 5.408247480555673e-06,
+      "loss": 0.9942,
+      "step": 52600
+    },
+    {
+      "epoch": 15.74,
+      "grad_norm": 4.1052141189575195,
+      "learning_rate": 5.4045984744064836e-06,
+      "loss": 0.9015,
+      "step": 52605
+    },
+    {
+      "epoch": 15.74,
+      "grad_norm": 2.427316665649414,
+      "learning_rate": 5.400950550490935e-06,
+      "loss": 1.129,
+      "step": 52610
+    },
+    {
+      "epoch": 15.74,
+      "grad_norm": 2.9681787490844727,
+      "learning_rate": 5.397303709010534e-06,
+      "loss": 1.1242,
+      "step": 52615
+    },
+    {
+      "epoch": 15.74,
+      "grad_norm": 2.8961920738220215,
+      "learning_rate": 5.393657950166659e-06,
+      "loss": 0.9106,
+      "step": 52620
+    },
+    {
+      "epoch": 15.74,
+      "grad_norm": 3.401855945587158,
+      "learning_rate": 5.390013274160685e-06,
+      "loss": 1.0304,
+      "step": 52625
+    },
+    {
+      "epoch": 15.75,
+      "grad_norm": 5.901130676269531,
+      "learning_rate": 5.386369681193884e-06,
+      "loss": 0.8025,
+      "step": 52630
+    },
+    {
+      "epoch": 15.75,
+      "grad_norm": 2.0417873859405518,
+      "learning_rate": 5.382727171467514e-06,
+      "loss": 1.0043,
+      "step": 52635
+    },
+    {
+      "epoch": 15.75,
+      "grad_norm": 17.817447662353516,
+      "learning_rate": 5.379085745182721e-06,
+      "loss": 0.8185,
+      "step": 52640
+    },
+    {
+      "epoch": 15.75,
+      "grad_norm": 1.411085844039917,
+      "learning_rate": 5.37544540254063e-06,
+      "loss": 1.1456,
+      "step": 52645
+    },
+    {
+      "epoch": 15.75,
+      "grad_norm": 3.324824810028076,
+      "learning_rate": 5.371806143742289e-06,
+      "loss": 0.9423,
+      "step": 52650
+    },
+    {
+      "epoch": 15.75,
+      "grad_norm": 2.4922614097595215,
+      "learning_rate": 5.368167968988691e-06,
+      "loss": 0.9954,
+      "step": 52655
+    },
+    {
+      "epoch": 15.76,
+      "grad_norm": 2.6019651889801025,
+      "learning_rate": 5.364530878480772e-06,
+      "loss": 0.9052,
+      "step": 52660
+    },
+    {
+      "epoch": 15.76,
+      "grad_norm": 5.232183456420898,
+      "learning_rate": 5.360894872419403e-06,
+      "loss": 0.7985,
+      "step": 52665
+    },
+    {
+      "epoch": 15.76,
+      "grad_norm": 4.952908515930176,
+      "learning_rate": 5.357259951005397e-06,
+      "loss": 1.1789,
+      "step": 52670
+    },
+    {
+      "epoch": 15.76,
+      "grad_norm": 2.208289861679077,
+      "learning_rate": 5.3536261144395054e-06,
+      "loss": 0.8794,
+      "step": 52675
+    },
+    {
+      "epoch": 15.76,
+      "grad_norm": 2.049973487854004,
+      "learning_rate": 5.3499933629224204e-06,
+      "loss": 0.7753,
+      "step": 52680
+    },
+    {
+      "epoch": 15.76,
+      "grad_norm": 2.3829286098480225,
+      "learning_rate": 5.346361696654781e-06,
+      "loss": 0.9712,
+      "step": 52685
+    },
+    {
+      "epoch": 15.76,
+      "grad_norm": 33.99540328979492,
+      "learning_rate": 5.3427311158371515e-06,
+      "loss": 1.0222,
+      "step": 52690
+    },
+    {
+      "epoch": 15.77,
+      "grad_norm": 1.5256500244140625,
+      "learning_rate": 5.339101620670054e-06,
+      "loss": 1.0975,
+      "step": 52695
+    },
+    {
+      "epoch": 15.77,
+      "grad_norm": 3.328197717666626,
+      "learning_rate": 5.3354732113539415e-06,
+      "loss": 0.872,
+      "step": 52700
+    },
+    {
+      "epoch": 15.77,
+      "grad_norm": 1.9230583906173706,
+      "learning_rate": 5.33184588808919e-06,
+      "loss": 0.9616,
+      "step": 52705
+    },
+    {
+      "epoch": 15.77,
+      "grad_norm": 4.239948749542236,
+      "learning_rate": 5.328219651076163e-06,
+      "loss": 0.9001,
+      "step": 52710
+    },
+    {
+      "epoch": 15.77,
+      "grad_norm": 3.32629656791687,
+      "learning_rate": 5.324594500515101e-06,
+      "loss": 1.0787,
+      "step": 52715
+    },
+    {
+      "epoch": 15.77,
+      "grad_norm": 1.680390477180481,
+      "learning_rate": 5.3209704366062455e-06,
+      "loss": 1.0742,
+      "step": 52720
+    },
+    {
+      "epoch": 15.77,
+      "grad_norm": 3.082519769668579,
+      "learning_rate": 5.317347459549726e-06,
+      "loss": 0.8436,
+      "step": 52725
+    },
+    {
+      "epoch": 15.78,
+      "grad_norm": 5.2578654289245605,
+      "learning_rate": 5.313725569545661e-06,
+      "loss": 1.0528,
+      "step": 52730
+    },
+    {
+      "epoch": 15.78,
+      "grad_norm": 3.4256768226623535,
+      "learning_rate": 5.310104766794064e-06,
+      "loss": 1.0051,
+      "step": 52735
+    },
+    {
+      "epoch": 15.78,
+      "grad_norm": 1.8970190286636353,
+      "learning_rate": 5.306485051494911e-06,
+      "loss": 1.1178,
+      "step": 52740
+    },
+    {
+      "epoch": 15.78,
+      "grad_norm": 6.223118305206299,
+      "learning_rate": 5.302866423848121e-06,
+      "loss": 1.0483,
+      "step": 52745
+    },
+    {
+      "epoch": 15.78,
+      "grad_norm": 3.551870822906494,
+      "learning_rate": 5.2992488840535425e-06,
+      "loss": 0.7908,
+      "step": 52750
+    },
+    {
+      "epoch": 15.78,
+      "grad_norm": 2.2895681858062744,
+      "learning_rate": 5.295632432310973e-06,
+      "loss": 0.9522,
+      "step": 52755
+    },
+    {
+      "epoch": 15.79,
+      "grad_norm": 3.403578281402588,
+      "learning_rate": 5.2920170688201405e-06,
+      "loss": 0.9253,
+      "step": 52760
+    },
+    {
+      "epoch": 15.79,
+      "grad_norm": 4.709850788116455,
+      "learning_rate": 5.2884027937807225e-06,
+      "loss": 0.8115,
+      "step": 52765
+    },
+    {
+      "epoch": 15.79,
+      "grad_norm": 2.8941822052001953,
+      "learning_rate": 5.284789607392326e-06,
+      "loss": 1.031,
+      "step": 52770
+    },
+    {
+      "epoch": 15.79,
+      "grad_norm": 3.633200168609619,
+      "learning_rate": 5.28117750985451e-06,
+      "loss": 1.0308,
+      "step": 52775
+    },
+    {
+      "epoch": 15.79,
+      "grad_norm": 2.2992610931396484,
+      "learning_rate": 5.27756650136676e-06,
+      "loss": 0.7602,
+      "step": 52780
+    },
+    {
+      "epoch": 15.79,
+      "grad_norm": 5.884945392608643,
+      "learning_rate": 5.2739565821285145e-06,
+      "loss": 0.8777,
+      "step": 52785
+    },
+    {
+      "epoch": 15.79,
+      "grad_norm": 2.312242031097412,
+      "learning_rate": 5.270347752339142e-06,
+      "loss": 0.8445,
+      "step": 52790
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 3.9331436157226562,
+      "learning_rate": 5.266740012197954e-06,
+      "loss": 1.1452,
+      "step": 52795
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 1.6924293041229248,
+      "learning_rate": 5.263133361904204e-06,
+      "loss": 0.9707,
+      "step": 52800
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 1.7212278842926025,
+      "learning_rate": 5.259527801657091e-06,
+      "loss": 1.1103,
+      "step": 52805
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 2.418543815612793,
+      "learning_rate": 5.255923331655724e-06,
+      "loss": 1.0522,
+      "step": 52810
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 1.562758445739746,
+      "learning_rate": 5.252319952099202e-06,
+      "loss": 0.9675,
+      "step": 52815
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 3.524245023727417,
+      "learning_rate": 5.248717663186511e-06,
+      "loss": 1.0859,
+      "step": 52820
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 3.3297927379608154,
+      "learning_rate": 5.245116465116625e-06,
+      "loss": 0.9522,
+      "step": 52825
+    },
+    {
+      "epoch": 15.81,
+      "grad_norm": 4.018186092376709,
+      "learning_rate": 5.241516358088417e-06,
+      "loss": 1.0692,
+      "step": 52830
+    },
+    {
+      "epoch": 15.81,
+      "grad_norm": 1.4995861053466797,
+      "learning_rate": 5.237917342300719e-06,
+      "loss": 1.0195,
+      "step": 52835
+    },
+    {
+      "epoch": 15.81,
+      "grad_norm": 2.2528059482574463,
+      "learning_rate": 5.234319417952308e-06,
+      "loss": 1.0619,
+      "step": 52840
+    },
+    {
+      "epoch": 15.81,
+      "grad_norm": 3.694645881652832,
+      "learning_rate": 5.23072258524189e-06,
+      "loss": 1.0082,
+      "step": 52845
+    },
+    {
+      "epoch": 15.81,
+      "grad_norm": 2.391193151473999,
+      "learning_rate": 5.227126844368113e-06,
+      "loss": 1.032,
+      "step": 52850
+    },
+    {
+      "epoch": 15.81,
+      "grad_norm": 2.8515758514404297,
+      "learning_rate": 5.223532195529571e-06,
+      "loss": 0.9227,
+      "step": 52855
+    },
+    {
+      "epoch": 15.82,
+      "grad_norm": 1.8766295909881592,
+      "learning_rate": 5.2199386389247876e-06,
+      "loss": 0.9314,
+      "step": 52860
+    },
+    {
+      "epoch": 15.82,
+      "grad_norm": 1.6634045839309692,
+      "learning_rate": 5.216346174752232e-06,
+      "loss": 0.9464,
+      "step": 52865
+    },
+    {
+      "epoch": 15.82,
+      "grad_norm": 2.1022915840148926,
+      "learning_rate": 5.212754803210312e-06,
+      "loss": 0.9514,
+      "step": 52870
+    },
+    {
+      "epoch": 15.82,
+      "grad_norm": 2.1778430938720703,
+      "learning_rate": 5.209164524497378e-06,
+      "loss": 0.9898,
+      "step": 52875
+    },
+    {
+      "epoch": 15.82,
+      "grad_norm": 1.7852226495742798,
+      "learning_rate": 5.205575338811719e-06,
+      "loss": 0.9918,
+      "step": 52880
+    },
+    {
+      "epoch": 15.82,
+      "grad_norm": 2.882033586502075,
+      "learning_rate": 5.201987246351556e-06,
+      "loss": 0.8887,
+      "step": 52885
+    },
+    {
+      "epoch": 15.82,
+      "grad_norm": 5.0355939865112305,
+      "learning_rate": 5.198400247315058e-06,
+      "loss": 1.0556,
+      "step": 52890
+    },
+    {
+      "epoch": 15.83,
+      "grad_norm": 1.298420786857605,
+      "learning_rate": 5.194814341900331e-06,
+      "loss": 1.1109,
+      "step": 52895
+    },
+    {
+      "epoch": 15.83,
+      "grad_norm": 1.944771409034729,
+      "learning_rate": 5.1912295303054305e-06,
+      "loss": 0.9752,
+      "step": 52900
+    },
+    {
+      "epoch": 15.83,
+      "grad_norm": 2.584871768951416,
+      "learning_rate": 5.187645812728317e-06,
+      "loss": 0.9854,
+      "step": 52905
+    },
+    {
+      "epoch": 15.83,
+      "grad_norm": 2.4717025756835938,
+      "learning_rate": 5.184063189366948e-06,
+      "loss": 1.0902,
+      "step": 52910
+    },
+    {
+      "epoch": 15.83,
+      "grad_norm": 1.5633115768432617,
+      "learning_rate": 5.1804816604191565e-06,
+      "loss": 0.9654,
+      "step": 52915
+    },
+    {
+      "epoch": 15.83,
+      "grad_norm": 2.15081787109375,
+      "learning_rate": 5.176901226082776e-06,
+      "loss": 0.9748,
+      "step": 52920
+    },
+    {
+      "epoch": 15.83,
+      "grad_norm": 1.1734721660614014,
+      "learning_rate": 5.173321886555527e-06,
+      "loss": 1.0343,
+      "step": 52925
+    },
+    {
+      "epoch": 15.84,
+      "grad_norm": 3.0947458744049072,
+      "learning_rate": 5.169743642035102e-06,
+      "loss": 0.9797,
+      "step": 52930
+    },
+    {
+      "epoch": 15.84,
+      "grad_norm": 5.485124588012695,
+      "learning_rate": 5.166166492719124e-06,
+      "loss": 0.874,
+      "step": 52935
+    },
+    {
+      "epoch": 15.84,
+      "grad_norm": 2.063061475753784,
+      "learning_rate": 5.162590438805156e-06,
+      "loss": 0.9525,
+      "step": 52940
+    },
+    {
+      "epoch": 15.84,
+      "grad_norm": 1.3920997381210327,
+      "learning_rate": 5.159015480490695e-06,
+      "loss": 1.0252,
+      "step": 52945
+    },
+    {
+      "epoch": 15.84,
+      "grad_norm": 1.3368240594863892,
+      "learning_rate": 5.155441617973189e-06,
+      "loss": 1.0572,
+      "step": 52950
+    },
+    {
+      "epoch": 15.84,
+      "grad_norm": 3.9613707065582275,
+      "learning_rate": 5.151868851450012e-06,
+      "loss": 0.9808,
+      "step": 52955
+    },
+    {
+      "epoch": 15.85,
+      "grad_norm": 3.1625585556030273,
+      "learning_rate": 5.148297181118489e-06,
+      "loss": 0.9818,
+      "step": 52960
+    },
+    {
+      "epoch": 15.85,
+      "grad_norm": 3.069246292114258,
+      "learning_rate": 5.1447266071758785e-06,
+      "loss": 0.9461,
+      "step": 52965
+    },
+    {
+      "epoch": 15.85,
+      "grad_norm": 3.266118288040161,
+      "learning_rate": 5.14115712981938e-06,
+      "loss": 0.958,
+      "step": 52970
+    },
+    {
+      "epoch": 15.85,
+      "grad_norm": 2.903211832046509,
+      "learning_rate": 5.137588749246128e-06,
+      "loss": 0.9459,
+      "step": 52975
+    },
+    {
+      "epoch": 15.85,
+      "grad_norm": 4.388901710510254,
+      "learning_rate": 5.134021465653205e-06,
+      "loss": 0.8478,
+      "step": 52980
+    },
+    {
+      "epoch": 15.85,
+      "grad_norm": 3.907933473587036,
+      "learning_rate": 5.130455279237625e-06,
+      "loss": 1.1191,
+      "step": 52985
+    },
+    {
+      "epoch": 15.85,
+      "grad_norm": 1.9974335432052612,
+      "learning_rate": 5.126890190196348e-06,
+      "loss": 1.1045,
+      "step": 52990
+    },
+    {
+      "epoch": 15.86,
+      "grad_norm": 4.463841438293457,
+      "learning_rate": 5.1233261987262775e-06,
+      "loss": 0.9843,
+      "step": 52995
+    },
+    {
+      "epoch": 15.86,
+      "grad_norm": 4.327840805053711,
+      "learning_rate": 5.119763305024225e-06,
+      "loss": 0.9954,
+      "step": 53000
+    },
+    {
+      "epoch": 15.86,
+      "grad_norm": 1.5290764570236206,
+      "learning_rate": 5.116201509286997e-06,
+      "loss": 1.2544,
+      "step": 53005
+    },
+    {
+      "epoch": 15.86,
+      "grad_norm": 2.252065420150757,
+      "learning_rate": 5.1126408117112755e-06,
+      "loss": 0.9876,
+      "step": 53010
+    },
+    {
+      "epoch": 15.86,
+      "grad_norm": 3.484567880630493,
+      "learning_rate": 5.109081212493744e-06,
+      "loss": 0.8204,
+      "step": 53015
+    },
+    {
+      "epoch": 15.86,
+      "grad_norm": 4.542111873626709,
+      "learning_rate": 5.105522711830968e-06,
+      "loss": 0.9088,
+      "step": 53020
+    },
+    {
+      "epoch": 15.86,
+      "grad_norm": 2.7944681644439697,
+      "learning_rate": 5.101965309919507e-06,
+      "loss": 0.7633,
+      "step": 53025
+    },
+    {
+      "epoch": 15.87,
+      "grad_norm": 3.740417718887329,
+      "learning_rate": 5.098409006955812e-06,
+      "loss": 0.8264,
+      "step": 53030
+    },
+    {
+      "epoch": 15.87,
+      "grad_norm": 2.1556479930877686,
+      "learning_rate": 5.094853803136296e-06,
+      "loss": 1.0365,
+      "step": 53035
+    },
+    {
+      "epoch": 15.87,
+      "grad_norm": 3.4103243350982666,
+      "learning_rate": 5.091299698657326e-06,
+      "loss": 1.164,
+      "step": 53040
+    },
+    {
+      "epoch": 15.87,
+      "grad_norm": 2.4072625637054443,
+      "learning_rate": 5.087746693715167e-06,
+      "loss": 0.7972,
+      "step": 53045
+    },
+    {
+      "epoch": 15.87,
+      "grad_norm": 2.5609822273254395,
+      "learning_rate": 5.084194788506075e-06,
+      "loss": 1.0668,
+      "step": 53050
+    },
+    {
+      "epoch": 15.87,
+      "grad_norm": 1.8932149410247803,
+      "learning_rate": 5.0806439832261935e-06,
+      "loss": 0.9741,
+      "step": 53055
+    },
+    {
+      "epoch": 15.87,
+      "grad_norm": 2.655104160308838,
+      "learning_rate": 5.077094278071643e-06,
+      "loss": 0.927,
+      "step": 53060
+    },
+    {
+      "epoch": 15.88,
+      "grad_norm": 2.491745948791504,
+      "learning_rate": 5.073545673238467e-06,
+      "loss": 0.8797,
+      "step": 53065
+    },
+    {
+      "epoch": 15.88,
+      "grad_norm": 2.7281858921051025,
+      "learning_rate": 5.069998168922649e-06,
+      "loss": 0.9462,
+      "step": 53070
+    },
+    {
+      "epoch": 15.88,
+      "grad_norm": 3.0326929092407227,
+      "learning_rate": 5.066451765320116e-06,
+      "loss": 0.997,
+      "step": 53075
+    },
+    {
+      "epoch": 15.88,
+      "grad_norm": 2.410046339035034,
+      "learning_rate": 5.062906462626732e-06,
+      "loss": 1.0823,
+      "step": 53080
+    },
+    {
+      "epoch": 15.88,
+      "grad_norm": 4.804044246673584,
+      "learning_rate": 5.059362261038303e-06,
+      "loss": 1.0974,
+      "step": 53085
+    },
+    {
+      "epoch": 15.88,
+      "grad_norm": 22.172578811645508,
+      "learning_rate": 5.0558191607505745e-06,
+      "loss": 0.944,
+      "step": 53090
+    },
+    {
+      "epoch": 15.89,
+      "grad_norm": 4.430202007293701,
+      "learning_rate": 5.052277161959209e-06,
+      "loss": 0.9268,
+      "step": 53095
+    },
+    {
+      "epoch": 15.89,
+      "grad_norm": 3.1318132877349854,
+      "learning_rate": 5.048736264859854e-06,
+      "loss": 0.8596,
+      "step": 53100
+    },
+    {
+      "epoch": 15.89,
+      "grad_norm": 2.395545244216919,
+      "learning_rate": 5.045196469648045e-06,
+      "loss": 0.8246,
+      "step": 53105
+    },
+    {
+      "epoch": 15.89,
+      "grad_norm": 1.1723501682281494,
+      "learning_rate": 5.041657776519307e-06,
+      "loss": 1.017,
+      "step": 53110
+    },
+    {
+      "epoch": 15.89,
+      "grad_norm": 3.7153756618499756,
+      "learning_rate": 5.038120185669051e-06,
+      "loss": 0.8919,
+      "step": 53115
+    },
+    {
+      "epoch": 15.89,
+      "grad_norm": 1.7875972986221313,
+      "learning_rate": 5.034583697292675e-06,
+      "loss": 0.9199,
+      "step": 53120
+    },
+    {
+      "epoch": 15.89,
+      "grad_norm": 3.715594530105591,
+      "learning_rate": 5.031048311585493e-06,
+      "loss": 1.0215,
+      "step": 53125
+    },
+    {
+      "epoch": 15.9,
+      "grad_norm": 3.1977784633636475,
+      "learning_rate": 5.027514028742744e-06,
+      "loss": 1.1078,
+      "step": 53130
+    },
+    {
+      "epoch": 15.9,
+      "grad_norm": 2.141249895095825,
+      "learning_rate": 5.023980848959647e-06,
+      "loss": 1.1468,
+      "step": 53135
+    },
+    {
+      "epoch": 15.9,
+      "grad_norm": 1.349517583847046,
+      "learning_rate": 5.020448772431313e-06,
+      "loss": 0.9661,
+      "step": 53140
+    },
+    {
+      "epoch": 15.9,
+      "grad_norm": 3.260608673095703,
+      "learning_rate": 5.016917799352838e-06,
+      "loss": 0.877,
+      "step": 53145
+    },
+    {
+      "epoch": 15.9,
+      "grad_norm": 3.2229676246643066,
+      "learning_rate": 5.0133879299192085e-06,
+      "loss": 0.9666,
+      "step": 53150
+    },
+    {
+      "epoch": 15.9,
+      "grad_norm": 4.101037979125977,
+      "learning_rate": 5.0098591643254e-06,
+      "loss": 1.0229,
+      "step": 53155
+    },
+    {
+      "epoch": 15.9,
+      "grad_norm": 2.1944096088409424,
+      "learning_rate": 5.006331502766287e-06,
+      "loss": 1.0606,
+      "step": 53160
+    },
+    {
+      "epoch": 15.91,
+      "grad_norm": 7.716952800750732,
+      "learning_rate": 5.002804945436701e-06,
+      "loss": 0.9529,
+      "step": 53165
+    },
+    {
+      "epoch": 15.91,
+      "grad_norm": 17.72467803955078,
+      "learning_rate": 4.999279492531414e-06,
+      "loss": 0.7256,
+      "step": 53170
+    },
+    {
+      "epoch": 15.91,
+      "grad_norm": 3.899148464202881,
+      "learning_rate": 4.995755144245129e-06,
+      "loss": 1.0014,
+      "step": 53175
+    },
+    {
+      "epoch": 15.91,
+      "grad_norm": 5.196731090545654,
+      "learning_rate": 4.992231900772495e-06,
+      "loss": 1.0583,
+      "step": 53180
+    },
+    {
+      "epoch": 15.91,
+      "grad_norm": 4.577493190765381,
+      "learning_rate": 4.9887097623081055e-06,
+      "loss": 0.94,
+      "step": 53185
+    },
+    {
+      "epoch": 15.91,
+      "grad_norm": 3.10968017578125,
+      "learning_rate": 4.985188729046458e-06,
+      "loss": 1.0676,
+      "step": 53190
+    },
+    {
+      "epoch": 15.92,
+      "grad_norm": 1.4285688400268555,
+      "learning_rate": 4.981668801182041e-06,
+      "loss": 1.0163,
+      "step": 53195
+    },
+    {
+      "epoch": 15.92,
+      "grad_norm": 2.724886894226074,
+      "learning_rate": 4.97814997890925e-06,
+      "loss": 0.9586,
+      "step": 53200
+    },
+    {
+      "epoch": 15.92,
+      "grad_norm": 2.3468189239501953,
+      "learning_rate": 4.974632262422421e-06,
+      "loss": 1.0864,
+      "step": 53205
+    },
+    {
+      "epoch": 15.92,
+      "grad_norm": 5.247642517089844,
+      "learning_rate": 4.971115651915839e-06,
+      "loss": 1.1944,
+      "step": 53210
+    },
+    {
+      "epoch": 15.92,
+      "grad_norm": 4.25636100769043,
+      "learning_rate": 4.967600147583718e-06,
+      "loss": 0.9301,
+      "step": 53215
+    },
+    {
+      "epoch": 15.92,
+      "grad_norm": 3.097277879714966,
+      "learning_rate": 4.964085749620228e-06,
+      "loss": 0.9153,
+      "step": 53220
+    },
+    {
+      "epoch": 15.92,
+      "grad_norm": 2.7797505855560303,
+      "learning_rate": 4.960572458219443e-06,
+      "loss": 1.0695,
+      "step": 53225
+    },
+    {
+      "epoch": 15.93,
+      "grad_norm": 3.9096181392669678,
+      "learning_rate": 4.957060273575423e-06,
+      "loss": 0.9084,
+      "step": 53230
+    },
+    {
+      "epoch": 15.93,
+      "grad_norm": 3.273726463317871,
+      "learning_rate": 4.953549195882115e-06,
+      "loss": 0.9885,
+      "step": 53235
+    },
+    {
+      "epoch": 15.93,
+      "grad_norm": 2.705861806869507,
+      "learning_rate": 4.950039225333464e-06,
+      "loss": 0.9194,
+      "step": 53240
+    },
+    {
+      "epoch": 15.93,
+      "grad_norm": 4.52183198928833,
+      "learning_rate": 4.946530362123291e-06,
+      "loss": 0.8444,
+      "step": 53245
+    },
+    {
+      "epoch": 15.93,
+      "grad_norm": 4.226823329925537,
+      "learning_rate": 4.943022606445416e-06,
+      "loss": 0.965,
+      "step": 53250
+    },
+    {
+      "epoch": 15.93,
+      "grad_norm": 2.7318599224090576,
+      "learning_rate": 4.939515958493549e-06,
+      "loss": 0.9715,
+      "step": 53255
+    },
+    {
+      "epoch": 15.93,
+      "grad_norm": 3.3146235942840576,
+      "learning_rate": 4.936010418461362e-06,
+      "loss": 0.8742,
+      "step": 53260
+    },
+    {
+      "epoch": 15.94,
+      "grad_norm": 9.107588768005371,
+      "learning_rate": 4.932505986542465e-06,
+      "loss": 0.8569,
+      "step": 53265
+    },
+    {
+      "epoch": 15.94,
+      "grad_norm": 4.029197692871094,
+      "learning_rate": 4.929002662930404e-06,
+      "loss": 0.9713,
+      "step": 53270
+    },
+    {
+      "epoch": 15.94,
+      "grad_norm": 3.5045926570892334,
+      "learning_rate": 4.9255004478186626e-06,
+      "loss": 0.9804,
+      "step": 53275
+    },
+    {
+      "epoch": 15.94,
+      "grad_norm": 2.6087844371795654,
+      "learning_rate": 4.921999341400666e-06,
+      "loss": 1.0268,
+      "step": 53280
+    },
+    {
+      "epoch": 15.94,
+      "grad_norm": 1.471195936203003,
+      "learning_rate": 4.918499343869773e-06,
+      "loss": 0.8482,
+      "step": 53285
+    },
+    {
+      "epoch": 15.94,
+      "grad_norm": 2.6033759117126465,
+      "learning_rate": 4.915000455419289e-06,
+      "loss": 1.0013,
+      "step": 53290
+    },
+    {
+      "epoch": 15.95,
+      "grad_norm": 2.508314609527588,
+      "learning_rate": 4.911502676242455e-06,
+      "loss": 0.9304,
+      "step": 53295
+    },
+    {
+      "epoch": 15.95,
+      "grad_norm": 3.0417144298553467,
+      "learning_rate": 4.908006006532445e-06,
+      "loss": 0.8015,
+      "step": 53300
+    },
+    {
+      "epoch": 15.95,
+      "grad_norm": 2.4495913982391357,
+      "learning_rate": 4.9045104464823795e-06,
+      "loss": 1.0227,
+      "step": 53305
+    },
+    {
+      "epoch": 15.95,
+      "grad_norm": 5.05440616607666,
+      "learning_rate": 4.901015996285313e-06,
+      "loss": 1.0307,
+      "step": 53310
+    },
+    {
+      "epoch": 15.95,
+      "grad_norm": 1.3113963603973389,
+      "learning_rate": 4.897522656134249e-06,
+      "loss": 1.0566,
+      "step": 53315
+    },
+    {
+      "epoch": 15.95,
+      "grad_norm": 2.597869634628296,
+      "learning_rate": 4.8940304262221015e-06,
+      "loss": 0.8975,
+      "step": 53320
+    },
+    {
+      "epoch": 15.95,
+      "grad_norm": 3.1354362964630127,
+      "learning_rate": 4.890539306741765e-06,
+      "loss": 0.8747,
+      "step": 53325
+    },
+    {
+      "epoch": 15.96,
+      "grad_norm": 3.2860591411590576,
+      "learning_rate": 4.887049297886029e-06,
+      "loss": 0.9559,
+      "step": 53330
+    },
+    {
+      "epoch": 15.96,
+      "grad_norm": 4.5994062423706055,
+      "learning_rate": 4.883560399847664e-06,
+      "loss": 0.9507,
+      "step": 53335
+    },
+    {
+      "epoch": 15.96,
+      "grad_norm": 4.300266742706299,
+      "learning_rate": 4.880072612819336e-06,
+      "loss": 0.9513,
+      "step": 53340
+    },
+    {
+      "epoch": 15.96,
+      "grad_norm": 5.296845436096191,
+      "learning_rate": 4.876585936993699e-06,
+      "loss": 1.0441,
+      "step": 53345
+    },
+    {
+      "epoch": 15.96,
+      "grad_norm": 2.162454843521118,
+      "learning_rate": 4.8731003725632965e-06,
+      "loss": 1.031,
+      "step": 53350
+    },
+    {
+      "epoch": 15.96,
+      "grad_norm": 3.4292469024658203,
+      "learning_rate": 4.869615919720641e-06,
+      "loss": 0.9578,
+      "step": 53355
+    },
+    {
+      "epoch": 15.96,
+      "grad_norm": 1.5867362022399902,
+      "learning_rate": 4.866132578658173e-06,
+      "loss": 1.0536,
+      "step": 53360
+    },
+    {
+      "epoch": 15.97,
+      "grad_norm": 2.8988234996795654,
+      "learning_rate": 4.862650349568274e-06,
+      "loss": 1.0964,
+      "step": 53365
+    },
+    {
+      "epoch": 15.97,
+      "grad_norm": 1.7928255796432495,
+      "learning_rate": 4.859169232643265e-06,
+      "loss": 0.9603,
+      "step": 53370
+    },
+    {
+      "epoch": 15.97,
+      "grad_norm": 2.7317888736724854,
+      "learning_rate": 4.855689228075402e-06,
+      "loss": 0.9083,
+      "step": 53375
+    },
+    {
+      "epoch": 15.97,
+      "grad_norm": 6.080751419067383,
+      "learning_rate": 4.852210336056887e-06,
+      "loss": 0.9522,
+      "step": 53380
+    },
+    {
+      "epoch": 15.97,
+      "grad_norm": 6.8323073387146,
+      "learning_rate": 4.848732556779853e-06,
+      "loss": 1.0334,
+      "step": 53385
+    },
+    {
+      "epoch": 15.97,
+      "grad_norm": 2.109210968017578,
+      "learning_rate": 4.84525589043637e-06,
+      "loss": 0.8857,
+      "step": 53390
+    },
+    {
+      "epoch": 15.98,
+      "grad_norm": 1.5655369758605957,
+      "learning_rate": 4.84178033721846e-06,
+      "loss": 0.8181,
+      "step": 53395
+    },
+    {
+      "epoch": 15.98,
+      "grad_norm": 2.1813764572143555,
+      "learning_rate": 4.838305897318063e-06,
+      "loss": 0.9369,
+      "step": 53400
+    },
+    {
+      "epoch": 15.98,
+      "grad_norm": 6.134399890899658,
+      "learning_rate": 4.8348325709270785e-06,
+      "loss": 0.9357,
+      "step": 53405
+    },
+    {
+      "epoch": 15.98,
+      "grad_norm": 2.2405998706817627,
+      "learning_rate": 4.831360358237336e-06,
+      "loss": 0.9124,
+      "step": 53410
+    },
+    {
+      "epoch": 15.98,
+      "grad_norm": 3.413001775741577,
+      "learning_rate": 4.8278892594405826e-06,
+      "loss": 1.0046,
+      "step": 53415
+    },
+    {
+      "epoch": 15.98,
+      "grad_norm": 2.7362916469573975,
+      "learning_rate": 4.82441927472855e-06,
+      "loss": 0.9897,
+      "step": 53420
+    },
+    {
+      "epoch": 15.98,
+      "grad_norm": 4.3867340087890625,
+      "learning_rate": 4.820950404292859e-06,
+      "loss": 1.0994,
+      "step": 53425
+    },
+    {
+      "epoch": 15.99,
+      "grad_norm": 7.300239086151123,
+      "learning_rate": 4.817482648325114e-06,
+      "loss": 0.9382,
+      "step": 53430
+    },
+    {
+      "epoch": 15.99,
+      "grad_norm": 5.3722124099731445,
+      "learning_rate": 4.814016007016811e-06,
+      "loss": 1.0042,
+      "step": 53435
+    },
+    {
+      "epoch": 15.99,
+      "grad_norm": 2.9729552268981934,
+      "learning_rate": 4.810550480559434e-06,
+      "loss": 0.8934,
+      "step": 53440
+    },
+    {
+      "epoch": 15.99,
+      "grad_norm": 2.5826375484466553,
+      "learning_rate": 4.807086069144364e-06,
+      "loss": 1.0011,
+      "step": 53445
+    },
+    {
+      "epoch": 15.99,
+      "grad_norm": 3.0726816654205322,
+      "learning_rate": 4.80362277296294e-06,
+      "loss": 0.9917,
+      "step": 53450
+    },
+    {
+      "epoch": 15.99,
+      "grad_norm": 2.546142339706421,
+      "learning_rate": 4.800160592206435e-06,
+      "loss": 0.9182,
+      "step": 53455
+    },
+    {
+      "epoch": 15.99,
+      "grad_norm": 1.8220473527908325,
+      "learning_rate": 4.796699527066065e-06,
+      "loss": 0.9448,
+      "step": 53460
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 2.17934250831604,
+      "learning_rate": 4.79323957773298e-06,
+      "loss": 1.1478,
+      "step": 53465
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 2.7372896671295166,
+      "learning_rate": 4.789780744398273e-06,
+      "loss": 0.8963,
+      "step": 53470
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 3.5111589431762695,
+      "learning_rate": 4.786323027252965e-06,
+      "loss": 0.9019,
+      "step": 53475
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 1.6906156539916992,
+      "learning_rate": 4.782866426488025e-06,
+      "loss": 0.9621,
+      "step": 53480
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 2.1876065731048584,
+      "learning_rate": 4.779410942294357e-06,
+      "loss": 1.0843,
+      "step": 53485
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 2.4555323123931885,
+      "learning_rate": 4.7759565748628055e-06,
+      "loss": 0.9055,
+      "step": 53490
+    },
+    {
+      "epoch": 16.01,
+      "grad_norm": 3.3051624298095703,
+      "learning_rate": 4.772503324384151e-06,
+      "loss": 0.7355,
+      "step": 53495
+    },
+    {
+      "epoch": 16.01,
+      "grad_norm": 1.634621500968933,
+      "learning_rate": 4.769051191049112e-06,
+      "loss": 1.1518,
+      "step": 53500
+    },
+    {
+      "epoch": 16.01,
+      "grad_norm": 5.108380317687988,
+      "learning_rate": 4.765600175048355e-06,
+      "loss": 0.7752,
+      "step": 53505
+    },
+    {
+      "epoch": 16.01,
+      "grad_norm": 3.3601999282836914,
+      "learning_rate": 4.762150276572452e-06,
+      "loss": 1.0098,
+      "step": 53510
+    },
+    {
+      "epoch": 16.01,
+      "grad_norm": 1.3551602363586426,
+      "learning_rate": 4.758701495811968e-06,
+      "loss": 0.9517,
+      "step": 53515
+    },
+    {
+      "epoch": 16.01,
+      "grad_norm": 3.776623249053955,
+      "learning_rate": 4.755253832957349e-06,
+      "loss": 0.9535,
+      "step": 53520
+    },
+    {
+      "epoch": 16.01,
+      "grad_norm": 1.849770188331604,
+      "learning_rate": 4.751807288199028e-06,
+      "loss": 1.0225,
+      "step": 53525
+    },
+    {
+      "epoch": 16.02,
+      "grad_norm": 2.972109794616699,
+      "learning_rate": 4.7483618617273325e-06,
+      "loss": 1.0399,
+      "step": 53530
+    },
+    {
+      "epoch": 16.02,
+      "grad_norm": 1.7321369647979736,
+      "learning_rate": 4.744917553732572e-06,
+      "loss": 1.0185,
+      "step": 53535
+    },
+    {
+      "epoch": 16.02,
+      "grad_norm": 2.0730857849121094,
+      "learning_rate": 4.741474364404955e-06,
+      "loss": 1.0239,
+      "step": 53540
+    },
+    {
+      "epoch": 16.02,
+      "grad_norm": 3.4221105575561523,
+      "learning_rate": 4.738032293934652e-06,
+      "loss": 0.8171,
+      "step": 53545
+    },
+    {
+      "epoch": 16.02,
+      "grad_norm": 1.468760371208191,
+      "learning_rate": 4.734591342511765e-06,
+      "loss": 0.9047,
+      "step": 53550
+    },
+    {
+      "epoch": 16.02,
+      "grad_norm": 2.599126100540161,
+      "learning_rate": 4.7311515103263315e-06,
+      "loss": 0.9684,
+      "step": 53555
+    },
+    {
+      "epoch": 16.02,
+      "grad_norm": 2.2176599502563477,
+      "learning_rate": 4.727712797568335e-06,
+      "loss": 0.9353,
+      "step": 53560
+    },
+    {
+      "epoch": 16.03,
+      "grad_norm": 5.144350528717041,
+      "learning_rate": 4.7242752044276865e-06,
+      "loss": 0.9837,
+      "step": 53565
+    },
+    {
+      "epoch": 16.03,
+      "grad_norm": 1.4376980066299438,
+      "learning_rate": 4.720838731094243e-06,
+      "loss": 1.0823,
+      "step": 53570
+    },
+    {
+      "epoch": 16.03,
+      "grad_norm": 3.1391711235046387,
+      "learning_rate": 4.717403377757798e-06,
+      "loss": 0.7762,
+      "step": 53575
+    },
+    {
+      "epoch": 16.03,
+      "grad_norm": 3.4208807945251465,
+      "learning_rate": 4.71396914460808e-06,
+      "loss": 0.8843,
+      "step": 53580
+    },
+    {
+      "epoch": 16.03,
+      "grad_norm": 3.4522769451141357,
+      "learning_rate": 4.710536031834761e-06,
+      "loss": 1.1328,
+      "step": 53585
+    },
+    {
+      "epoch": 16.03,
+      "grad_norm": 1.8125996589660645,
+      "learning_rate": 4.707104039627447e-06,
+      "loss": 0.9399,
+      "step": 53590
+    },
+    {
+      "epoch": 16.04,
+      "grad_norm": 4.709026336669922,
+      "learning_rate": 4.703673168175684e-06,
+      "loss": 0.9658,
+      "step": 53595
+    },
+    {
+      "epoch": 16.04,
+      "grad_norm": 1.6594618558883667,
+      "learning_rate": 4.700243417668957e-06,
+      "loss": 1.03,
+      "step": 53600
+    },
+    {
+      "epoch": 16.04,
+      "grad_norm": 2.7412757873535156,
+      "learning_rate": 4.696814788296683e-06,
+      "loss": 0.945,
+      "step": 53605
+    },
+    {
+      "epoch": 16.04,
+      "grad_norm": 4.177542686462402,
+      "learning_rate": 4.693387280248232e-06,
+      "loss": 0.9507,
+      "step": 53610
+    },
+    {
+      "epoch": 16.04,
+      "grad_norm": 1.866605520248413,
+      "learning_rate": 4.68996089371288e-06,
+      "loss": 0.8753,
+      "step": 53615
+    },
+    {
+      "epoch": 16.04,
+      "grad_norm": 2.461459159851074,
+      "learning_rate": 4.686535628879893e-06,
+      "loss": 0.8427,
+      "step": 53620
+    },
+    {
+      "epoch": 16.04,
+      "grad_norm": 2.5235543251037598,
+      "learning_rate": 4.68311148593841e-06,
+      "loss": 1.0127,
+      "step": 53625
+    },
+    {
+      "epoch": 16.05,
+      "grad_norm": 3.0015721321105957,
+      "learning_rate": 4.679688465077578e-06,
+      "loss": 1.0528,
+      "step": 53630
+    },
+    {
+      "epoch": 16.05,
+      "grad_norm": 1.9430887699127197,
+      "learning_rate": 4.676266566486423e-06,
+      "loss": 0.9704,
+      "step": 53635
+    },
+    {
+      "epoch": 16.05,
+      "grad_norm": 3.479346752166748,
+      "learning_rate": 4.672845790353941e-06,
+      "loss": 0.8397,
+      "step": 53640
+    },
+    {
+      "epoch": 16.05,
+      "grad_norm": 5.18800163269043,
+      "learning_rate": 4.669426136869056e-06,
+      "loss": 0.8035,
+      "step": 53645
+    },
+    {
+      "epoch": 16.05,
+      "grad_norm": 4.256208419799805,
+      "learning_rate": 4.666007606220635e-06,
+      "loss": 0.922,
+      "step": 53650
+    },
+    {
+      "epoch": 16.05,
+      "grad_norm": 3.6765284538269043,
+      "learning_rate": 4.662590198597477e-06,
+      "loss": 0.9406,
+      "step": 53655
+    },
+    {
+      "epoch": 16.05,
+      "grad_norm": 1.5843440294265747,
+      "learning_rate": 4.659173914188322e-06,
+      "loss": 1.2731,
+      "step": 53660
+    },
+    {
+      "epoch": 16.06,
+      "grad_norm": 2.01247501373291,
+      "learning_rate": 4.655758753181846e-06,
+      "loss": 0.95,
+      "step": 53665
+    },
+    {
+      "epoch": 16.06,
+      "grad_norm": 2.1680808067321777,
+      "learning_rate": 4.652344715766671e-06,
+      "loss": 1.0874,
+      "step": 53670
+    },
+    {
+      "epoch": 16.06,
+      "grad_norm": 1.926619291305542,
+      "learning_rate": 4.648931802131348e-06,
+      "loss": 0.935,
+      "step": 53675
+    },
+    {
+      "epoch": 16.06,
+      "grad_norm": 4.823452472686768,
+      "learning_rate": 4.645520012464366e-06,
+      "loss": 0.7357,
+      "step": 53680
+    },
+    {
+      "epoch": 16.06,
+      "grad_norm": 3.968709945678711,
+      "learning_rate": 4.6421093469541545e-06,
+      "loss": 1.1243,
+      "step": 53685
+    },
+    {
+      "epoch": 16.06,
+      "grad_norm": 2.4009251594543457,
+      "learning_rate": 4.638699805789088e-06,
+      "loss": 0.9297,
+      "step": 53690
+    },
+    {
+      "epoch": 16.06,
+      "grad_norm": 2.19075345993042,
+      "learning_rate": 4.635291389157462e-06,
+      "loss": 0.8262,
+      "step": 53695
+    },
+    {
+      "epoch": 16.07,
+      "grad_norm": 5.621512413024902,
+      "learning_rate": 4.631884097247527e-06,
+      "loss": 1.0455,
+      "step": 53700
+    },
+    {
+      "epoch": 16.07,
+      "grad_norm": 1.1728134155273438,
+      "learning_rate": 4.628477930247466e-06,
+      "loss": 0.7797,
+      "step": 53705
+    },
+    {
+      "epoch": 16.07,
+      "grad_norm": 2.9061529636383057,
+      "learning_rate": 4.6250728883453835e-06,
+      "loss": 0.9716,
+      "step": 53710
+    },
+    {
+      "epoch": 16.07,
+      "grad_norm": 2.0948145389556885,
+      "learning_rate": 4.621668971729359e-06,
+      "loss": 0.9648,
+      "step": 53715
+    },
+    {
+      "epoch": 16.07,
+      "grad_norm": 1.2624660730361938,
+      "learning_rate": 4.618266180587363e-06,
+      "loss": 1.1212,
+      "step": 53720
+    },
+    {
+      "epoch": 16.07,
+      "grad_norm": 1.8535141944885254,
+      "learning_rate": 4.61486451510735e-06,
+      "loss": 0.9889,
+      "step": 53725
+    },
+    {
+      "epoch": 16.08,
+      "grad_norm": 2.1382455825805664,
+      "learning_rate": 4.61146397547717e-06,
+      "loss": 0.8759,
+      "step": 53730
+    },
+    {
+      "epoch": 16.08,
+      "grad_norm": 2.7965381145477295,
+      "learning_rate": 4.608064561884656e-06,
+      "loss": 0.8569,
+      "step": 53735
+    },
+    {
+      "epoch": 16.08,
+      "grad_norm": 3.074979066848755,
+      "learning_rate": 4.6046662745175325e-06,
+      "loss": 1.1403,
+      "step": 53740
+    },
+    {
+      "epoch": 16.08,
+      "grad_norm": 2.496306896209717,
+      "learning_rate": 4.601269113563492e-06,
+      "loss": 0.8768,
+      "step": 53745
+    },
+    {
+      "epoch": 16.08,
+      "grad_norm": 1.5231736898422241,
+      "learning_rate": 4.597873079210152e-06,
+      "loss": 1.005,
+      "step": 53750
+    },
+    {
+      "epoch": 16.08,
+      "grad_norm": 3.4601387977600098,
+      "learning_rate": 4.594478171645078e-06,
+      "loss": 1.0404,
+      "step": 53755
+    },
+    {
+      "epoch": 16.08,
+      "grad_norm": 2.981828451156616,
+      "learning_rate": 4.591084391055761e-06,
+      "loss": 1.1838,
+      "step": 53760
+    },
+    {
+      "epoch": 16.09,
+      "grad_norm": 1.5750380754470825,
+      "learning_rate": 4.587691737629643e-06,
+      "loss": 1.0259,
+      "step": 53765
+    },
+    {
+      "epoch": 16.09,
+      "grad_norm": 1.4507620334625244,
+      "learning_rate": 4.58430021155409e-06,
+      "loss": 0.9411,
+      "step": 53770
+    },
+    {
+      "epoch": 16.09,
+      "grad_norm": 1.5799695253372192,
+      "learning_rate": 4.580909813016418e-06,
+      "loss": 1.0206,
+      "step": 53775
+    },
+    {
+      "epoch": 16.09,
+      "grad_norm": 2.9777636528015137,
+      "learning_rate": 4.57752054220387e-06,
+      "loss": 0.9011,
+      "step": 53780
+    },
+    {
+      "epoch": 16.09,
+      "grad_norm": 2.887299060821533,
+      "learning_rate": 4.574132399303638e-06,
+      "loss": 0.9952,
+      "step": 53785
+    },
+    {
+      "epoch": 16.09,
+      "grad_norm": 2.431195020675659,
+      "learning_rate": 4.5707453845028365e-06,
+      "loss": 1.0887,
+      "step": 53790
+    },
+    {
+      "epoch": 16.09,
+      "grad_norm": 5.010396480560303,
+      "learning_rate": 4.567359497988538e-06,
+      "loss": 1.1466,
+      "step": 53795
+    },
+    {
+      "epoch": 16.1,
+      "grad_norm": 4.0482611656188965,
+      "learning_rate": 4.563974739947738e-06,
+      "loss": 0.7066,
+      "step": 53800
+    },
+    {
+      "epoch": 16.1,
+      "grad_norm": 1.8453367948532104,
+      "learning_rate": 4.560591110567361e-06,
+      "loss": 0.9911,
+      "step": 53805
+    },
+    {
+      "epoch": 16.1,
+      "grad_norm": 4.799133777618408,
+      "learning_rate": 4.557208610034302e-06,
+      "loss": 0.9197,
+      "step": 53810
+    },
+    {
+      "epoch": 16.1,
+      "grad_norm": 2.3460564613342285,
+      "learning_rate": 4.553827238535352e-06,
+      "loss": 1.1286,
+      "step": 53815
+    },
+    {
+      "epoch": 16.1,
+      "grad_norm": 2.8893086910247803,
+      "learning_rate": 4.550446996257282e-06,
+      "loss": 0.8731,
+      "step": 53820
+    },
+    {
+      "epoch": 16.1,
+      "grad_norm": 3.3876943588256836,
+      "learning_rate": 4.5470678833867575e-06,
+      "loss": 1.0628,
+      "step": 53825
+    },
+    {
+      "epoch": 16.11,
+      "grad_norm": 3.1112098693847656,
+      "learning_rate": 4.543689900110423e-06,
+      "loss": 0.7802,
+      "step": 53830
+    },
+    {
+      "epoch": 16.11,
+      "grad_norm": 3.280080795288086,
+      "learning_rate": 4.54031304661483e-06,
+      "loss": 0.9001,
+      "step": 53835
+    },
+    {
+      "epoch": 16.11,
+      "grad_norm": 2.949982166290283,
+      "learning_rate": 4.536937323086479e-06,
+      "loss": 0.8499,
+      "step": 53840
+    },
+    {
+      "epoch": 16.11,
+      "grad_norm": 1.0892285108566284,
+      "learning_rate": 4.53356272971181e-06,
+      "loss": 1.1195,
+      "step": 53845
+    },
+    {
+      "epoch": 16.11,
+      "grad_norm": 1.7821409702301025,
+      "learning_rate": 4.53018926667719e-06,
+      "loss": 1.0122,
+      "step": 53850
+    },
+    {
+      "epoch": 16.11,
+      "grad_norm": 3.348202705383301,
+      "learning_rate": 4.526816934168954e-06,
+      "loss": 0.9711,
+      "step": 53855
+    },
+    {
+      "epoch": 16.11,
+      "grad_norm": 3.279318332672119,
+      "learning_rate": 4.52344573237333e-06,
+      "loss": 0.9613,
+      "step": 53860
+    },
+    {
+      "epoch": 16.12,
+      "grad_norm": 1.0132023096084595,
+      "learning_rate": 4.520075661476517e-06,
+      "loss": 0.9816,
+      "step": 53865
+    },
+    {
+      "epoch": 16.12,
+      "grad_norm": 2.115264892578125,
+      "learning_rate": 4.516706721664635e-06,
+      "loss": 1.0883,
+      "step": 53870
+    },
+    {
+      "epoch": 16.12,
+      "grad_norm": 4.187863826751709,
+      "learning_rate": 4.5133389131237495e-06,
+      "loss": 1.0543,
+      "step": 53875
+    },
+    {
+      "epoch": 16.12,
+      "grad_norm": 3.177401065826416,
+      "learning_rate": 4.509972236039861e-06,
+      "loss": 0.8548,
+      "step": 53880
+    },
+    {
+      "epoch": 16.12,
+      "grad_norm": 3.7143056392669678,
+      "learning_rate": 4.50660669059891e-06,
+      "loss": 1.0591,
+      "step": 53885
+    },
+    {
+      "epoch": 16.12,
+      "grad_norm": 2.0598580837249756,
+      "learning_rate": 4.503242276986769e-06,
+      "loss": 1.0228,
+      "step": 53890
+    },
+    {
+      "epoch": 16.12,
+      "grad_norm": 2.2336246967315674,
+      "learning_rate": 4.4998789953892566e-06,
+      "loss": 0.8827,
+      "step": 53895
+    },
+    {
+      "epoch": 16.13,
+      "grad_norm": 2.2673490047454834,
+      "learning_rate": 4.496516845992108e-06,
+      "loss": 1.0858,
+      "step": 53900
+    },
+    {
+      "epoch": 16.13,
+      "grad_norm": 1.8698546886444092,
+      "learning_rate": 4.493155828981033e-06,
+      "loss": 1.0087,
+      "step": 53905
+    },
+    {
+      "epoch": 16.13,
+      "grad_norm": 1.9772906303405762,
+      "learning_rate": 4.489795944541633e-06,
+      "loss": 0.9655,
+      "step": 53910
+    },
+    {
+      "epoch": 16.13,
+      "grad_norm": 1.3239150047302246,
+      "learning_rate": 4.486437192859496e-06,
+      "loss": 0.8522,
+      "step": 53915
+    },
+    {
+      "epoch": 16.13,
+      "grad_norm": 2.097325086593628,
+      "learning_rate": 4.483079574120097e-06,
+      "loss": 1.1089,
+      "step": 53920
+    },
+    {
+      "epoch": 16.13,
+      "grad_norm": 2.80251407623291,
+      "learning_rate": 4.479723088508897e-06,
+      "loss": 0.9511,
+      "step": 53925
+    },
+    {
+      "epoch": 16.14,
+      "grad_norm": 2.2428815364837646,
+      "learning_rate": 4.476367736211265e-06,
+      "loss": 0.7491,
+      "step": 53930
+    },
+    {
+      "epoch": 16.14,
+      "grad_norm": 4.145894527435303,
+      "learning_rate": 4.4730135174124974e-06,
+      "loss": 1.0072,
+      "step": 53935
+    },
+    {
+      "epoch": 16.14,
+      "grad_norm": 1.747147560119629,
+      "learning_rate": 4.469660432297868e-06,
+      "loss": 0.9186,
+      "step": 53940
+    },
+    {
+      "epoch": 16.14,
+      "grad_norm": 2.3156204223632812,
+      "learning_rate": 4.466308481052542e-06,
+      "loss": 1.0476,
+      "step": 53945
+    },
+    {
+      "epoch": 16.14,
+      "grad_norm": 1.5770155191421509,
+      "learning_rate": 4.4636277365666055e-06,
+      "loss": 1.0194,
+      "step": 53950
+    },
+    {
+      "epoch": 16.14,
+      "grad_norm": 5.032956123352051,
+      "learning_rate": 4.460277826752518e-06,
+      "loss": 0.9442,
+      "step": 53955
+    },
+    {
+      "epoch": 16.14,
+      "grad_norm": 2.544370651245117,
+      "learning_rate": 4.456929051325948e-06,
+      "loss": 0.9561,
+      "step": 53960
+    },
+    {
+      "epoch": 16.15,
+      "grad_norm": 2.7542896270751953,
+      "learning_rate": 4.453581410471821e-06,
+      "loss": 0.9285,
+      "step": 53965
+    },
+    {
+      "epoch": 16.15,
+      "grad_norm": 4.545336723327637,
+      "learning_rate": 4.450234904375047e-06,
+      "loss": 0.8249,
+      "step": 53970
+    },
+    {
+      "epoch": 16.15,
+      "grad_norm": 6.846313953399658,
+      "learning_rate": 4.446889533220427e-06,
+      "loss": 0.9482,
+      "step": 53975
+    },
+    {
+      "epoch": 16.15,
+      "grad_norm": 2.253056764602661,
+      "learning_rate": 4.443545297192747e-06,
+      "loss": 1.0561,
+      "step": 53980
+    },
+    {
+      "epoch": 16.15,
+      "grad_norm": 3.439204216003418,
+      "learning_rate": 4.440202196476687e-06,
+      "loss": 0.8451,
+      "step": 53985
+    },
+    {
+      "epoch": 16.15,
+      "grad_norm": 2.593958854675293,
+      "learning_rate": 4.43686023125689e-06,
+      "loss": 0.8443,
+      "step": 53990
+    },
+    {
+      "epoch": 16.15,
+      "grad_norm": 3.0037314891815186,
+      "learning_rate": 4.433519401717925e-06,
+      "loss": 0.8752,
+      "step": 53995
+    },
+    {
+      "epoch": 16.16,
+      "grad_norm": 1.4025530815124512,
+      "learning_rate": 4.43017970804431e-06,
+      "loss": 0.8974,
+      "step": 54000
+    },
+    {
+      "epoch": 16.16,
+      "grad_norm": 3.922529935836792,
+      "learning_rate": 4.426841150420485e-06,
+      "loss": 1.1543,
+      "step": 54005
+    },
+    {
+      "epoch": 16.16,
+      "grad_norm": 2.439948797225952,
+      "learning_rate": 4.4235037290308425e-06,
+      "loss": 0.8726,
+      "step": 54010
+    },
+    {
+      "epoch": 16.16,
+      "grad_norm": 2.4692089557647705,
+      "learning_rate": 4.420167444059698e-06,
+      "loss": 0.8118,
+      "step": 54015
+    },
+    {
+      "epoch": 16.16,
+      "grad_norm": 2.5722415447235107,
+      "learning_rate": 4.416832295691314e-06,
+      "loss": 1.0299,
+      "step": 54020
+    },
+    {
+      "epoch": 16.16,
+      "grad_norm": 2.7503256797790527,
+      "learning_rate": 4.413498284109888e-06,
+      "loss": 1.172,
+      "step": 54025
+    },
+    {
+      "epoch": 16.17,
+      "grad_norm": 3.138927459716797,
+      "learning_rate": 4.410165409499553e-06,
+      "loss": 0.941,
+      "step": 54030
+    },
+    {
+      "epoch": 16.17,
+      "grad_norm": 2.402808666229248,
+      "learning_rate": 4.40683367204438e-06,
+      "loss": 0.9587,
+      "step": 54035
+    },
+    {
+      "epoch": 16.17,
+      "grad_norm": 3.9878361225128174,
+      "learning_rate": 4.403503071928378e-06,
+      "loss": 0.8567,
+      "step": 54040
+    },
+    {
+      "epoch": 16.17,
+      "grad_norm": 1.6606813669204712,
+      "learning_rate": 4.4001736093355005e-06,
+      "loss": 1.0092,
+      "step": 54045
+    },
+    {
+      "epoch": 16.17,
+      "grad_norm": 3.9329001903533936,
+      "learning_rate": 4.396845284449608e-06,
+      "loss": 1.008,
+      "step": 54050
+    },
+    {
+      "epoch": 16.17,
+      "grad_norm": 5.513614654541016,
+      "learning_rate": 4.393518097454546e-06,
+      "loss": 1.0288,
+      "step": 54055
+    },
+    {
+      "epoch": 16.17,
+      "grad_norm": 1.570499062538147,
+      "learning_rate": 4.39019204853405e-06,
+      "loss": 1.0363,
+      "step": 54060
+    },
+    {
+      "epoch": 16.18,
+      "grad_norm": 2.9202938079833984,
+      "learning_rate": 4.386867137871839e-06,
+      "loss": 0.8739,
+      "step": 54065
+    },
+    {
+      "epoch": 16.18,
+      "grad_norm": 1.5912072658538818,
+      "learning_rate": 4.383543365651513e-06,
+      "loss": 0.9977,
+      "step": 54070
+    },
+    {
+      "epoch": 16.18,
+      "grad_norm": 1.2433182001113892,
+      "learning_rate": 4.380220732056673e-06,
+      "loss": 0.9558,
+      "step": 54075
+    },
+    {
+      "epoch": 16.18,
+      "grad_norm": 3.465158700942993,
+      "learning_rate": 4.376899237270798e-06,
+      "loss": 0.9604,
+      "step": 54080
+    },
+    {
+      "epoch": 16.18,
+      "grad_norm": 4.610307216644287,
+      "learning_rate": 4.373578881477347e-06,
+      "loss": 0.9055,
+      "step": 54085
+    },
+    {
+      "epoch": 16.18,
+      "grad_norm": 1.9522101879119873,
+      "learning_rate": 4.370259664859691e-06,
+      "loss": 0.9228,
+      "step": 54090
+    },
+    {
+      "epoch": 16.18,
+      "grad_norm": 1.3069242238998413,
+      "learning_rate": 4.3669415876011535e-06,
+      "loss": 1.0042,
+      "step": 54095
+    },
+    {
+      "epoch": 16.19,
+      "grad_norm": 2.4760279655456543,
+      "learning_rate": 4.363624649884982e-06,
+      "loss": 1.0021,
+      "step": 54100
+    },
+    {
+      "epoch": 16.19,
+      "grad_norm": 1.8461346626281738,
+      "learning_rate": 4.360308851894366e-06,
+      "loss": 1.0572,
+      "step": 54105
+    },
+    {
+      "epoch": 16.19,
+      "grad_norm": 3.235544443130493,
+      "learning_rate": 4.35699419381245e-06,
+      "loss": 0.9269,
+      "step": 54110
+    },
+    {
+      "epoch": 16.19,
+      "grad_norm": 2.0938470363616943,
+      "learning_rate": 4.353680675822281e-06,
+      "loss": 0.9334,
+      "step": 54115
+    },
+    {
+      "epoch": 16.19,
+      "grad_norm": 1.0838735103607178,
+      "learning_rate": 4.350368298106869e-06,
+      "loss": 1.0252,
+      "step": 54120
+    },
+    {
+      "epoch": 16.19,
+      "grad_norm": 4.055896282196045,
+      "learning_rate": 4.34705706084915e-06,
+      "loss": 0.9531,
+      "step": 54125
+    },
+    {
+      "epoch": 16.2,
+      "grad_norm": 1.868022084236145,
+      "learning_rate": 4.343746964232004e-06,
+      "loss": 1.0052,
+      "step": 54130
+    },
+    {
+      "epoch": 16.2,
+      "grad_norm": 2.870447874069214,
+      "learning_rate": 4.340438008438241e-06,
+      "loss": 1.0222,
+      "step": 54135
+    },
+    {
+      "epoch": 16.2,
+      "grad_norm": 3.2282488346099854,
+      "learning_rate": 4.3371301936506205e-06,
+      "loss": 1.0202,
+      "step": 54140
+    },
+    {
+      "epoch": 16.2,
+      "grad_norm": 2.0114264488220215,
+      "learning_rate": 4.333823520051805e-06,
+      "loss": 1.038,
+      "step": 54145
+    },
+    {
+      "epoch": 16.2,
+      "grad_norm": 1.6153552532196045,
+      "learning_rate": 4.330517987824451e-06,
+      "loss": 1.0861,
+      "step": 54150
+    },
+    {
+      "epoch": 16.2,
+      "grad_norm": 4.407323360443115,
+      "learning_rate": 4.327213597151092e-06,
+      "loss": 1.1782,
+      "step": 54155
+    },
+    {
+      "epoch": 16.2,
+      "grad_norm": 2.3100712299346924,
+      "learning_rate": 4.323910348214247e-06,
+      "loss": 0.7662,
+      "step": 54160
+    },
+    {
+      "epoch": 16.21,
+      "grad_norm": 2.331082582473755,
+      "learning_rate": 4.320608241196333e-06,
+      "loss": 0.9865,
+      "step": 54165
+    },
+    {
+      "epoch": 16.21,
+      "grad_norm": 2.015458822250366,
+      "learning_rate": 4.317307276279742e-06,
+      "loss": 1.0186,
+      "step": 54170
+    },
+    {
+      "epoch": 16.21,
+      "grad_norm": 2.7874059677124023,
+      "learning_rate": 4.3140074536467625e-06,
+      "loss": 1.0223,
+      "step": 54175
+    },
+    {
+      "epoch": 16.21,
+      "grad_norm": 1.4887959957122803,
+      "learning_rate": 4.310708773479652e-06,
+      "loss": 1.0826,
+      "step": 54180
+    },
+    {
+      "epoch": 16.21,
+      "grad_norm": 3.373190402984619,
+      "learning_rate": 4.307411235960593e-06,
+      "loss": 1.1612,
+      "step": 54185
+    },
+    {
+      "epoch": 16.21,
+      "grad_norm": 5.217267990112305,
+      "learning_rate": 4.304114841271692e-06,
+      "loss": 1.0709,
+      "step": 54190
+    },
+    {
+      "epoch": 16.21,
+      "grad_norm": 1.9477370977401733,
+      "learning_rate": 4.3008195895950315e-06,
+      "loss": 0.9157,
+      "step": 54195
+    },
+    {
+      "epoch": 16.22,
+      "grad_norm": 3.2760956287384033,
+      "learning_rate": 4.297525481112577e-06,
+      "loss": 0.9145,
+      "step": 54200
+    },
+    {
+      "epoch": 16.22,
+      "grad_norm": 4.605449199676514,
+      "learning_rate": 4.2942325160062826e-06,
+      "loss": 0.9861,
+      "step": 54205
+    },
+    {
+      "epoch": 16.22,
+      "grad_norm": 1.7385135889053345,
+      "learning_rate": 4.290940694457995e-06,
+      "loss": 1.0439,
+      "step": 54210
+    },
+    {
+      "epoch": 16.22,
+      "grad_norm": 2.1554768085479736,
+      "learning_rate": 4.287650016649533e-06,
+      "loss": 1.2434,
+      "step": 54215
+    },
+    {
+      "epoch": 16.22,
+      "grad_norm": 3.2271509170532227,
+      "learning_rate": 4.284360482762626e-06,
+      "loss": 0.9879,
+      "step": 54220
+    },
+    {
+      "epoch": 16.22,
+      "grad_norm": 2.956493616104126,
+      "learning_rate": 4.28107209297896e-06,
+      "loss": 1.2336,
+      "step": 54225
+    },
+    {
+      "epoch": 16.22,
+      "grad_norm": 2.8822336196899414,
+      "learning_rate": 4.277784847480145e-06,
+      "loss": 0.9283,
+      "step": 54230
+    },
+    {
+      "epoch": 16.23,
+      "grad_norm": 3.9838624000549316,
+      "learning_rate": 4.274498746447739e-06,
+      "loss": 0.999,
+      "step": 54235
+    },
+    {
+      "epoch": 16.23,
+      "grad_norm": 1.797528862953186,
+      "learning_rate": 4.271213790063213e-06,
+      "loss": 0.9987,
+      "step": 54240
+    },
+    {
+      "epoch": 16.23,
+      "grad_norm": 8.245445251464844,
+      "learning_rate": 4.267929978508014e-06,
+      "loss": 0.8547,
+      "step": 54245
+    },
+    {
+      "epoch": 16.23,
+      "grad_norm": 1.7248831987380981,
+      "learning_rate": 4.264647311963482e-06,
+      "loss": 1.2987,
+      "step": 54250
+    },
+    {
+      "epoch": 16.23,
+      "grad_norm": 1.602921485900879,
+      "learning_rate": 4.261365790610935e-06,
+      "loss": 1.0701,
+      "step": 54255
+    },
+    {
+      "epoch": 16.23,
+      "grad_norm": 1.774390459060669,
+      "learning_rate": 4.258085414631591e-06,
+      "loss": 0.9237,
+      "step": 54260
+    },
+    {
+      "epoch": 16.24,
+      "grad_norm": 2.390256881713867,
+      "learning_rate": 4.254806184206634e-06,
+      "loss": 1.0289,
+      "step": 54265
+    },
+    {
+      "epoch": 16.24,
+      "grad_norm": 2.603825330734253,
+      "learning_rate": 4.2515280995171715e-06,
+      "loss": 0.9717,
+      "step": 54270
+    },
+    {
+      "epoch": 16.24,
+      "grad_norm": 1.769459843635559,
+      "learning_rate": 4.248251160744235e-06,
+      "loss": 1.237,
+      "step": 54275
+    },
+    {
+      "epoch": 16.24,
+      "grad_norm": 2.9963274002075195,
+      "learning_rate": 4.2449753680688295e-06,
+      "loss": 0.7686,
+      "step": 54280
+    },
+    {
+      "epoch": 16.24,
+      "grad_norm": 1.8369311094284058,
+      "learning_rate": 4.241700721671849e-06,
+      "loss": 1.0841,
+      "step": 54285
+    },
+    {
+      "epoch": 16.24,
+      "grad_norm": 1.0869977474212646,
+      "learning_rate": 4.23842722173417e-06,
+      "loss": 0.7985,
+      "step": 54290
+    },
+    {
+      "epoch": 16.24,
+      "grad_norm": 2.742431402206421,
+      "learning_rate": 4.235154868436564e-06,
+      "loss": 1.0202,
+      "step": 54295
+    },
+    {
+      "epoch": 16.25,
+      "grad_norm": 2.4266834259033203,
+      "learning_rate": 4.231883661959785e-06,
+      "loss": 0.8518,
+      "step": 54300
+    },
+    {
+      "epoch": 16.25,
+      "grad_norm": 3.0461745262145996,
+      "learning_rate": 4.228613602484477e-06,
+      "loss": 0.8847,
+      "step": 54305
+    },
+    {
+      "epoch": 16.25,
+      "grad_norm": 2.476036787033081,
+      "learning_rate": 4.225344690191247e-06,
+      "loss": 0.8996,
+      "step": 54310
+    },
+    {
+      "epoch": 16.25,
+      "grad_norm": 3.7024502754211426,
+      "learning_rate": 4.222076925260637e-06,
+      "loss": 0.9061,
+      "step": 54315
+    },
+    {
+      "epoch": 16.25,
+      "grad_norm": 3.387453556060791,
+      "learning_rate": 4.218810307873122e-06,
+      "loss": 1.0476,
+      "step": 54320
+    },
+    {
+      "epoch": 16.25,
+      "grad_norm": 4.558333396911621,
+      "learning_rate": 4.215544838209112e-06,
+      "loss": 0.797,
+      "step": 54325
+    },
+    {
+      "epoch": 16.25,
+      "grad_norm": 3.4511423110961914,
+      "learning_rate": 4.2122805164489575e-06,
+      "loss": 0.7751,
+      "step": 54330
+    },
+    {
+      "epoch": 16.26,
+      "grad_norm": 2.4084949493408203,
+      "learning_rate": 4.209017342772939e-06,
+      "loss": 0.8006,
+      "step": 54335
+    },
+    {
+      "epoch": 16.26,
+      "grad_norm": 3.032461643218994,
+      "learning_rate": 4.205755317361293e-06,
+      "loss": 0.9037,
+      "step": 54340
+    },
+    {
+      "epoch": 16.26,
+      "grad_norm": 1.9977115392684937,
+      "learning_rate": 4.20249444039415e-06,
+      "loss": 0.9293,
+      "step": 54345
+    },
+    {
+      "epoch": 16.26,
+      "grad_norm": 1.7754746675491333,
+      "learning_rate": 4.199234712051628e-06,
+      "loss": 1.0857,
+      "step": 54350
+    },
+    {
+      "epoch": 16.26,
+      "grad_norm": 3.022590398788452,
+      "learning_rate": 4.195976132513754e-06,
+      "loss": 0.8392,
+      "step": 54355
+    },
+    {
+      "epoch": 16.26,
+      "grad_norm": 3.174025535583496,
+      "learning_rate": 4.192718701960491e-06,
+      "loss": 0.9006,
+      "step": 54360
+    },
+    {
+      "epoch": 16.27,
+      "grad_norm": 12.422464370727539,
+      "learning_rate": 4.189462420571752e-06,
+      "loss": 1.0187,
+      "step": 54365
+    },
+    {
+      "epoch": 16.27,
+      "grad_norm": 1.5121108293533325,
+      "learning_rate": 4.186207288527361e-06,
+      "loss": 0.9843,
+      "step": 54370
+    },
+    {
+      "epoch": 16.27,
+      "grad_norm": 2.849782705307007,
+      "learning_rate": 4.18295330600712e-06,
+      "loss": 1.0467,
+      "step": 54375
+    },
+    {
+      "epoch": 16.27,
+      "grad_norm": 3.5961849689483643,
+      "learning_rate": 4.179700473190717e-06,
+      "loss": 0.9365,
+      "step": 54380
+    },
+    {
+      "epoch": 16.27,
+      "grad_norm": 2.3234174251556396,
+      "learning_rate": 4.176448790257828e-06,
+      "loss": 0.8933,
+      "step": 54385
+    },
+    {
+      "epoch": 16.27,
+      "grad_norm": 1.903601050376892,
+      "learning_rate": 4.173198257388014e-06,
+      "loss": 0.8867,
+      "step": 54390
+    },
+    {
+      "epoch": 16.27,
+      "grad_norm": 5.895204067230225,
+      "learning_rate": 4.169948874760823e-06,
+      "loss": 1.0914,
+      "step": 54395
+    },
+    {
+      "epoch": 16.28,
+      "grad_norm": 2.9874160289764404,
+      "learning_rate": 4.166700642555702e-06,
+      "loss": 1.0449,
+      "step": 54400
+    },
+    {
+      "epoch": 16.28,
+      "grad_norm": 2.241297721862793,
+      "learning_rate": 4.1634535609520465e-06,
+      "loss": 1.0472,
+      "step": 54405
+    },
+    {
+      "epoch": 16.28,
+      "grad_norm": 3.7421207427978516,
+      "learning_rate": 4.160207630129192e-06,
+      "loss": 1.1745,
+      "step": 54410
+    },
+    {
+      "epoch": 16.28,
+      "grad_norm": 4.566219329833984,
+      "learning_rate": 4.15696285026641e-06,
+      "loss": 1.0118,
+      "step": 54415
+    },
+    {
+      "epoch": 16.28,
+      "grad_norm": 1.1942859888076782,
+      "learning_rate": 4.153719221542904e-06,
+      "loss": 1.0707,
+      "step": 54420
+    },
+    {
+      "epoch": 16.28,
+      "grad_norm": 3.381263494491577,
+      "learning_rate": 4.150476744137818e-06,
+      "loss": 0.9244,
+      "step": 54425
+    },
+    {
+      "epoch": 16.28,
+      "grad_norm": 2.5632987022399902,
+      "learning_rate": 4.147235418230227e-06,
+      "loss": 0.9353,
+      "step": 54430
+    },
+    {
+      "epoch": 16.29,
+      "grad_norm": 4.676177978515625,
+      "learning_rate": 4.143995243999152e-06,
+      "loss": 1.0719,
+      "step": 54435
+    },
+    {
+      "epoch": 16.29,
+      "grad_norm": 3.268036365509033,
+      "learning_rate": 4.140756221623537e-06,
+      "loss": 0.7861,
+      "step": 54440
+    },
+    {
+      "epoch": 16.29,
+      "grad_norm": 1.5229823589324951,
+      "learning_rate": 4.137518351282277e-06,
+      "loss": 0.874,
+      "step": 54445
+    },
+    {
+      "epoch": 16.29,
+      "grad_norm": 4.277475833892822,
+      "learning_rate": 4.1342816331541915e-06,
+      "loss": 0.9358,
+      "step": 54450
+    },
+    {
+      "epoch": 16.29,
+      "grad_norm": 3.253185510635376,
+      "learning_rate": 4.131046067418043e-06,
+      "loss": 1.0229,
+      "step": 54455
+    },
+    {
+      "epoch": 16.29,
+      "grad_norm": 1.8826074600219727,
+      "learning_rate": 4.127811654252531e-06,
+      "loss": 0.8511,
+      "step": 54460
+    },
+    {
+      "epoch": 16.3,
+      "grad_norm": 4.963164806365967,
+      "learning_rate": 4.124578393836282e-06,
+      "loss": 1.0294,
+      "step": 54465
+    },
+    {
+      "epoch": 16.3,
+      "grad_norm": 4.152764320373535,
+      "learning_rate": 4.121346286347877e-06,
+      "loss": 1.0621,
+      "step": 54470
+    },
+    {
+      "epoch": 16.3,
+      "grad_norm": 4.597795009613037,
+      "learning_rate": 4.1181153319658025e-06,
+      "loss": 0.8713,
+      "step": 54475
+    },
+    {
+      "epoch": 16.3,
+      "grad_norm": 2.7429773807525635,
+      "learning_rate": 4.1148855308685256e-06,
+      "loss": 0.9384,
+      "step": 54480
+    },
+    {
+      "epoch": 16.3,
+      "grad_norm": 3.036653518676758,
+      "learning_rate": 4.111656883234396e-06,
+      "loss": 1.026,
+      "step": 54485
+    },
+    {
+      "epoch": 16.3,
+      "grad_norm": 3.560821533203125,
+      "learning_rate": 4.108429389241761e-06,
+      "loss": 0.9886,
+      "step": 54490
+    },
+    {
+      "epoch": 16.3,
+      "grad_norm": 3.0521786212921143,
+      "learning_rate": 4.105203049068848e-06,
+      "loss": 1.0105,
+      "step": 54495
+    },
+    {
+      "epoch": 16.31,
+      "grad_norm": 2.9421958923339844,
+      "learning_rate": 4.101977862893852e-06,
+      "loss": 0.8541,
+      "step": 54500
+    },
+    {
+      "epoch": 16.31,
+      "grad_norm": 2.5392541885375977,
+      "learning_rate": 4.098753830894894e-06,
+      "loss": 1.1314,
+      "step": 54505
+    },
+    {
+      "epoch": 16.31,
+      "grad_norm": 2.7192037105560303,
+      "learning_rate": 4.09553095325004e-06,
+      "loss": 1.0789,
+      "step": 54510
+    },
+    {
+      "epoch": 16.31,
+      "grad_norm": 1.7262064218521118,
+      "learning_rate": 4.092309230137282e-06,
+      "loss": 0.901,
+      "step": 54515
+    },
+    {
+      "epoch": 16.31,
+      "grad_norm": 6.73134708404541,
+      "learning_rate": 4.08908866173455e-06,
+      "loss": 0.8987,
+      "step": 54520
+    },
+    {
+      "epoch": 16.31,
+      "grad_norm": 2.3271663188934326,
+      "learning_rate": 4.085869248219715e-06,
+      "loss": 1.0085,
+      "step": 54525
+    },
+    {
+      "epoch": 16.31,
+      "grad_norm": 4.21439266204834,
+      "learning_rate": 4.082650989770584e-06,
+      "loss": 0.9719,
+      "step": 54530
+    },
+    {
+      "epoch": 16.32,
+      "grad_norm": 3.639162540435791,
+      "learning_rate": 4.079433886564899e-06,
+      "loss": 0.9036,
+      "step": 54535
+    },
+    {
+      "epoch": 16.32,
+      "grad_norm": 2.1096203327178955,
+      "learning_rate": 4.0762179387803305e-06,
+      "loss": 1.0381,
+      "step": 54540
+    },
+    {
+      "epoch": 16.32,
+      "grad_norm": 4.696827411651611,
+      "learning_rate": 4.073003146594498e-06,
+      "loss": 0.8693,
+      "step": 54545
+    },
+    {
+      "epoch": 16.32,
+      "grad_norm": 4.135863304138184,
+      "learning_rate": 4.069789510184951e-06,
+      "loss": 0.7749,
+      "step": 54550
+    },
+    {
+      "epoch": 16.32,
+      "grad_norm": 2.406681537628174,
+      "learning_rate": 4.0665770297291714e-06,
+      "loss": 1.0704,
+      "step": 54555
+    },
+    {
+      "epoch": 16.32,
+      "grad_norm": 2.2193562984466553,
+      "learning_rate": 4.063365705404584e-06,
+      "loss": 1.1637,
+      "step": 54560
+    },
+    {
+      "epoch": 16.33,
+      "grad_norm": 2.6856563091278076,
+      "learning_rate": 4.060155537388552e-06,
+      "loss": 1.0761,
+      "step": 54565
+    },
+    {
+      "epoch": 16.33,
+      "grad_norm": 5.369014263153076,
+      "learning_rate": 4.056946525858352e-06,
+      "loss": 1.0351,
+      "step": 54570
+    },
+    {
+      "epoch": 16.33,
+      "grad_norm": 2.9913430213928223,
+      "learning_rate": 4.053738670991239e-06,
+      "loss": 1.006,
+      "step": 54575
+    },
+    {
+      "epoch": 16.33,
+      "grad_norm": 1.8716338872909546,
+      "learning_rate": 4.0505319729643554e-06,
+      "loss": 0.9395,
+      "step": 54580
+    },
+    {
+      "epoch": 16.33,
+      "grad_norm": 1.9372947216033936,
+      "learning_rate": 4.0473264319548256e-06,
+      "loss": 0.8675,
+      "step": 54585
+    },
+    {
+      "epoch": 16.33,
+      "grad_norm": 2.8180017471313477,
+      "learning_rate": 4.044122048139673e-06,
+      "loss": 0.9045,
+      "step": 54590
+    },
+    {
+      "epoch": 16.33,
+      "grad_norm": 2.123150110244751,
+      "learning_rate": 4.0409188216958775e-06,
+      "loss": 0.8954,
+      "step": 54595
+    },
+    {
+      "epoch": 16.34,
+      "grad_norm": 1.7818152904510498,
+      "learning_rate": 4.037716752800347e-06,
+      "loss": 1.0991,
+      "step": 54600
+    },
+    {
+      "epoch": 16.34,
+      "grad_norm": 1.8431493043899536,
+      "learning_rate": 4.034515841629932e-06,
+      "loss": 1.1051,
+      "step": 54605
+    },
+    {
+      "epoch": 16.34,
+      "grad_norm": 1.9420349597930908,
+      "learning_rate": 4.031316088361417e-06,
+      "loss": 0.7161,
+      "step": 54610
+    },
+    {
+      "epoch": 16.34,
+      "grad_norm": 2.2753307819366455,
+      "learning_rate": 4.0281174931715185e-06,
+      "loss": 0.8588,
+      "step": 54615
+    },
+    {
+      "epoch": 16.34,
+      "grad_norm": 1.6886123418807983,
+      "learning_rate": 4.024920056236889e-06,
+      "loss": 0.9475,
+      "step": 54620
+    },
+    {
+      "epoch": 16.34,
+      "grad_norm": 2.501573324203491,
+      "learning_rate": 4.021723777734124e-06,
+      "loss": 0.935,
+      "step": 54625
+    },
+    {
+      "epoch": 16.34,
+      "grad_norm": 2.7062265872955322,
+      "learning_rate": 4.018528657839752e-06,
+      "loss": 1.1356,
+      "step": 54630
+    },
+    {
+      "epoch": 16.35,
+      "grad_norm": 1.7512439489364624,
+      "learning_rate": 4.015334696730227e-06,
+      "loss": 1.0477,
+      "step": 54635
+    },
+    {
+      "epoch": 16.35,
+      "grad_norm": 3.424471139907837,
+      "learning_rate": 4.012141894581958e-06,
+      "loss": 0.9916,
+      "step": 54640
+    },
+    {
+      "epoch": 16.35,
+      "grad_norm": 2.5888991355895996,
+      "learning_rate": 4.008950251571278e-06,
+      "loss": 0.8972,
+      "step": 54645
+    },
+    {
+      "epoch": 16.35,
+      "grad_norm": 2.482107639312744,
+      "learning_rate": 4.005759767874453e-06,
+      "loss": 0.9857,
+      "step": 54650
+    },
+    {
+      "epoch": 16.35,
+      "grad_norm": 4.3171796798706055,
+      "learning_rate": 4.002570443667697e-06,
+      "loss": 1.0612,
+      "step": 54655
+    },
+    {
+      "epoch": 16.35,
+      "grad_norm": 2.3370325565338135,
+      "learning_rate": 3.999382279127153e-06,
+      "loss": 1.032,
+      "step": 54660
+    },
+    {
+      "epoch": 16.36,
+      "grad_norm": 2.1718616485595703,
+      "learning_rate": 3.996195274428888e-06,
+      "loss": 1.2058,
+      "step": 54665
+    },
+    {
+      "epoch": 16.36,
+      "grad_norm": 4.225613594055176,
+      "learning_rate": 3.993009429748937e-06,
+      "loss": 0.8797,
+      "step": 54670
+    },
+    {
+      "epoch": 16.36,
+      "grad_norm": 3.9194295406341553,
+      "learning_rate": 3.989824745263224e-06,
+      "loss": 0.8541,
+      "step": 54675
+    },
+    {
+      "epoch": 16.36,
+      "grad_norm": 3.645948886871338,
+      "learning_rate": 3.986641221147666e-06,
+      "loss": 1.0261,
+      "step": 54680
+    },
+    {
+      "epoch": 16.36,
+      "grad_norm": 1.8858450651168823,
+      "learning_rate": 3.9834588575780595e-06,
+      "loss": 1.023,
+      "step": 54685
+    },
+    {
+      "epoch": 16.36,
+      "grad_norm": 2.733396291732788,
+      "learning_rate": 3.9802776547301884e-06,
+      "loss": 1.2168,
+      "step": 54690
+    },
+    {
+      "epoch": 16.36,
+      "grad_norm": 1.9093108177185059,
+      "learning_rate": 3.977097612779723e-06,
+      "loss": 0.9631,
+      "step": 54695
+    },
+    {
+      "epoch": 16.37,
+      "grad_norm": 3.2848117351531982,
+      "learning_rate": 3.9739187319023066e-06,
+      "loss": 0.972,
+      "step": 54700
+    },
+    {
+      "epoch": 16.37,
+      "grad_norm": 1.4223105907440186,
+      "learning_rate": 3.970741012273507e-06,
+      "loss": 0.9396,
+      "step": 54705
+    },
+    {
+      "epoch": 16.37,
+      "grad_norm": 1.8752068281173706,
+      "learning_rate": 3.967564454068817e-06,
+      "loss": 0.8828,
+      "step": 54710
+    },
+    {
+      "epoch": 16.37,
+      "grad_norm": 2.8616607189178467,
+      "learning_rate": 3.964389057463683e-06,
+      "loss": 1.0098,
+      "step": 54715
+    },
+    {
+      "epoch": 16.37,
+      "grad_norm": 2.8053674697875977,
+      "learning_rate": 3.961214822633477e-06,
+      "loss": 0.9985,
+      "step": 54720
+    },
+    {
+      "epoch": 16.37,
+      "grad_norm": 3.857592821121216,
+      "learning_rate": 3.958041749753505e-06,
+      "loss": 0.9801,
+      "step": 54725
+    },
+    {
+      "epoch": 16.37,
+      "grad_norm": 1.627759337425232,
+      "learning_rate": 3.954869838999017e-06,
+      "loss": 0.9103,
+      "step": 54730
+    },
+    {
+      "epoch": 16.38,
+      "grad_norm": 3.5982964038848877,
+      "learning_rate": 3.951699090545194e-06,
+      "loss": 0.7888,
+      "step": 54735
+    },
+    {
+      "epoch": 16.38,
+      "grad_norm": 3.797572135925293,
+      "learning_rate": 3.948529504567147e-06,
+      "loss": 0.9846,
+      "step": 54740
+    },
+    {
+      "epoch": 16.38,
+      "grad_norm": 2.0291848182678223,
+      "learning_rate": 3.9453610812399375e-06,
+      "loss": 0.8557,
+      "step": 54745
+    },
+    {
+      "epoch": 16.38,
+      "grad_norm": 1.8980909585952759,
+      "learning_rate": 3.9421938207385485e-06,
+      "loss": 1.0815,
+      "step": 54750
+    },
+    {
+      "epoch": 16.38,
+      "grad_norm": 3.711595058441162,
+      "learning_rate": 3.939027723237914e-06,
+      "loss": 0.9014,
+      "step": 54755
+    },
+    {
+      "epoch": 16.38,
+      "grad_norm": 1.4966007471084595,
+      "learning_rate": 3.935862788912875e-06,
+      "loss": 0.9629,
+      "step": 54760
+    },
+    {
+      "epoch": 16.39,
+      "grad_norm": 4.108902931213379,
+      "learning_rate": 3.932699017938251e-06,
+      "loss": 1.071,
+      "step": 54765
+    },
+    {
+      "epoch": 16.39,
+      "grad_norm": 2.2687480449676514,
+      "learning_rate": 3.929536410488749e-06,
+      "loss": 0.8898,
+      "step": 54770
+    },
+    {
+      "epoch": 16.39,
+      "grad_norm": 1.9844911098480225,
+      "learning_rate": 3.926374966739063e-06,
+      "loss": 0.9253,
+      "step": 54775
+    },
+    {
+      "epoch": 16.39,
+      "grad_norm": 1.7941007614135742,
+      "learning_rate": 3.923214686863769e-06,
+      "loss": 0.9916,
+      "step": 54780
+    },
+    {
+      "epoch": 16.39,
+      "grad_norm": 3.715179681777954,
+      "learning_rate": 3.920055571037431e-06,
+      "loss": 0.9925,
+      "step": 54785
+    },
+    {
+      "epoch": 16.39,
+      "grad_norm": 3.1290266513824463,
+      "learning_rate": 3.916897619434509e-06,
+      "loss": 0.9698,
+      "step": 54790
+    },
+    {
+      "epoch": 16.39,
+      "grad_norm": 2.829307794570923,
+      "learning_rate": 3.913740832229415e-06,
+      "loss": 0.9611,
+      "step": 54795
+    },
+    {
+      "epoch": 16.4,
+      "grad_norm": 4.010441780090332,
+      "learning_rate": 3.910585209596499e-06,
+      "loss": 0.8892,
+      "step": 54800
+    },
+    {
+      "epoch": 16.4,
+      "grad_norm": 5.990212917327881,
+      "learning_rate": 3.907430751710039e-06,
+      "loss": 0.9162,
+      "step": 54805
+    },
+    {
+      "epoch": 16.4,
+      "grad_norm": 5.119846820831299,
+      "learning_rate": 3.904277458744254e-06,
+      "loss": 1.0017,
+      "step": 54810
+    },
+    {
+      "epoch": 16.4,
+      "grad_norm": 2.6883440017700195,
+      "learning_rate": 3.901125330873298e-06,
+      "loss": 0.9282,
+      "step": 54815
+    },
+    {
+      "epoch": 16.4,
+      "grad_norm": 4.281408309936523,
+      "learning_rate": 3.897974368271256e-06,
+      "loss": 0.9491,
+      "step": 54820
+    },
+    {
+      "epoch": 16.4,
+      "grad_norm": 1.7802320718765259,
+      "learning_rate": 3.89482457111216e-06,
+      "loss": 1.0643,
+      "step": 54825
+    },
+    {
+      "epoch": 16.4,
+      "grad_norm": 1.374393343925476,
+      "learning_rate": 3.891675939569961e-06,
+      "loss": 1.0454,
+      "step": 54830
+    },
+    {
+      "epoch": 16.41,
+      "grad_norm": 1.9680904150009155,
+      "learning_rate": 3.888528473818562e-06,
+      "loss": 0.8646,
+      "step": 54835
+    },
+    {
+      "epoch": 16.41,
+      "grad_norm": 3.9535000324249268,
+      "learning_rate": 3.88538217403179e-06,
+      "loss": 0.914,
+      "step": 54840
+    },
+    {
+      "epoch": 16.41,
+      "grad_norm": 4.569047451019287,
+      "learning_rate": 3.882237040383413e-06,
+      "loss": 0.8554,
+      "step": 54845
+    },
+    {
+      "epoch": 16.41,
+      "grad_norm": 2.8582353591918945,
+      "learning_rate": 3.879093073047141e-06,
+      "loss": 1.1247,
+      "step": 54850
+    },
+    {
+      "epoch": 16.41,
+      "grad_norm": 3.682445764541626,
+      "learning_rate": 3.87595027219659e-06,
+      "loss": 0.9073,
+      "step": 54855
+    },
+    {
+      "epoch": 16.41,
+      "grad_norm": 2.520266056060791,
+      "learning_rate": 3.872808638005362e-06,
+      "loss": 0.8065,
+      "step": 54860
+    },
+    {
+      "epoch": 16.41,
+      "grad_norm": 4.994065284729004,
+      "learning_rate": 3.8696681706469425e-06,
+      "loss": 0.9488,
+      "step": 54865
+    },
+    {
+      "epoch": 16.42,
+      "grad_norm": 2.4194717407226562,
+      "learning_rate": 3.866528870294795e-06,
+      "loss": 0.8722,
+      "step": 54870
+    },
+    {
+      "epoch": 16.42,
+      "grad_norm": 3.143922805786133,
+      "learning_rate": 3.86339073712228e-06,
+      "loss": 1.0721,
+      "step": 54875
+    },
+    {
+      "epoch": 16.42,
+      "grad_norm": 3.6272706985473633,
+      "learning_rate": 3.860253771302736e-06,
+      "loss": 1.1331,
+      "step": 54880
+    },
+    {
+      "epoch": 16.42,
+      "grad_norm": 2.595782518386841,
+      "learning_rate": 3.857117973009397e-06,
+      "loss": 0.7396,
+      "step": 54885
+    },
+    {
+      "epoch": 16.42,
+      "grad_norm": 2.1412410736083984,
+      "learning_rate": 3.853983342415454e-06,
+      "loss": 0.8936,
+      "step": 54890
+    },
+    {
+      "epoch": 16.42,
+      "grad_norm": 2.1500375270843506,
+      "learning_rate": 3.850849879694032e-06,
+      "loss": 0.9859,
+      "step": 54895
+    },
+    {
+      "epoch": 16.43,
+      "grad_norm": 2.9890248775482178,
+      "learning_rate": 3.84771758501819e-06,
+      "loss": 1.1094,
+      "step": 54900
+    },
+    {
+      "epoch": 16.43,
+      "grad_norm": 3.5275800228118896,
+      "learning_rate": 3.844586458560917e-06,
+      "loss": 1.0972,
+      "step": 54905
+    },
+    {
+      "epoch": 16.43,
+      "grad_norm": 3.5730512142181396,
+      "learning_rate": 3.841456500495144e-06,
+      "loss": 0.9291,
+      "step": 54910
+    },
+    {
+      "epoch": 16.43,
+      "grad_norm": 2.170400619506836,
+      "learning_rate": 3.838327710993736e-06,
+      "loss": 0.9088,
+      "step": 54915
+    },
+    {
+      "epoch": 16.43,
+      "grad_norm": 3.1143369674682617,
+      "learning_rate": 3.835200090229493e-06,
+      "loss": 0.9509,
+      "step": 54920
+    },
+    {
+      "epoch": 16.43,
+      "grad_norm": 2.1064987182617188,
+      "learning_rate": 3.832073638375147e-06,
+      "loss": 1.1217,
+      "step": 54925
+    },
+    {
+      "epoch": 16.43,
+      "grad_norm": 1.3867061138153076,
+      "learning_rate": 3.828948355603374e-06,
+      "loss": 0.9994,
+      "step": 54930
+    },
+    {
+      "epoch": 16.44,
+      "grad_norm": 3.5733249187469482,
+      "learning_rate": 3.8258242420867755e-06,
+      "loss": 0.9443,
+      "step": 54935
+    },
+    {
+      "epoch": 16.44,
+      "grad_norm": 4.986148357391357,
+      "learning_rate": 3.8227012979978935e-06,
+      "loss": 1.0024,
+      "step": 54940
+    },
+    {
+      "epoch": 16.44,
+      "grad_norm": 3.2679810523986816,
+      "learning_rate": 3.819579523509214e-06,
+      "loss": 0.8478,
+      "step": 54945
+    },
+    {
+      "epoch": 16.44,
+      "grad_norm": 1.716213345527649,
+      "learning_rate": 3.816458918793131e-06,
+      "loss": 0.7901,
+      "step": 54950
+    },
+    {
+      "epoch": 16.44,
+      "grad_norm": 2.6446051597595215,
+      "learning_rate": 3.813339484022013e-06,
+      "loss": 0.9754,
+      "step": 54955
+    },
+    {
+      "epoch": 16.44,
+      "grad_norm": 3.2618935108184814,
+      "learning_rate": 3.810221219368121e-06,
+      "loss": 1.1121,
+      "step": 54960
+    },
+    {
+      "epoch": 16.44,
+      "grad_norm": 2.8837664127349854,
+      "learning_rate": 3.8071041250036964e-06,
+      "loss": 0.9444,
+      "step": 54965
+    },
+    {
+      "epoch": 16.45,
+      "grad_norm": 3.107837438583374,
+      "learning_rate": 3.803988201100872e-06,
+      "loss": 1.2149,
+      "step": 54970
+    },
+    {
+      "epoch": 16.45,
+      "grad_norm": 1.7038583755493164,
+      "learning_rate": 3.80087344783176e-06,
+      "loss": 0.9996,
+      "step": 54975
+    },
+    {
+      "epoch": 16.45,
+      "grad_norm": 4.2032365798950195,
+      "learning_rate": 3.797759865368364e-06,
+      "loss": 0.9182,
+      "step": 54980
+    },
+    {
+      "epoch": 16.45,
+      "grad_norm": 3.1628682613372803,
+      "learning_rate": 3.794647453882652e-06,
+      "loss": 0.9363,
+      "step": 54985
+    },
+    {
+      "epoch": 16.45,
+      "grad_norm": 2.4282500743865967,
+      "learning_rate": 3.7915362135465205e-06,
+      "loss": 1.1558,
+      "step": 54990
+    },
+    {
+      "epoch": 16.45,
+      "grad_norm": 3.435473918914795,
+      "learning_rate": 3.7884261445318007e-06,
+      "loss": 0.959,
+      "step": 54995
+    },
+    {
+      "epoch": 16.46,
+      "grad_norm": 3.692786455154419,
+      "learning_rate": 3.7853172470102533e-06,
+      "loss": 0.9526,
+      "step": 55000
+    },
+    {
+      "epoch": 16.46,
+      "grad_norm": 1.5530604124069214,
+      "learning_rate": 3.782209521153579e-06,
+      "loss": 0.9287,
+      "step": 55005
+    },
+    {
+      "epoch": 16.46,
+      "grad_norm": 2.799910306930542,
+      "learning_rate": 3.77910296713343e-06,
+      "loss": 1.0503,
+      "step": 55010
+    },
+    {
+      "epoch": 16.46,
+      "grad_norm": 1.5998826026916504,
+      "learning_rate": 3.7759975851213603e-06,
+      "loss": 0.8919,
+      "step": 55015
+    },
+    {
+      "epoch": 16.46,
+      "grad_norm": 1.571890115737915,
+      "learning_rate": 3.772893375288883e-06,
+      "loss": 0.9862,
+      "step": 55020
+    },
+    {
+      "epoch": 16.46,
+      "grad_norm": 4.059834003448486,
+      "learning_rate": 3.7697903378074412e-06,
+      "loss": 1.0683,
+      "step": 55025
+    },
+    {
+      "epoch": 16.46,
+      "grad_norm": 1.3430641889572144,
+      "learning_rate": 3.7666884728484094e-06,
+      "loss": 1.0935,
+      "step": 55030
+    },
+    {
+      "epoch": 16.47,
+      "grad_norm": 3.6715023517608643,
+      "learning_rate": 3.7635877805831026e-06,
+      "loss": 0.9086,
+      "step": 55035
+    },
+    {
+      "epoch": 16.47,
+      "grad_norm": 2.4204914569854736,
+      "learning_rate": 3.760488261182768e-06,
+      "loss": 1.0692,
+      "step": 55040
+    },
+    {
+      "epoch": 16.47,
+      "grad_norm": 6.810561180114746,
+      "learning_rate": 3.7573899148185932e-06,
+      "loss": 1.2002,
+      "step": 55045
+    },
+    {
+      "epoch": 16.47,
+      "grad_norm": 3.171234607696533,
+      "learning_rate": 3.754292741661694e-06,
+      "loss": 1.0991,
+      "step": 55050
+    },
+    {
+      "epoch": 16.47,
+      "grad_norm": 3.974795341491699,
+      "learning_rate": 3.7511967418831145e-06,
+      "loss": 1.072,
+      "step": 55055
+    },
+    {
+      "epoch": 16.47,
+      "grad_norm": 1.6691629886627197,
+      "learning_rate": 3.748101915653862e-06,
+      "loss": 0.9038,
+      "step": 55060
+    },
+    {
+      "epoch": 16.47,
+      "grad_norm": 2.60886287689209,
+      "learning_rate": 3.745008263144842e-06,
+      "loss": 0.9588,
+      "step": 55065
+    },
+    {
+      "epoch": 16.48,
+      "grad_norm": 1.6360840797424316,
+      "learning_rate": 3.741915784526931e-06,
+      "loss": 0.9272,
+      "step": 55070
+    },
+    {
+      "epoch": 16.48,
+      "grad_norm": 1.671827793121338,
+      "learning_rate": 3.7388244799709095e-06,
+      "loss": 1.0059,
+      "step": 55075
+    },
+    {
+      "epoch": 16.48,
+      "grad_norm": 1.6585466861724854,
+      "learning_rate": 3.735734349647507e-06,
+      "loss": 1.0757,
+      "step": 55080
+    },
+    {
+      "epoch": 16.48,
+      "grad_norm": 2.521653890609741,
+      "learning_rate": 3.7326453937274047e-06,
+      "loss": 1.0236,
+      "step": 55085
+    },
+    {
+      "epoch": 16.48,
+      "grad_norm": 2.377997636795044,
+      "learning_rate": 3.7295576123811795e-06,
+      "loss": 0.9756,
+      "step": 55090
+    },
+    {
+      "epoch": 16.48,
+      "grad_norm": 2.9460082054138184,
+      "learning_rate": 3.726471005779389e-06,
+      "loss": 1.1004,
+      "step": 55095
+    },
+    {
+      "epoch": 16.49,
+      "grad_norm": 3.480149507522583,
+      "learning_rate": 3.7233855740924805e-06,
+      "loss": 0.8099,
+      "step": 55100
+    },
+    {
+      "epoch": 16.49,
+      "grad_norm": 1.956493616104126,
+      "learning_rate": 3.7203013174908816e-06,
+      "loss": 0.8011,
+      "step": 55105
+    },
+    {
+      "epoch": 16.49,
+      "grad_norm": 3.8703737258911133,
+      "learning_rate": 3.7172182361449174e-06,
+      "loss": 1.013,
+      "step": 55110
+    },
+    {
+      "epoch": 16.49,
+      "grad_norm": 1.4395601749420166,
+      "learning_rate": 3.714136330224868e-06,
+      "loss": 0.9137,
+      "step": 55115
+    },
+    {
+      "epoch": 16.49,
+      "grad_norm": 0.9838583469390869,
+      "learning_rate": 3.711055599900945e-06,
+      "loss": 0.941,
+      "step": 55120
+    },
+    {
+      "epoch": 16.49,
+      "grad_norm": 2.752509355545044,
+      "learning_rate": 3.7079760453432895e-06,
+      "loss": 0.9234,
+      "step": 55125
+    },
+    {
+      "epoch": 16.49,
+      "grad_norm": 3.263653516769409,
+      "learning_rate": 3.7048976667219877e-06,
+      "loss": 0.974,
+      "step": 55130
+    },
+    {
+      "epoch": 16.5,
+      "grad_norm": 3.148515462875366,
+      "learning_rate": 3.701820464207054e-06,
+      "loss": 0.8743,
+      "step": 55135
+    },
+    {
+      "epoch": 16.5,
+      "grad_norm": 4.454446315765381,
+      "learning_rate": 3.6987444379684348e-06,
+      "loss": 1.0195,
+      "step": 55140
+    },
+    {
+      "epoch": 16.5,
+      "grad_norm": 3.798888921737671,
+      "learning_rate": 3.695669588176026e-06,
+      "loss": 0.9871,
+      "step": 55145
+    },
+    {
+      "epoch": 16.5,
+      "grad_norm": 1.551338791847229,
+      "learning_rate": 3.6925959149996274e-06,
+      "loss": 0.9701,
+      "step": 55150
+    },
+    {
+      "epoch": 16.5,
+      "grad_norm": 4.081212997436523,
+      "learning_rate": 3.689523418609017e-06,
+      "loss": 1.0122,
+      "step": 55155
+    },
+    {
+      "epoch": 16.5,
+      "grad_norm": 2.6100282669067383,
+      "learning_rate": 3.686452099173876e-06,
+      "loss": 0.8688,
+      "step": 55160
+    },
+    {
+      "epoch": 16.5,
+      "grad_norm": 3.0267417430877686,
+      "learning_rate": 3.683381956863832e-06,
+      "loss": 0.9878,
+      "step": 55165
+    },
+    {
+      "epoch": 16.51,
+      "grad_norm": 3.891383409500122,
+      "learning_rate": 3.680312991848445e-06,
+      "loss": 1.0131,
+      "step": 55170
+    },
+    {
+      "epoch": 16.51,
+      "grad_norm": 2.7012784481048584,
+      "learning_rate": 3.677245204297211e-06,
+      "loss": 1.1283,
+      "step": 55175
+    },
+    {
+      "epoch": 16.51,
+      "grad_norm": 1.8227554559707642,
+      "learning_rate": 3.674178594379565e-06,
+      "loss": 1.0563,
+      "step": 55180
+    },
+    {
+      "epoch": 16.51,
+      "grad_norm": 3.7222025394439697,
+      "learning_rate": 3.671113162264861e-06,
+      "loss": 0.9301,
+      "step": 55185
+    },
+    {
+      "epoch": 16.51,
+      "grad_norm": 2.2444727420806885,
+      "learning_rate": 3.668048908122415e-06,
+      "loss": 0.954,
+      "step": 55190
+    },
+    {
+      "epoch": 16.51,
+      "grad_norm": 3.4596662521362305,
+      "learning_rate": 3.6649858321214424e-06,
+      "loss": 0.9375,
+      "step": 55195
+    },
+    {
+      "epoch": 16.52,
+      "grad_norm": 4.218637466430664,
+      "learning_rate": 3.6619239344311383e-06,
+      "loss": 1.0127,
+      "step": 55200
+    },
+    {
+      "epoch": 16.52,
+      "grad_norm": 2.6548755168914795,
+      "learning_rate": 3.6588632152205897e-06,
+      "loss": 0.9714,
+      "step": 55205
+    },
+    {
+      "epoch": 16.52,
+      "grad_norm": 2.7811079025268555,
+      "learning_rate": 3.655803674658842e-06,
+      "loss": 0.8688,
+      "step": 55210
+    },
+    {
+      "epoch": 16.52,
+      "grad_norm": 1.8681191205978394,
+      "learning_rate": 3.6527453129148738e-06,
+      "loss": 1.0711,
+      "step": 55215
+    },
+    {
+      "epoch": 16.52,
+      "grad_norm": 1.8448801040649414,
+      "learning_rate": 3.6496881301575887e-06,
+      "loss": 0.9248,
+      "step": 55220
+    },
+    {
+      "epoch": 16.52,
+      "grad_norm": 3.9367079734802246,
+      "learning_rate": 3.6466321265558406e-06,
+      "loss": 0.994,
+      "step": 55225
+    },
+    {
+      "epoch": 16.52,
+      "grad_norm": 3.2302794456481934,
+      "learning_rate": 3.6435773022784004e-06,
+      "loss": 0.8619,
+      "step": 55230
+    },
+    {
+      "epoch": 16.53,
+      "grad_norm": 4.870388984680176,
+      "learning_rate": 3.640523657493991e-06,
+      "loss": 0.9849,
+      "step": 55235
+    },
+    {
+      "epoch": 16.53,
+      "grad_norm": 1.335530400276184,
+      "learning_rate": 3.637471192371256e-06,
+      "loss": 1.0656,
+      "step": 55240
+    },
+    {
+      "epoch": 16.53,
+      "grad_norm": 7.603557586669922,
+      "learning_rate": 3.6344199070787815e-06,
+      "loss": 1.1671,
+      "step": 55245
+    },
+    {
+      "epoch": 16.53,
+      "grad_norm": 2.9642961025238037,
+      "learning_rate": 3.6313698017850866e-06,
+      "loss": 1.0257,
+      "step": 55250
+    },
+    {
+      "epoch": 16.53,
+      "grad_norm": 2.462388277053833,
+      "learning_rate": 3.628320876658628e-06,
+      "loss": 0.8544,
+      "step": 55255
+    },
+    {
+      "epoch": 16.53,
+      "grad_norm": 1.824250340461731,
+      "learning_rate": 3.6252731318677902e-06,
+      "loss": 1.0885,
+      "step": 55260
+    },
+    {
+      "epoch": 16.53,
+      "grad_norm": 6.232416152954102,
+      "learning_rate": 3.6222265675809025e-06,
+      "loss": 0.9456,
+      "step": 55265
+    },
+    {
+      "epoch": 16.54,
+      "grad_norm": 3.9859626293182373,
+      "learning_rate": 3.6191811839662194e-06,
+      "loss": 0.7835,
+      "step": 55270
+    },
+    {
+      "epoch": 16.54,
+      "grad_norm": 3.1337802410125732,
+      "learning_rate": 3.616136981191939e-06,
+      "loss": 1.0781,
+      "step": 55275
+    },
+    {
+      "epoch": 16.54,
+      "grad_norm": 2.0644679069519043,
+      "learning_rate": 3.6130939594261783e-06,
+      "loss": 0.9863,
+      "step": 55280
+    },
+    {
+      "epoch": 16.54,
+      "grad_norm": 4.25192403793335,
+      "learning_rate": 3.610052118837015e-06,
+      "loss": 0.8807,
+      "step": 55285
+    },
+    {
+      "epoch": 16.54,
+      "grad_norm": 1.9776525497436523,
+      "learning_rate": 3.6070114595924327e-06,
+      "loss": 0.9195,
+      "step": 55290
+    },
+    {
+      "epoch": 16.54,
+      "grad_norm": 1.4639649391174316,
+      "learning_rate": 3.6039719818603797e-06,
+      "loss": 1.0648,
+      "step": 55295
+    },
+    {
+      "epoch": 16.55,
+      "grad_norm": 1.7178945541381836,
+      "learning_rate": 3.600933685808708e-06,
+      "loss": 0.921,
+      "step": 55300
+    },
+    {
+      "epoch": 16.55,
+      "grad_norm": 2.7617597579956055,
+      "learning_rate": 3.5978965716052248e-06,
+      "loss": 0.9255,
+      "step": 55305
+    },
+    {
+      "epoch": 16.55,
+      "grad_norm": 1.6579577922821045,
+      "learning_rate": 3.5948606394176687e-06,
+      "loss": 0.8896,
+      "step": 55310
+    },
+    {
+      "epoch": 16.55,
+      "grad_norm": 5.555087089538574,
+      "learning_rate": 3.5918258894137107e-06,
+      "loss": 0.7572,
+      "step": 55315
+    },
+    {
+      "epoch": 16.55,
+      "grad_norm": 3.7583398818969727,
+      "learning_rate": 3.5887923217609586e-06,
+      "loss": 0.8979,
+      "step": 55320
+    },
+    {
+      "epoch": 16.55,
+      "grad_norm": 2.396233081817627,
+      "learning_rate": 3.585759936626948e-06,
+      "loss": 0.9145,
+      "step": 55325
+    },
+    {
+      "epoch": 16.55,
+      "grad_norm": 2.1706974506378174,
+      "learning_rate": 3.5827287341791583e-06,
+      "loss": 0.9833,
+      "step": 55330
+    },
+    {
+      "epoch": 16.56,
+      "grad_norm": 21.9448184967041,
+      "learning_rate": 3.5796987145849974e-06,
+      "loss": 0.9713,
+      "step": 55335
+    },
+    {
+      "epoch": 16.56,
+      "grad_norm": 1.5266534090042114,
+      "learning_rate": 3.5766698780118125e-06,
+      "loss": 0.9322,
+      "step": 55340
+    },
+    {
+      "epoch": 16.56,
+      "grad_norm": 3.1922240257263184,
+      "learning_rate": 3.5736422246268804e-06,
+      "loss": 0.9115,
+      "step": 55345
+    },
+    {
+      "epoch": 16.56,
+      "grad_norm": 1.6911910772323608,
+      "learning_rate": 3.5706157545974174e-06,
+      "loss": 1.0724,
+      "step": 55350
+    },
+    {
+      "epoch": 16.56,
+      "grad_norm": 1.6051517724990845,
+      "learning_rate": 3.5675904680905704e-06,
+      "loss": 0.9671,
+      "step": 55355
+    },
+    {
+      "epoch": 16.56,
+      "grad_norm": 3.0182387828826904,
+      "learning_rate": 3.5645663652734277e-06,
+      "loss": 0.9052,
+      "step": 55360
+    },
+    {
+      "epoch": 16.56,
+      "grad_norm": 2.883270740509033,
+      "learning_rate": 3.561543446313001e-06,
+      "loss": 1.1312,
+      "step": 55365
+    },
+    {
+      "epoch": 16.57,
+      "grad_norm": 5.493057727813721,
+      "learning_rate": 3.5585217113762533e-06,
+      "loss": 0.889,
+      "step": 55370
+    },
+    {
+      "epoch": 16.57,
+      "grad_norm": 2.2365972995758057,
+      "learning_rate": 3.5555011606300507e-06,
+      "loss": 0.9428,
+      "step": 55375
+    },
+    {
+      "epoch": 16.57,
+      "grad_norm": 2.719250440597534,
+      "learning_rate": 3.5524817942412414e-06,
+      "loss": 0.9899,
+      "step": 55380
+    },
+    {
+      "epoch": 16.57,
+      "grad_norm": 1.566593050956726,
+      "learning_rate": 3.549463612376558e-06,
+      "loss": 0.8925,
+      "step": 55385
+    },
+    {
+      "epoch": 16.57,
+      "grad_norm": 3.2841618061065674,
+      "learning_rate": 3.5464466152027113e-06,
+      "loss": 1.0915,
+      "step": 55390
+    },
+    {
+      "epoch": 16.57,
+      "grad_norm": 3.8649485111236572,
+      "learning_rate": 3.5434308028863107e-06,
+      "loss": 0.9006,
+      "step": 55395
+    },
+    {
+      "epoch": 16.58,
+      "grad_norm": 3.333514928817749,
+      "learning_rate": 3.540416175593933e-06,
+      "loss": 1.0511,
+      "step": 55400
+    },
+    {
+      "epoch": 16.58,
+      "grad_norm": 3.3573429584503174,
+      "learning_rate": 3.5374027334920595e-06,
+      "loss": 1.0593,
+      "step": 55405
+    },
+    {
+      "epoch": 16.58,
+      "grad_norm": 3.672663450241089,
+      "learning_rate": 3.5343904767471235e-06,
+      "loss": 0.8411,
+      "step": 55410
+    },
+    {
+      "epoch": 16.58,
+      "grad_norm": 3.1080899238586426,
+      "learning_rate": 3.5313794055254894e-06,
+      "loss": 0.9778,
+      "step": 55415
+    },
+    {
+      "epoch": 16.58,
+      "grad_norm": 3.76132869720459,
+      "learning_rate": 3.5283695199934545e-06,
+      "loss": 1.039,
+      "step": 55420
+    },
+    {
+      "epoch": 16.58,
+      "grad_norm": 1.2220946550369263,
+      "learning_rate": 3.5253608203172527e-06,
+      "loss": 0.9653,
+      "step": 55425
+    },
+    {
+      "epoch": 16.58,
+      "grad_norm": 2.680553674697876,
+      "learning_rate": 3.5223533066630538e-06,
+      "loss": 0.9479,
+      "step": 55430
+    },
+    {
+      "epoch": 16.59,
+      "grad_norm": 2.850705862045288,
+      "learning_rate": 3.5193469791969523e-06,
+      "loss": 0.9687,
+      "step": 55435
+    },
+    {
+      "epoch": 16.59,
+      "grad_norm": 6.794698715209961,
+      "learning_rate": 3.516341838084994e-06,
+      "loss": 0.9051,
+      "step": 55440
+    },
+    {
+      "epoch": 16.59,
+      "grad_norm": 1.4583183526992798,
+      "learning_rate": 3.5133378834931424e-06,
+      "loss": 0.9477,
+      "step": 55445
+    },
+    {
+      "epoch": 16.59,
+      "grad_norm": 1.7493852376937866,
+      "learning_rate": 3.5103351155873044e-06,
+      "loss": 1.0363,
+      "step": 55450
+    },
+    {
+      "epoch": 16.59,
+      "grad_norm": 2.9901397228240967,
+      "learning_rate": 3.5073335345333247e-06,
+      "loss": 1.0936,
+      "step": 55455
+    },
+    {
+      "epoch": 16.59,
+      "grad_norm": 2.2844088077545166,
+      "learning_rate": 3.504333140496968e-06,
+      "loss": 0.9662,
+      "step": 55460
+    },
+    {
+      "epoch": 16.59,
+      "grad_norm": 1.7796694040298462,
+      "learning_rate": 3.5013339336439576e-06,
+      "loss": 0.9429,
+      "step": 55465
+    },
+    {
+      "epoch": 16.6,
+      "grad_norm": 1.1826192140579224,
+      "learning_rate": 3.4983359141399164e-06,
+      "loss": 1.0261,
+      "step": 55470
+    },
+    {
+      "epoch": 16.6,
+      "grad_norm": 2.0418243408203125,
+      "learning_rate": 3.4953390821504423e-06,
+      "loss": 0.9578,
+      "step": 55475
+    },
+    {
+      "epoch": 16.6,
+      "grad_norm": 6.461709022521973,
+      "learning_rate": 3.492343437841028e-06,
+      "loss": 0.9947,
+      "step": 55480
+    },
+    {
+      "epoch": 16.6,
+      "grad_norm": 3.240283489227295,
+      "learning_rate": 3.4893489813771413e-06,
+      "loss": 0.8213,
+      "step": 55485
+    },
+    {
+      "epoch": 16.6,
+      "grad_norm": 2.8579697608947754,
+      "learning_rate": 3.486355712924139e-06,
+      "loss": 0.912,
+      "step": 55490
+    },
+    {
+      "epoch": 16.6,
+      "grad_norm": 2.917512893676758,
+      "learning_rate": 3.48336363264736e-06,
+      "loss": 0.9478,
+      "step": 55495
+    },
+    {
+      "epoch": 16.6,
+      "grad_norm": 2.837340831756592,
+      "learning_rate": 3.4803727407120375e-06,
+      "loss": 1.032,
+      "step": 55500
+    },
+    {
+      "epoch": 16.61,
+      "grad_norm": 3.3056883811950684,
+      "learning_rate": 3.4773830372833576e-06,
+      "loss": 1.1411,
+      "step": 55505
+    },
+    {
+      "epoch": 16.61,
+      "grad_norm": 4.646909713745117,
+      "learning_rate": 3.474394522526442e-06,
+      "loss": 0.9487,
+      "step": 55510
+    },
+    {
+      "epoch": 16.61,
+      "grad_norm": 5.3782124519348145,
+      "learning_rate": 3.4714071966063436e-06,
+      "loss": 0.9774,
+      "step": 55515
+    },
+    {
+      "epoch": 16.61,
+      "grad_norm": 1.5361804962158203,
+      "learning_rate": 3.4684210596880486e-06,
+      "loss": 1.0604,
+      "step": 55520
+    },
+    {
+      "epoch": 16.61,
+      "grad_norm": 2.3509507179260254,
+      "learning_rate": 3.4654361119364764e-06,
+      "loss": 0.9274,
+      "step": 55525
+    },
+    {
+      "epoch": 16.61,
+      "grad_norm": 4.507537841796875,
+      "learning_rate": 3.4624523535164823e-06,
+      "loss": 0.8698,
+      "step": 55530
+    },
+    {
+      "epoch": 16.62,
+      "grad_norm": 2.660539388656616,
+      "learning_rate": 3.459469784592859e-06,
+      "loss": 0.9137,
+      "step": 55535
+    },
+    {
+      "epoch": 16.62,
+      "grad_norm": 5.054265022277832,
+      "learning_rate": 3.4564884053303298e-06,
+      "loss": 0.9331,
+      "step": 55540
+    },
+    {
+      "epoch": 16.62,
+      "grad_norm": 1.3174716234207153,
+      "learning_rate": 3.4535082158935527e-06,
+      "loss": 1.135,
+      "step": 55545
+    },
+    {
+      "epoch": 16.62,
+      "grad_norm": 2.0269131660461426,
+      "learning_rate": 3.450529216447121e-06,
+      "loss": 1.0076,
+      "step": 55550
+    },
+    {
+      "epoch": 16.62,
+      "grad_norm": 5.588700771331787,
+      "learning_rate": 3.447551407155561e-06,
+      "loss": 0.9909,
+      "step": 55555
+    },
+    {
+      "epoch": 16.62,
+      "grad_norm": 2.787294864654541,
+      "learning_rate": 3.4445747881833386e-06,
+      "loss": 1.0891,
+      "step": 55560
+    },
+    {
+      "epoch": 16.62,
+      "grad_norm": 2.2700955867767334,
+      "learning_rate": 3.441599359694836e-06,
+      "loss": 0.9488,
+      "step": 55565
+    },
+    {
+      "epoch": 16.63,
+      "grad_norm": 1.9791069030761719,
+      "learning_rate": 3.4386251218543996e-06,
+      "loss": 1.0003,
+      "step": 55570
+    },
+    {
+      "epoch": 16.63,
+      "grad_norm": 2.2911648750305176,
+      "learning_rate": 3.435652074826279e-06,
+      "loss": 1.0334,
+      "step": 55575
+    },
+    {
+      "epoch": 16.63,
+      "grad_norm": 5.44551420211792,
+      "learning_rate": 3.4326802187746864e-06,
+      "loss": 0.9583,
+      "step": 55580
+    },
+    {
+      "epoch": 16.63,
+      "grad_norm": 1.6152199506759644,
+      "learning_rate": 3.4297095538637388e-06,
+      "loss": 1.0084,
+      "step": 55585
+    },
+    {
+      "epoch": 16.63,
+      "grad_norm": 3.9968864917755127,
+      "learning_rate": 3.4267400802575233e-06,
+      "loss": 0.888,
+      "step": 55590
+    },
+    {
+      "epoch": 16.63,
+      "grad_norm": 1.834970474243164,
+      "learning_rate": 3.423771798120021e-06,
+      "loss": 0.8411,
+      "step": 55595
+    },
+    {
+      "epoch": 16.63,
+      "grad_norm": 1.1965527534484863,
+      "learning_rate": 3.4208047076151773e-06,
+      "loss": 0.9757,
+      "step": 55600
+    },
+    {
+      "epoch": 16.64,
+      "grad_norm": 2.324232578277588,
+      "learning_rate": 3.417838808906859e-06,
+      "loss": 0.9249,
+      "step": 55605
+    },
+    {
+      "epoch": 16.64,
+      "grad_norm": 1.6368885040283203,
+      "learning_rate": 3.4148741021588686e-06,
+      "loss": 1.0691,
+      "step": 55610
+    },
+    {
+      "epoch": 16.64,
+      "grad_norm": 2.840993881225586,
+      "learning_rate": 3.4119105875349444e-06,
+      "loss": 0.998,
+      "step": 55615
+    },
+    {
+      "epoch": 16.64,
+      "grad_norm": 4.377618789672852,
+      "learning_rate": 3.4089482651987607e-06,
+      "loss": 1.0532,
+      "step": 55620
+    },
+    {
+      "epoch": 16.64,
+      "grad_norm": 3.657803535461426,
+      "learning_rate": 3.4059871353139207e-06,
+      "loss": 0.8328,
+      "step": 55625
+    },
+    {
+      "epoch": 16.64,
+      "grad_norm": 2.706447124481201,
+      "learning_rate": 3.403027198043968e-06,
+      "loss": 0.8507,
+      "step": 55630
+    },
+    {
+      "epoch": 16.65,
+      "grad_norm": 2.529573917388916,
+      "learning_rate": 3.4000684535523714e-06,
+      "loss": 0.9334,
+      "step": 55635
+    },
+    {
+      "epoch": 16.65,
+      "grad_norm": 3.0248026847839355,
+      "learning_rate": 3.397110902002543e-06,
+      "loss": 0.9053,
+      "step": 55640
+    },
+    {
+      "epoch": 16.65,
+      "grad_norm": 1.9483550786972046,
+      "learning_rate": 3.394154543557826e-06,
+      "loss": 0.8746,
+      "step": 55645
+    },
+    {
+      "epoch": 16.65,
+      "grad_norm": 6.750313758850098,
+      "learning_rate": 3.3911993783814934e-06,
+      "loss": 0.9609,
+      "step": 55650
+    },
+    {
+      "epoch": 16.65,
+      "grad_norm": 4.349567890167236,
+      "learning_rate": 3.388245406636764e-06,
+      "loss": 0.9785,
+      "step": 55655
+    },
+    {
+      "epoch": 16.65,
+      "grad_norm": 4.089972019195557,
+      "learning_rate": 3.3852926284867665e-06,
+      "loss": 1.1758,
+      "step": 55660
+    },
+    {
+      "epoch": 16.65,
+      "grad_norm": 1.6726969480514526,
+      "learning_rate": 3.3823410440946e-06,
+      "loss": 0.8968,
+      "step": 55665
+    },
+    {
+      "epoch": 16.66,
+      "grad_norm": 12.856719970703125,
+      "learning_rate": 3.379390653623257e-06,
+      "loss": 0.9887,
+      "step": 55670
+    },
+    {
+      "epoch": 16.66,
+      "grad_norm": 4.076183795928955,
+      "learning_rate": 3.3764414572357046e-06,
+      "loss": 0.8843,
+      "step": 55675
+    },
+    {
+      "epoch": 16.66,
+      "grad_norm": 3.290557384490967,
+      "learning_rate": 3.3734934550948034e-06,
+      "loss": 0.9424,
+      "step": 55680
+    },
+    {
+      "epoch": 16.66,
+      "grad_norm": 1.8041436672210693,
+      "learning_rate": 3.37054664736339e-06,
+      "loss": 1.054,
+      "step": 55685
+    },
+    {
+      "epoch": 16.66,
+      "grad_norm": 2.115232467651367,
+      "learning_rate": 3.3676010342041988e-06,
+      "loss": 1.0038,
+      "step": 55690
+    },
+    {
+      "epoch": 16.66,
+      "grad_norm": 2.6543099880218506,
+      "learning_rate": 3.364656615779918e-06,
+      "loss": 0.9418,
+      "step": 55695
+    },
+    {
+      "epoch": 16.66,
+      "grad_norm": 1.230863332748413,
+      "learning_rate": 3.36171339225316e-06,
+      "loss": 0.9834,
+      "step": 55700
+    },
+    {
+      "epoch": 16.67,
+      "grad_norm": 4.0604047775268555,
+      "learning_rate": 3.358771363786481e-06,
+      "loss": 0.9245,
+      "step": 55705
+    },
+    {
+      "epoch": 16.67,
+      "grad_norm": 15.559052467346191,
+      "learning_rate": 3.355830530542367e-06,
+      "loss": 0.9499,
+      "step": 55710
+    },
+    {
+      "epoch": 16.67,
+      "grad_norm": 7.960972309112549,
+      "learning_rate": 3.3528908926832304e-06,
+      "loss": 0.9759,
+      "step": 55715
+    },
+    {
+      "epoch": 16.67,
+      "grad_norm": 3.3675858974456787,
+      "learning_rate": 3.3499524503714326e-06,
+      "loss": 0.9966,
+      "step": 55720
+    },
+    {
+      "epoch": 16.67,
+      "grad_norm": 2.5648207664489746,
+      "learning_rate": 3.3470152037692547e-06,
+      "loss": 1.1776,
+      "step": 55725
+    },
+    {
+      "epoch": 16.67,
+      "grad_norm": 3.301900625228882,
+      "learning_rate": 3.3440791530389232e-06,
+      "loss": 1.0791,
+      "step": 55730
+    },
+    {
+      "epoch": 16.68,
+      "grad_norm": 1.4749406576156616,
+      "learning_rate": 3.3411442983425885e-06,
+      "loss": 1.0181,
+      "step": 55735
+    },
+    {
+      "epoch": 16.68,
+      "grad_norm": 1.6755149364471436,
+      "learning_rate": 3.33821063984234e-06,
+      "loss": 1.0547,
+      "step": 55740
+    },
+    {
+      "epoch": 16.68,
+      "grad_norm": 3.6287951469421387,
+      "learning_rate": 3.335278177700202e-06,
+      "loss": 0.9052,
+      "step": 55745
+    },
+    {
+      "epoch": 16.68,
+      "grad_norm": 2.7663354873657227,
+      "learning_rate": 3.3323469120781326e-06,
+      "loss": 0.8988,
+      "step": 55750
+    },
+    {
+      "epoch": 16.68,
+      "grad_norm": 2.6993236541748047,
+      "learning_rate": 3.329416843138017e-06,
+      "loss": 0.983,
+      "step": 55755
+    },
+    {
+      "epoch": 16.68,
+      "grad_norm": 6.778196811676025,
+      "learning_rate": 3.3264879710416917e-06,
+      "loss": 1.0198,
+      "step": 55760
+    },
+    {
+      "epoch": 16.68,
+      "grad_norm": 1.4923733472824097,
+      "learning_rate": 3.3235602959508977e-06,
+      "loss": 0.9077,
+      "step": 55765
+    },
+    {
+      "epoch": 16.69,
+      "grad_norm": 2.7575523853302,
+      "learning_rate": 3.3206338180273465e-06,
+      "loss": 0.8404,
+      "step": 55770
+    },
+    {
+      "epoch": 16.69,
+      "grad_norm": 4.008522033691406,
+      "learning_rate": 3.3177085374326453e-06,
+      "loss": 0.7967,
+      "step": 55775
+    },
+    {
+      "epoch": 16.69,
+      "grad_norm": 3.0780680179595947,
+      "learning_rate": 3.314784454328376e-06,
+      "loss": 0.9908,
+      "step": 55780
+    },
+    {
+      "epoch": 16.69,
+      "grad_norm": 3.2288169860839844,
+      "learning_rate": 3.3118615688760124e-06,
+      "loss": 0.9111,
+      "step": 55785
+    },
+    {
+      "epoch": 16.69,
+      "grad_norm": 2.5853915214538574,
+      "learning_rate": 3.3089398812369942e-06,
+      "loss": 1.0616,
+      "step": 55790
+    },
+    {
+      "epoch": 16.69,
+      "grad_norm": 2.06427264213562,
+      "learning_rate": 3.3060193915726763e-06,
+      "loss": 0.997,
+      "step": 55795
+    },
+    {
+      "epoch": 16.69,
+      "grad_norm": 9.891775131225586,
+      "learning_rate": 3.3031001000443597e-06,
+      "loss": 1.0209,
+      "step": 55800
+    },
+    {
+      "epoch": 16.7,
+      "grad_norm": 2.132046937942505,
+      "learning_rate": 3.300182006813271e-06,
+      "loss": 0.9908,
+      "step": 55805
+    },
+    {
+      "epoch": 16.7,
+      "grad_norm": 2.466611862182617,
+      "learning_rate": 3.297265112040568e-06,
+      "loss": 1.0584,
+      "step": 55810
+    },
+    {
+      "epoch": 16.7,
+      "grad_norm": 1.6372896432876587,
+      "learning_rate": 3.2943494158873684e-06,
+      "loss": 0.8243,
+      "step": 55815
+    },
+    {
+      "epoch": 16.7,
+      "grad_norm": 2.080674886703491,
+      "learning_rate": 3.29143491851468e-06,
+      "loss": 0.9725,
+      "step": 55820
+    },
+    {
+      "epoch": 16.7,
+      "grad_norm": 2.6015894412994385,
+      "learning_rate": 3.288521620083479e-06,
+      "loss": 1.0789,
+      "step": 55825
+    },
+    {
+      "epoch": 16.7,
+      "grad_norm": 2.357999086380005,
+      "learning_rate": 3.2856095207546595e-06,
+      "loss": 0.9111,
+      "step": 55830
+    },
+    {
+      "epoch": 16.71,
+      "grad_norm": 2.4748330116271973,
+      "learning_rate": 3.2826986206890537e-06,
+      "loss": 1.0326,
+      "step": 55835
+    },
+    {
+      "epoch": 16.71,
+      "grad_norm": 3.481816291809082,
+      "learning_rate": 3.279788920047433e-06,
+      "loss": 0.9831,
+      "step": 55840
+    },
+    {
+      "epoch": 16.71,
+      "grad_norm": 2.6234841346740723,
+      "learning_rate": 3.2768804189904913e-06,
+      "loss": 1.0191,
+      "step": 55845
+    },
+    {
+      "epoch": 16.71,
+      "grad_norm": 1.8331263065338135,
+      "learning_rate": 3.2739731176788636e-06,
+      "loss": 0.9873,
+      "step": 55850
+    },
+    {
+      "epoch": 16.71,
+      "grad_norm": 1.4677926301956177,
+      "learning_rate": 3.271067016273124e-06,
+      "loss": 0.9661,
+      "step": 55855
+    },
+    {
+      "epoch": 16.71,
+      "grad_norm": 2.3680524826049805,
+      "learning_rate": 3.2681621149337538e-06,
+      "loss": 0.862,
+      "step": 55860
+    },
+    {
+      "epoch": 16.71,
+      "grad_norm": 1.7516663074493408,
+      "learning_rate": 3.265258413821215e-06,
+      "loss": 0.9544,
+      "step": 55865
+    },
+    {
+      "epoch": 16.72,
+      "grad_norm": 1.8315908908843994,
+      "learning_rate": 3.262355913095852e-06,
+      "loss": 0.9942,
+      "step": 55870
+    },
+    {
+      "epoch": 16.72,
+      "grad_norm": 2.081782341003418,
+      "learning_rate": 3.259454612917984e-06,
+      "loss": 1.2045,
+      "step": 55875
+    },
+    {
+      "epoch": 16.72,
+      "grad_norm": 2.6945362091064453,
+      "learning_rate": 3.256554513447832e-06,
+      "loss": 1.0241,
+      "step": 55880
+    },
+    {
+      "epoch": 16.72,
+      "grad_norm": 3.111546277999878,
+      "learning_rate": 3.253655614845583e-06,
+      "loss": 0.9431,
+      "step": 55885
+    },
+    {
+      "epoch": 16.72,
+      "grad_norm": 3.409693717956543,
+      "learning_rate": 3.250757917271324e-06,
+      "loss": 1.1223,
+      "step": 55890
+    },
+    {
+      "epoch": 16.72,
+      "grad_norm": 3.733633279800415,
+      "learning_rate": 3.2478614208850928e-06,
+      "loss": 1.0297,
+      "step": 55895
+    },
+    {
+      "epoch": 16.72,
+      "grad_norm": 2.1389636993408203,
+      "learning_rate": 3.2449661258468768e-06,
+      "loss": 0.9982,
+      "step": 55900
+    },
+    {
+      "epoch": 16.73,
+      "grad_norm": 2.501753568649292,
+      "learning_rate": 3.2420720323165566e-06,
+      "loss": 0.9889,
+      "step": 55905
+    },
+    {
+      "epoch": 16.73,
+      "grad_norm": 1.5202560424804688,
+      "learning_rate": 3.239179140453996e-06,
+      "loss": 0.8041,
+      "step": 55910
+    },
+    {
+      "epoch": 16.73,
+      "grad_norm": 4.239963054656982,
+      "learning_rate": 3.2362874504189428e-06,
+      "loss": 0.9234,
+      "step": 55915
+    },
+    {
+      "epoch": 16.73,
+      "grad_norm": 3.403322458267212,
+      "learning_rate": 3.2333969623711157e-06,
+      "loss": 0.935,
+      "step": 55920
+    },
+    {
+      "epoch": 16.73,
+      "grad_norm": 1.6759166717529297,
+      "learning_rate": 3.2305076764701458e-06,
+      "loss": 0.895,
+      "step": 55925
+    },
+    {
+      "epoch": 16.73,
+      "grad_norm": 3.4392189979553223,
+      "learning_rate": 3.2276195928756107e-06,
+      "loss": 0.9788,
+      "step": 55930
+    },
+    {
+      "epoch": 16.74,
+      "grad_norm": 2.1375179290771484,
+      "learning_rate": 3.224732711747014e-06,
+      "loss": 0.8672,
+      "step": 55935
+    },
+    {
+      "epoch": 16.74,
+      "grad_norm": 1.1759334802627563,
+      "learning_rate": 3.221847033243794e-06,
+      "loss": 1.0013,
+      "step": 55940
+    },
+    {
+      "epoch": 16.74,
+      "grad_norm": 2.3137335777282715,
+      "learning_rate": 3.2189625575253265e-06,
+      "loss": 0.8176,
+      "step": 55945
+    },
+    {
+      "epoch": 16.74,
+      "grad_norm": 1.971378207206726,
+      "learning_rate": 3.2160792847509225e-06,
+      "loss": 0.8394,
+      "step": 55950
+    },
+    {
+      "epoch": 16.74,
+      "grad_norm": 1.7172585725784302,
+      "learning_rate": 3.213197215079805e-06,
+      "loss": 0.9224,
+      "step": 55955
+    },
+    {
+      "epoch": 16.74,
+      "grad_norm": 6.083642482757568,
+      "learning_rate": 3.210316348671169e-06,
+      "loss": 1.0735,
+      "step": 55960
+    },
+    {
+      "epoch": 16.74,
+      "grad_norm": 1.318853497505188,
+      "learning_rate": 3.2074366856841036e-06,
+      "loss": 1.0313,
+      "step": 55965
+    },
+    {
+      "epoch": 16.75,
+      "grad_norm": 2.593428134918213,
+      "learning_rate": 3.2045582262776626e-06,
+      "loss": 0.9487,
+      "step": 55970
+    },
+    {
+      "epoch": 16.75,
+      "grad_norm": 3.650310516357422,
+      "learning_rate": 3.2016809706108153e-06,
+      "loss": 0.9554,
+      "step": 55975
+    },
+    {
+      "epoch": 16.75,
+      "grad_norm": 1.8704103231430054,
+      "learning_rate": 3.1988049188424683e-06,
+      "loss": 1.0365,
+      "step": 55980
+    },
+    {
+      "epoch": 16.75,
+      "grad_norm": 1.763230562210083,
+      "learning_rate": 3.1959300711314727e-06,
+      "loss": 1.0271,
+      "step": 55985
+    },
+    {
+      "epoch": 16.75,
+      "grad_norm": 2.3225176334381104,
+      "learning_rate": 3.1930564276365842e-06,
+      "loss": 0.9846,
+      "step": 55990
+    },
+    {
+      "epoch": 16.75,
+      "grad_norm": 3.4815707206726074,
+      "learning_rate": 3.190183988516532e-06,
+      "loss": 0.9911,
+      "step": 55995
+    },
+    {
+      "epoch": 16.75,
+      "grad_norm": 2.6258528232574463,
+      "learning_rate": 3.187312753929936e-06,
+      "loss": 1.1604,
+      "step": 56000
+    },
+    {
+      "epoch": 16.76,
+      "grad_norm": 1.8864493370056152,
+      "learning_rate": 3.184442724035397e-06,
+      "loss": 1.023,
+      "step": 56005
+    },
+    {
+      "epoch": 16.76,
+      "grad_norm": 3.0884039402008057,
+      "learning_rate": 3.1815738989914022e-06,
+      "loss": 0.9337,
+      "step": 56010
+    },
+    {
+      "epoch": 16.76,
+      "grad_norm": 2.619140625,
+      "learning_rate": 3.178706278956403e-06,
+      "loss": 0.9569,
+      "step": 56015
+    },
+    {
+      "epoch": 16.76,
+      "grad_norm": 2.4161531925201416,
+      "learning_rate": 3.1758398640887753e-06,
+      "loss": 1.0311,
+      "step": 56020
+    },
+    {
+      "epoch": 16.76,
+      "grad_norm": 2.404909133911133,
+      "learning_rate": 3.172974654546826e-06,
+      "loss": 0.8572,
+      "step": 56025
+    },
+    {
+      "epoch": 16.76,
+      "grad_norm": 3.482473850250244,
+      "learning_rate": 3.1701106504888e-06,
+      "loss": 0.9448,
+      "step": 56030
+    },
+    {
+      "epoch": 16.77,
+      "grad_norm": 1.7155938148498535,
+      "learning_rate": 3.167247852072869e-06,
+      "loss": 0.8405,
+      "step": 56035
+    },
+    {
+      "epoch": 16.77,
+      "grad_norm": 5.21242094039917,
+      "learning_rate": 3.164386259457147e-06,
+      "loss": 0.999,
+      "step": 56040
+    },
+    {
+      "epoch": 16.77,
+      "grad_norm": 0.9593510627746582,
+      "learning_rate": 3.1615258727996812e-06,
+      "loss": 0.9095,
+      "step": 56045
+    },
+    {
+      "epoch": 16.77,
+      "grad_norm": 3.08491849899292,
+      "learning_rate": 3.15866669225843e-06,
+      "loss": 0.9706,
+      "step": 56050
+    },
+    {
+      "epoch": 16.77,
+      "grad_norm": 4.684256076812744,
+      "learning_rate": 3.1558087179913183e-06,
+      "loss": 0.9068,
+      "step": 56055
+    },
+    {
+      "epoch": 16.77,
+      "grad_norm": 1.6469522714614868,
+      "learning_rate": 3.1529519501561883e-06,
+      "loss": 0.8806,
+      "step": 56060
+    },
+    {
+      "epoch": 16.77,
+      "grad_norm": 2.585136651992798,
+      "learning_rate": 3.1500963889108113e-06,
+      "loss": 0.986,
+      "step": 56065
+    },
+    {
+      "epoch": 16.78,
+      "grad_norm": 1.8152649402618408,
+      "learning_rate": 3.1472420344129005e-06,
+      "loss": 0.9496,
+      "step": 56070
+    },
+    {
+      "epoch": 16.78,
+      "grad_norm": 4.420323371887207,
+      "learning_rate": 3.144388886820096e-06,
+      "loss": 1.0925,
+      "step": 56075
+    },
+    {
+      "epoch": 16.78,
+      "grad_norm": 1.4763715267181396,
+      "learning_rate": 3.1415369462899803e-06,
+      "loss": 1.0199,
+      "step": 56080
+    },
+    {
+      "epoch": 16.78,
+      "grad_norm": 5.959677219390869,
+      "learning_rate": 3.1386862129800494e-06,
+      "loss": 0.9485,
+      "step": 56085
+    },
+    {
+      "epoch": 16.78,
+      "grad_norm": 3.459981918334961,
+      "learning_rate": 3.135836687047766e-06,
+      "loss": 0.8902,
+      "step": 56090
+    },
+    {
+      "epoch": 16.78,
+      "grad_norm": 3.747342348098755,
+      "learning_rate": 3.132988368650486e-06,
+      "loss": 0.9345,
+      "step": 56095
+    },
+    {
+      "epoch": 16.78,
+      "grad_norm": 2.7175967693328857,
+      "learning_rate": 3.1301412579455373e-06,
+      "loss": 0.9902,
+      "step": 56100
+    },
+    {
+      "epoch": 16.79,
+      "grad_norm": 2.1608591079711914,
+      "learning_rate": 3.127295355090146e-06,
+      "loss": 0.9676,
+      "step": 56105
+    },
+    {
+      "epoch": 16.79,
+      "grad_norm": 4.2502336502075195,
+      "learning_rate": 3.124450660241507e-06,
+      "loss": 1.024,
+      "step": 56110
+    },
+    {
+      "epoch": 16.79,
+      "grad_norm": 3.1099443435668945,
+      "learning_rate": 3.121607173556712e-06,
+      "loss": 0.8469,
+      "step": 56115
+    },
+    {
+      "epoch": 16.79,
+      "grad_norm": 2.466075897216797,
+      "learning_rate": 3.1187648951928134e-06,
+      "loss": 0.8405,
+      "step": 56120
+    },
+    {
+      "epoch": 16.79,
+      "grad_norm": 1.5878918170928955,
+      "learning_rate": 3.115923825306785e-06,
+      "loss": 1.0979,
+      "step": 56125
+    },
+    {
+      "epoch": 16.79,
+      "grad_norm": 6.672008037567139,
+      "learning_rate": 3.113083964055538e-06,
+      "loss": 1.0864,
+      "step": 56130
+    },
+    {
+      "epoch": 16.79,
+      "grad_norm": 3.1133804321289062,
+      "learning_rate": 3.1102453115959106e-06,
+      "loss": 0.8562,
+      "step": 56135
+    },
+    {
+      "epoch": 16.8,
+      "grad_norm": 2.506171226501465,
+      "learning_rate": 3.1074078680846858e-06,
+      "loss": 0.8097,
+      "step": 56140
+    },
+    {
+      "epoch": 16.8,
+      "grad_norm": 3.1147594451904297,
+      "learning_rate": 3.104571633678563e-06,
+      "loss": 0.8618,
+      "step": 56145
+    },
+    {
+      "epoch": 16.8,
+      "grad_norm": 2.9119558334350586,
+      "learning_rate": 3.101736608534195e-06,
+      "loss": 0.9775,
+      "step": 56150
+    },
+    {
+      "epoch": 16.8,
+      "grad_norm": 2.8269081115722656,
+      "learning_rate": 3.098902792808148e-06,
+      "loss": 0.8513,
+      "step": 56155
+    },
+    {
+      "epoch": 16.8,
+      "grad_norm": 3.424445867538452,
+      "learning_rate": 3.096070186656938e-06,
+      "loss": 0.9278,
+      "step": 56160
+    },
+    {
+      "epoch": 16.8,
+      "grad_norm": 2.689589500427246,
+      "learning_rate": 3.0932387902370017e-06,
+      "loss": 0.9281,
+      "step": 56165
+    },
+    {
+      "epoch": 16.81,
+      "grad_norm": 2.5684525966644287,
+      "learning_rate": 3.090408603704717e-06,
+      "loss": 1.0208,
+      "step": 56170
+    },
+    {
+      "epoch": 16.81,
+      "grad_norm": 1.9679588079452515,
+      "learning_rate": 3.0875796272163972e-06,
+      "loss": 0.9015,
+      "step": 56175
+    },
+    {
+      "epoch": 16.81,
+      "grad_norm": 1.7874964475631714,
+      "learning_rate": 3.084751860928264e-06,
+      "loss": 1.0641,
+      "step": 56180
+    },
+    {
+      "epoch": 16.81,
+      "grad_norm": 3.177513599395752,
+      "learning_rate": 3.081925304996522e-06,
+      "loss": 1.0432,
+      "step": 56185
+    },
+    {
+      "epoch": 16.81,
+      "grad_norm": 3.6404199600219727,
+      "learning_rate": 3.0790999595772473e-06,
+      "loss": 0.9549,
+      "step": 56190
+    },
+    {
+      "epoch": 16.81,
+      "grad_norm": 11.20604419708252,
+      "learning_rate": 3.0762758248265107e-06,
+      "loss": 1.0167,
+      "step": 56195
+    },
+    {
+      "epoch": 16.81,
+      "grad_norm": 2.8811817169189453,
+      "learning_rate": 3.073452900900259e-06,
+      "loss": 0.8696,
+      "step": 56200
+    },
+    {
+      "epoch": 16.82,
+      "grad_norm": 3.1575052738189697,
+      "learning_rate": 3.070631187954423e-06,
+      "loss": 0.7686,
+      "step": 56205
+    },
+    {
+      "epoch": 16.82,
+      "grad_norm": 1.944756269454956,
+      "learning_rate": 3.067810686144826e-06,
+      "loss": 1.0258,
+      "step": 56210
+    },
+    {
+      "epoch": 16.82,
+      "grad_norm": 3.3357532024383545,
+      "learning_rate": 3.064991395627248e-06,
+      "loss": 0.8491,
+      "step": 56215
+    },
+    {
+      "epoch": 16.82,
+      "grad_norm": 4.347214221954346,
+      "learning_rate": 3.062173316557396e-06,
+      "loss": 0.9459,
+      "step": 56220
+    },
+    {
+      "epoch": 16.82,
+      "grad_norm": 2.0537803173065186,
+      "learning_rate": 3.0593564490909084e-06,
+      "loss": 0.9565,
+      "step": 56225
+    },
+    {
+      "epoch": 16.82,
+      "grad_norm": 2.428102493286133,
+      "learning_rate": 3.0565407933833586e-06,
+      "loss": 0.7407,
+      "step": 56230
+    },
+    {
+      "epoch": 16.82,
+      "grad_norm": 4.625239372253418,
+      "learning_rate": 3.0537263495902522e-06,
+      "loss": 0.9934,
+      "step": 56235
+    },
+    {
+      "epoch": 16.83,
+      "grad_norm": 1.8983107805252075,
+      "learning_rate": 3.050913117867027e-06,
+      "loss": 1.0148,
+      "step": 56240
+    },
+    {
+      "epoch": 16.83,
+      "grad_norm": 1.6342809200286865,
+      "learning_rate": 3.048101098369055e-06,
+      "loss": 1.0129,
+      "step": 56245
+    },
+    {
+      "epoch": 16.83,
+      "grad_norm": 1.8513437509536743,
+      "learning_rate": 3.0452902912516403e-06,
+      "loss": 0.855,
+      "step": 56250
+    },
+    {
+      "epoch": 16.83,
+      "grad_norm": 3.056840658187866,
+      "learning_rate": 3.0424806966700246e-06,
+      "loss": 0.9762,
+      "step": 56255
+    },
+    {
+      "epoch": 16.83,
+      "grad_norm": 9.707741737365723,
+      "learning_rate": 3.0396723147793737e-06,
+      "loss": 0.9429,
+      "step": 56260
+    },
+    {
+      "epoch": 16.83,
+      "grad_norm": 8.954275131225586,
+      "learning_rate": 3.036865145734796e-06,
+      "loss": 1.0287,
+      "step": 56265
+    },
+    {
+      "epoch": 16.84,
+      "grad_norm": 2.4749701023101807,
+      "learning_rate": 3.034059189691335e-06,
+      "loss": 1.0522,
+      "step": 56270
+    },
+    {
+      "epoch": 16.84,
+      "grad_norm": 3.8181324005126953,
+      "learning_rate": 3.0312544468039383e-06,
+      "loss": 0.9914,
+      "step": 56275
+    },
+    {
+      "epoch": 16.84,
+      "grad_norm": 6.021519660949707,
+      "learning_rate": 3.028450917227535e-06,
+      "loss": 0.7978,
+      "step": 56280
+    },
+    {
+      "epoch": 16.84,
+      "grad_norm": 2.9637115001678467,
+      "learning_rate": 3.02564860111694e-06,
+      "loss": 1.0615,
+      "step": 56285
+    },
+    {
+      "epoch": 16.84,
+      "grad_norm": 4.300103187561035,
+      "learning_rate": 3.0228474986269406e-06,
+      "loss": 0.7859,
+      "step": 56290
+    },
+    {
+      "epoch": 16.84,
+      "grad_norm": 1.1417111158370972,
+      "learning_rate": 3.0200476099122186e-06,
+      "loss": 1.0462,
+      "step": 56295
+    },
+    {
+      "epoch": 16.84,
+      "grad_norm": 1.8585431575775146,
+      "learning_rate": 3.0172489351274313e-06,
+      "loss": 1.1015,
+      "step": 56300
+    },
+    {
+      "epoch": 16.85,
+      "grad_norm": 1.4056024551391602,
+      "learning_rate": 3.0144514744271325e-06,
+      "loss": 1.066,
+      "step": 56305
+    },
+    {
+      "epoch": 16.85,
+      "grad_norm": 2.6178314685821533,
+      "learning_rate": 3.011655227965823e-06,
+      "loss": 1.0539,
+      "step": 56310
+    },
+    {
+      "epoch": 16.85,
+      "grad_norm": 4.641908645629883,
+      "learning_rate": 3.0088601958979416e-06,
+      "loss": 0.7809,
+      "step": 56315
+    },
+    {
+      "epoch": 16.85,
+      "grad_norm": 8.58937931060791,
+      "learning_rate": 3.006066378377853e-06,
+      "loss": 1.0699,
+      "step": 56320
+    },
+    {
+      "epoch": 16.85,
+      "grad_norm": 1.9301339387893677,
+      "learning_rate": 3.003273775559856e-06,
+      "loss": 0.9594,
+      "step": 56325
+    },
+    {
+      "epoch": 16.85,
+      "grad_norm": 3.1416869163513184,
+      "learning_rate": 3.0004823875981857e-06,
+      "loss": 0.8427,
+      "step": 56330
+    },
+    {
+      "epoch": 16.85,
+      "grad_norm": 1.7465052604675293,
+      "learning_rate": 2.997692214647005e-06,
+      "loss": 1.0006,
+      "step": 56335
+    },
+    {
+      "epoch": 16.86,
+      "grad_norm": 1.5967586040496826,
+      "learning_rate": 2.9949032568604125e-06,
+      "loss": 1.0248,
+      "step": 56340
+    },
+    {
+      "epoch": 16.86,
+      "grad_norm": 2.2583372592926025,
+      "learning_rate": 2.992115514392441e-06,
+      "loss": 1.0231,
+      "step": 56345
+    },
+    {
+      "epoch": 16.86,
+      "grad_norm": 4.796255111694336,
+      "learning_rate": 2.9893289873970528e-06,
+      "loss": 0.9619,
+      "step": 56350
+    },
+    {
+      "epoch": 16.86,
+      "grad_norm": 6.1636152267456055,
+      "learning_rate": 2.9865436760281445e-06,
+      "loss": 0.9783,
+      "step": 56355
+    },
+    {
+      "epoch": 16.86,
+      "grad_norm": 2.931490421295166,
+      "learning_rate": 2.983759580439549e-06,
+      "loss": 0.9008,
+      "step": 56360
+    },
+    {
+      "epoch": 16.86,
+      "grad_norm": 2.367924213409424,
+      "learning_rate": 2.980976700785035e-06,
+      "loss": 1.0938,
+      "step": 56365
+    },
+    {
+      "epoch": 16.87,
+      "grad_norm": 1.7075245380401611,
+      "learning_rate": 2.978195037218276e-06,
+      "loss": 1.027,
+      "step": 56370
+    },
+    {
+      "epoch": 16.87,
+      "grad_norm": 2.3319573402404785,
+      "learning_rate": 2.9754145898929275e-06,
+      "loss": 0.9178,
+      "step": 56375
+    },
+    {
+      "epoch": 16.87,
+      "grad_norm": 2.598543643951416,
+      "learning_rate": 2.972635358962525e-06,
+      "loss": 0.8888,
+      "step": 56380
+    },
+    {
+      "epoch": 16.87,
+      "grad_norm": 1.452913522720337,
+      "learning_rate": 2.9698573445805873e-06,
+      "loss": 1.0465,
+      "step": 56385
+    },
+    {
+      "epoch": 16.87,
+      "grad_norm": 6.2622833251953125,
+      "learning_rate": 2.967080546900519e-06,
+      "loss": 0.9591,
+      "step": 56390
+    },
+    {
+      "epoch": 16.87,
+      "grad_norm": 2.436224937438965,
+      "learning_rate": 2.964304966075701e-06,
+      "loss": 0.935,
+      "step": 56395
+    },
+    {
+      "epoch": 16.87,
+      "grad_norm": 2.5119998455047607,
+      "learning_rate": 2.9615306022594104e-06,
+      "loss": 0.9041,
+      "step": 56400
+    },
+    {
+      "epoch": 16.88,
+      "grad_norm": 2.9698565006256104,
+      "learning_rate": 2.958757455604874e-06,
+      "loss": 1.0233,
+      "step": 56405
+    },
+    {
+      "epoch": 16.88,
+      "grad_norm": 1.585296630859375,
+      "learning_rate": 2.9559855262652564e-06,
+      "loss": 1.0708,
+      "step": 56410
+    },
+    {
+      "epoch": 16.88,
+      "grad_norm": 2.456385850906372,
+      "learning_rate": 2.9532148143936426e-06,
+      "loss": 0.8779,
+      "step": 56415
+    },
+    {
+      "epoch": 16.88,
+      "grad_norm": 2.774512767791748,
+      "learning_rate": 2.950445320143058e-06,
+      "loss": 1.0221,
+      "step": 56420
+    },
+    {
+      "epoch": 16.88,
+      "grad_norm": 1.0109182596206665,
+      "learning_rate": 2.9476770436664582e-06,
+      "loss": 0.9762,
+      "step": 56425
+    },
+    {
+      "epoch": 16.88,
+      "grad_norm": 1.9075700044631958,
+      "learning_rate": 2.9449099851167365e-06,
+      "loss": 0.8601,
+      "step": 56430
+    },
+    {
+      "epoch": 16.88,
+      "grad_norm": 1.4174439907073975,
+      "learning_rate": 2.9421441446467078e-06,
+      "loss": 1.0847,
+      "step": 56435
+    },
+    {
+      "epoch": 16.89,
+      "grad_norm": 2.0944745540618896,
+      "learning_rate": 2.9393795224091332e-06,
+      "loss": 0.7858,
+      "step": 56440
+    },
+    {
+      "epoch": 16.89,
+      "grad_norm": 3.1148488521575928,
+      "learning_rate": 2.9366161185566952e-06,
+      "loss": 1.08,
+      "step": 56445
+    },
+    {
+      "epoch": 16.89,
+      "grad_norm": 2.154017448425293,
+      "learning_rate": 2.9338539332420143e-06,
+      "loss": 1.025,
+      "step": 56450
+    },
+    {
+      "epoch": 16.89,
+      "grad_norm": 1.8690617084503174,
+      "learning_rate": 2.931092966617646e-06,
+      "loss": 1.0529,
+      "step": 56455
+    },
+    {
+      "epoch": 16.89,
+      "grad_norm": 1.549531102180481,
+      "learning_rate": 2.928333218836074e-06,
+      "loss": 0.8698,
+      "step": 56460
+    },
+    {
+      "epoch": 16.89,
+      "grad_norm": 4.329528331756592,
+      "learning_rate": 2.9255746900497148e-06,
+      "loss": 0.8416,
+      "step": 56465
+    },
+    {
+      "epoch": 16.9,
+      "grad_norm": 4.210263729095459,
+      "learning_rate": 2.9228173804109276e-06,
+      "loss": 1.1474,
+      "step": 56470
+    },
+    {
+      "epoch": 16.9,
+      "grad_norm": 2.099457263946533,
+      "learning_rate": 2.9200612900719764e-06,
+      "loss": 1.1059,
+      "step": 56475
+    },
+    {
+      "epoch": 16.9,
+      "grad_norm": 4.597415924072266,
+      "learning_rate": 2.917306419185098e-06,
+      "loss": 0.8395,
+      "step": 56480
+    },
+    {
+      "epoch": 16.9,
+      "grad_norm": 2.6783969402313232,
+      "learning_rate": 2.914552767902426e-06,
+      "loss": 1.0578,
+      "step": 56485
+    },
+    {
+      "epoch": 16.9,
+      "grad_norm": 1.4600439071655273,
+      "learning_rate": 2.9118003363760553e-06,
+      "loss": 0.8978,
+      "step": 56490
+    },
+    {
+      "epoch": 16.9,
+      "grad_norm": 3.4358808994293213,
+      "learning_rate": 2.90904912475799e-06,
+      "loss": 0.9432,
+      "step": 56495
+    },
+    {
+      "epoch": 16.9,
+      "grad_norm": 2.708552360534668,
+      "learning_rate": 2.906299133200177e-06,
+      "loss": 0.9166,
+      "step": 56500
+    },
+    {
+      "epoch": 16.91,
+      "grad_norm": 2.577993869781494,
+      "learning_rate": 2.9035503618544986e-06,
+      "loss": 1.0563,
+      "step": 56505
+    },
+    {
+      "epoch": 16.91,
+      "grad_norm": 2.4056396484375,
+      "learning_rate": 2.900802810872766e-06,
+      "loss": 1.03,
+      "step": 56510
+    },
+    {
+      "epoch": 16.91,
+      "grad_norm": 4.068645000457764,
+      "learning_rate": 2.898056480406722e-06,
+      "loss": 0.8931,
+      "step": 56515
+    },
+    {
+      "epoch": 16.91,
+      "grad_norm": 2.8915958404541016,
+      "learning_rate": 2.8953113706080477e-06,
+      "loss": 1.1591,
+      "step": 56520
+    },
+    {
+      "epoch": 16.91,
+      "grad_norm": 1.79038667678833,
+      "learning_rate": 2.892567481628347e-06,
+      "loss": 1.0253,
+      "step": 56525
+    },
+    {
+      "epoch": 16.91,
+      "grad_norm": 2.4151062965393066,
+      "learning_rate": 2.889824813619166e-06,
+      "loss": 0.8595,
+      "step": 56530
+    },
+    {
+      "epoch": 16.91,
+      "grad_norm": 1.7418419122695923,
+      "learning_rate": 2.8870833667319795e-06,
+      "loss": 0.8801,
+      "step": 56535
+    },
+    {
+      "epoch": 16.92,
+      "grad_norm": 4.192162036895752,
+      "learning_rate": 2.8843431411181926e-06,
+      "loss": 1.1202,
+      "step": 56540
+    },
+    {
+      "epoch": 16.92,
+      "grad_norm": 4.1534881591796875,
+      "learning_rate": 2.8816041369291446e-06,
+      "loss": 0.8828,
+      "step": 56545
+    },
+    {
+      "epoch": 16.92,
+      "grad_norm": 2.742586374282837,
+      "learning_rate": 2.878866354316112e-06,
+      "loss": 0.9694,
+      "step": 56550
+    },
+    {
+      "epoch": 16.92,
+      "grad_norm": 3.240999460220337,
+      "learning_rate": 2.8761297934302934e-06,
+      "loss": 0.8857,
+      "step": 56555
+    },
+    {
+      "epoch": 16.92,
+      "grad_norm": 4.671759128570557,
+      "learning_rate": 2.873394454422834e-06,
+      "loss": 0.9096,
+      "step": 56560
+    },
+    {
+      "epoch": 16.92,
+      "grad_norm": 2.9793546199798584,
+      "learning_rate": 2.8706603374448027e-06,
+      "loss": 1.1287,
+      "step": 56565
+    },
+    {
+      "epoch": 16.93,
+      "grad_norm": 1.7608052492141724,
+      "learning_rate": 2.8679274426471864e-06,
+      "loss": 0.9546,
+      "step": 56570
+    },
+    {
+      "epoch": 16.93,
+      "grad_norm": 3.5491206645965576,
+      "learning_rate": 2.865195770180945e-06,
+      "loss": 0.9436,
+      "step": 56575
+    },
+    {
+      "epoch": 16.93,
+      "grad_norm": 4.213696002960205,
+      "learning_rate": 2.8624653201969247e-06,
+      "loss": 0.8552,
+      "step": 56580
+    },
+    {
+      "epoch": 16.93,
+      "grad_norm": 2.7237231731414795,
+      "learning_rate": 2.859736092845941e-06,
+      "loss": 1.0671,
+      "step": 56585
+    },
+    {
+      "epoch": 16.93,
+      "grad_norm": 2.797983407974243,
+      "learning_rate": 2.8570080882787175e-06,
+      "loss": 1.0807,
+      "step": 56590
+    },
+    {
+      "epoch": 16.93,
+      "grad_norm": 4.629161834716797,
+      "learning_rate": 2.8542813066459173e-06,
+      "loss": 0.9689,
+      "step": 56595
+    },
+    {
+      "epoch": 16.93,
+      "grad_norm": 4.058006763458252,
+      "learning_rate": 2.8515557480981448e-06,
+      "loss": 1.0349,
+      "step": 56600
+    },
+    {
+      "epoch": 16.94,
+      "grad_norm": 2.309382677078247,
+      "learning_rate": 2.848831412785924e-06,
+      "loss": 0.9884,
+      "step": 56605
+    },
+    {
+      "epoch": 16.94,
+      "grad_norm": 4.170457363128662,
+      "learning_rate": 2.8461083008597204e-06,
+      "loss": 0.8189,
+      "step": 56610
+    },
+    {
+      "epoch": 16.94,
+      "grad_norm": 2.4919357299804688,
+      "learning_rate": 2.8433864124699284e-06,
+      "loss": 0.9594,
+      "step": 56615
+    },
+    {
+      "epoch": 16.94,
+      "grad_norm": 4.131077289581299,
+      "learning_rate": 2.840665747766874e-06,
+      "loss": 0.9989,
+      "step": 56620
+    },
+    {
+      "epoch": 16.94,
+      "grad_norm": 4.024702072143555,
+      "learning_rate": 2.8379463069008153e-06,
+      "loss": 0.9219,
+      "step": 56625
+    },
+    {
+      "epoch": 16.94,
+      "grad_norm": 2.887530565261841,
+      "learning_rate": 2.8352280900219462e-06,
+      "loss": 0.8609,
+      "step": 56630
+    },
+    {
+      "epoch": 16.94,
+      "grad_norm": 3.2231643199920654,
+      "learning_rate": 2.8325110972803936e-06,
+      "loss": 1.0832,
+      "step": 56635
+    },
+    {
+      "epoch": 16.95,
+      "grad_norm": 2.5507965087890625,
+      "learning_rate": 2.829795328826207e-06,
+      "loss": 1.023,
+      "step": 56640
+    },
+    {
+      "epoch": 16.95,
+      "grad_norm": 1.9942423105239868,
+      "learning_rate": 2.827080784809383e-06,
+      "loss": 1.0267,
+      "step": 56645
+    },
+    {
+      "epoch": 16.95,
+      "grad_norm": 3.263000011444092,
+      "learning_rate": 2.8243674653798376e-06,
+      "loss": 0.9424,
+      "step": 56650
+    },
+    {
+      "epoch": 16.95,
+      "grad_norm": 2.1129324436187744,
+      "learning_rate": 2.8216553706874288e-06,
+      "loss": 1.119,
+      "step": 56655
+    },
+    {
+      "epoch": 16.95,
+      "grad_norm": 1.207990050315857,
+      "learning_rate": 2.8189445008819454e-06,
+      "loss": 0.762,
+      "step": 56660
+    },
+    {
+      "epoch": 16.95,
+      "grad_norm": 2.738921880722046,
+      "learning_rate": 2.8162348561130895e-06,
+      "loss": 1.012,
+      "step": 56665
+    },
+    {
+      "epoch": 16.96,
+      "grad_norm": 4.169528007507324,
+      "learning_rate": 2.8135264365305366e-06,
+      "loss": 1.0023,
+      "step": 56670
+    },
+    {
+      "epoch": 16.96,
+      "grad_norm": 3.962857723236084,
+      "learning_rate": 2.8108192422838437e-06,
+      "loss": 0.901,
+      "step": 56675
+    },
+    {
+      "epoch": 16.96,
+      "grad_norm": 2.816699743270874,
+      "learning_rate": 2.808113273522553e-06,
+      "loss": 0.969,
+      "step": 56680
+    },
+    {
+      "epoch": 16.96,
+      "grad_norm": 4.528411388397217,
+      "learning_rate": 2.8054085303960863e-06,
+      "loss": 1.0662,
+      "step": 56685
+    },
+    {
+      "epoch": 16.96,
+      "grad_norm": 1.3069534301757812,
+      "learning_rate": 2.8027050130538494e-06,
+      "loss": 1.0393,
+      "step": 56690
+    },
+    {
+      "epoch": 16.96,
+      "grad_norm": 5.835165977478027,
+      "learning_rate": 2.800002721645134e-06,
+      "loss": 1.1026,
+      "step": 56695
+    },
+    {
+      "epoch": 16.96,
+      "grad_norm": 2.2448999881744385,
+      "learning_rate": 2.79730165631919e-06,
+      "loss": 0.8728,
+      "step": 56700
+    },
+    {
+      "epoch": 16.97,
+      "grad_norm": 3.5195298194885254,
+      "learning_rate": 2.7946018172252066e-06,
+      "loss": 1.0797,
+      "step": 56705
+    },
+    {
+      "epoch": 16.97,
+      "grad_norm": 1.8925422430038452,
+      "learning_rate": 2.7919032045122725e-06,
+      "loss": 0.9078,
+      "step": 56710
+    },
+    {
+      "epoch": 16.97,
+      "grad_norm": 1.31459379196167,
+      "learning_rate": 2.789205818329449e-06,
+      "loss": 0.9306,
+      "step": 56715
+    },
+    {
+      "epoch": 16.97,
+      "grad_norm": 3.2655656337738037,
+      "learning_rate": 2.786509658825698e-06,
+      "loss": 1.1181,
+      "step": 56720
+    },
+    {
+      "epoch": 16.97,
+      "grad_norm": 1.9709197282791138,
+      "learning_rate": 2.7838147261499305e-06,
+      "loss": 0.8847,
+      "step": 56725
+    },
+    {
+      "epoch": 16.97,
+      "grad_norm": 2.601952314376831,
+      "learning_rate": 2.78112102045098e-06,
+      "loss": 0.911,
+      "step": 56730
+    },
+    {
+      "epoch": 16.97,
+      "grad_norm": 1.5735349655151367,
+      "learning_rate": 2.7784285418776227e-06,
+      "loss": 0.8787,
+      "step": 56735
+    },
+    {
+      "epoch": 16.98,
+      "grad_norm": 1.383877158164978,
+      "learning_rate": 2.775737290578559e-06,
+      "loss": 0.8761,
+      "step": 56740
+    },
+    {
+      "epoch": 16.98,
+      "grad_norm": 2.030069351196289,
+      "learning_rate": 2.773047266702422e-06,
+      "loss": 1.0923,
+      "step": 56745
+    },
+    {
+      "epoch": 16.98,
+      "grad_norm": 2.9204518795013428,
+      "learning_rate": 2.7703584703977824e-06,
+      "loss": 0.9418,
+      "step": 56750
+    },
+    {
+      "epoch": 16.98,
+      "grad_norm": 2.140024423599243,
+      "learning_rate": 2.7676709018131436e-06,
+      "loss": 0.9529,
+      "step": 56755
+    },
+    {
+      "epoch": 16.98,
+      "grad_norm": 1.8030258417129517,
+      "learning_rate": 2.76498456109692e-06,
+      "loss": 1.0081,
+      "step": 56760
+    },
+    {
+      "epoch": 16.98,
+      "grad_norm": 4.3402509689331055,
+      "learning_rate": 2.7622994483974985e-06,
+      "loss": 0.8433,
+      "step": 56765
+    },
+    {
+      "epoch": 16.98,
+      "grad_norm": 2.835216999053955,
+      "learning_rate": 2.759615563863152e-06,
+      "loss": 1.0517,
+      "step": 56770
+    },
+    {
+      "epoch": 16.99,
+      "grad_norm": 2.009197473526001,
+      "learning_rate": 2.7569329076421317e-06,
+      "loss": 0.985,
+      "step": 56775
+    },
+    {
+      "epoch": 16.99,
+      "grad_norm": 1.9505784511566162,
+      "learning_rate": 2.7542514798825773e-06,
+      "loss": 0.8046,
+      "step": 56780
+    },
+    {
+      "epoch": 16.99,
+      "grad_norm": 1.5446016788482666,
+      "learning_rate": 2.7515712807325955e-06,
+      "loss": 0.9255,
+      "step": 56785
+    },
+    {
+      "epoch": 16.99,
+      "grad_norm": 1.5326980352401733,
+      "learning_rate": 2.7488923103402094e-06,
+      "loss": 1.1662,
+      "step": 56790
+    },
+    {
+      "epoch": 16.99,
+      "grad_norm": 2.3964767456054688,
+      "learning_rate": 2.7462145688533615e-06,
+      "loss": 0.9944,
+      "step": 56795
+    },
+    {
+      "epoch": 16.99,
+      "grad_norm": 5.0350189208984375,
+      "learning_rate": 2.743538056419964e-06,
+      "loss": 0.9068,
+      "step": 56800
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 2.6177234649658203,
+      "learning_rate": 2.7408627731878133e-06,
+      "loss": 0.9081,
+      "step": 56805
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 2.825147867202759,
+      "learning_rate": 2.738188719304685e-06,
+      "loss": 0.9299,
+      "step": 56810
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 3.711740732192993,
+      "learning_rate": 2.735515894918245e-06,
+      "loss": 0.9777,
+      "step": 56815
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 2.3973779678344727,
+      "learning_rate": 2.7328443001761266e-06,
+      "loss": 1.0681,
+      "step": 56820
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 3.3198635578155518,
+      "learning_rate": 2.7301739352258716e-06,
+      "loss": 0.9094,
+      "step": 56825
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 2.2214438915252686,
+      "learning_rate": 2.7275048002149585e-06,
+      "loss": 0.9812,
+      "step": 56830
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 6.447070598602295,
+      "learning_rate": 2.7248368952908053e-06,
+      "loss": 0.76,
+      "step": 56835
+    },
+    {
+      "epoch": 17.01,
+      "grad_norm": 4.4163713455200195,
+      "learning_rate": 2.722170220600756e-06,
+      "loss": 0.8238,
+      "step": 56840
+    },
+    {
+      "epoch": 17.01,
+      "grad_norm": 2.432135820388794,
+      "learning_rate": 2.7195047762920895e-06,
+      "loss": 0.8192,
+      "step": 56845
+    },
+    {
+      "epoch": 17.01,
+      "grad_norm": 1.874724268913269,
+      "learning_rate": 2.716840562512021e-06,
+      "loss": 0.916,
+      "step": 56850
+    },
+    {
+      "epoch": 17.01,
+      "grad_norm": 3.546983003616333,
+      "learning_rate": 2.714177579407673e-06,
+      "loss": 0.858,
+      "step": 56855
+    },
+    {
+      "epoch": 17.01,
+      "grad_norm": 3.305762529373169,
+      "learning_rate": 2.7115158271261403e-06,
+      "loss": 0.9272,
+      "step": 56860
+    },
+    {
+      "epoch": 17.01,
+      "grad_norm": 4.167823791503906,
+      "learning_rate": 2.70885530581442e-06,
+      "loss": 0.9225,
+      "step": 56865
+    },
+    {
+      "epoch": 17.01,
+      "grad_norm": 2.542015790939331,
+      "learning_rate": 2.7061960156194526e-06,
+      "loss": 0.972,
+      "step": 56870
+    },
+    {
+      "epoch": 17.02,
+      "grad_norm": 2.959282398223877,
+      "learning_rate": 2.7035379566881042e-06,
+      "loss": 1.1611,
+      "step": 56875
+    },
+    {
+      "epoch": 17.02,
+      "grad_norm": 2.539849281311035,
+      "learning_rate": 2.700881129167179e-06,
+      "loss": 0.8913,
+      "step": 56880
+    },
+    {
+      "epoch": 17.02,
+      "grad_norm": 3.6170997619628906,
+      "learning_rate": 2.6982255332034133e-06,
+      "loss": 1.0476,
+      "step": 56885
+    },
+    {
+      "epoch": 17.02,
+      "grad_norm": 4.399633407592773,
+      "learning_rate": 2.6955711689434614e-06,
+      "loss": 0.7257,
+      "step": 56890
+    },
+    {
+      "epoch": 17.02,
+      "grad_norm": 2.9377009868621826,
+      "learning_rate": 2.6929180365339423e-06,
+      "loss": 1.2028,
+      "step": 56895
+    },
+    {
+      "epoch": 17.02,
+      "grad_norm": 3.030989408493042,
+      "learning_rate": 2.690266136121361e-06,
+      "loss": 1.0568,
+      "step": 56900
+    },
+    {
+      "epoch": 17.03,
+      "grad_norm": 2.5369529724121094,
+      "learning_rate": 2.6876154678522005e-06,
+      "loss": 1.0388,
+      "step": 56905
+    },
+    {
+      "epoch": 17.03,
+      "grad_norm": 2.8671159744262695,
+      "learning_rate": 2.6849660318728376e-06,
+      "loss": 0.9034,
+      "step": 56910
+    },
+    {
+      "epoch": 17.03,
+      "grad_norm": 1.3632619380950928,
+      "learning_rate": 2.682317828329614e-06,
+      "loss": 0.735,
+      "step": 56915
+    },
+    {
+      "epoch": 17.03,
+      "grad_norm": 2.6106884479522705,
+      "learning_rate": 2.679670857368774e-06,
+      "loss": 0.9862,
+      "step": 56920
+    },
+    {
+      "epoch": 17.03,
+      "grad_norm": 2.3712689876556396,
+      "learning_rate": 2.6770251191365113e-06,
+      "loss": 1.0196,
+      "step": 56925
+    },
+    {
+      "epoch": 17.03,
+      "grad_norm": 5.395451545715332,
+      "learning_rate": 2.6743806137789485e-06,
+      "loss": 1.0626,
+      "step": 56930
+    },
+    {
+      "epoch": 17.03,
+      "grad_norm": 4.0020856857299805,
+      "learning_rate": 2.6717373414421377e-06,
+      "loss": 0.9506,
+      "step": 56935
+    },
+    {
+      "epoch": 17.04,
+      "grad_norm": 1.977827548980713,
+      "learning_rate": 2.669095302272065e-06,
+      "loss": 0.7509,
+      "step": 56940
+    },
+    {
+      "epoch": 17.04,
+      "grad_norm": 4.007442474365234,
+      "learning_rate": 2.666454496414647e-06,
+      "loss": 0.9768,
+      "step": 56945
+    },
+    {
+      "epoch": 17.04,
+      "grad_norm": 2.781235933303833,
+      "learning_rate": 2.6638149240157313e-06,
+      "loss": 0.9895,
+      "step": 56950
+    },
+    {
+      "epoch": 17.04,
+      "grad_norm": 3.7630717754364014,
+      "learning_rate": 2.6611765852211037e-06,
+      "loss": 1.0876,
+      "step": 56955
+    },
+    {
+      "epoch": 17.04,
+      "grad_norm": 6.746951580047607,
+      "learning_rate": 2.6585394801764697e-06,
+      "loss": 0.9619,
+      "step": 56960
+    },
+    {
+      "epoch": 17.04,
+      "grad_norm": 1.971835732460022,
+      "learning_rate": 2.6559036090274796e-06,
+      "loss": 1.147,
+      "step": 56965
+    },
+    {
+      "epoch": 17.04,
+      "grad_norm": 1.7732189893722534,
+      "learning_rate": 2.653268971919706e-06,
+      "loss": 1.0708,
+      "step": 56970
+    },
+    {
+      "epoch": 17.05,
+      "grad_norm": 2.1703782081604004,
+      "learning_rate": 2.6506355689986607e-06,
+      "loss": 0.9244,
+      "step": 56975
+    },
+    {
+      "epoch": 17.05,
+      "grad_norm": 3.1191794872283936,
+      "learning_rate": 2.6480034004097877e-06,
+      "loss": 0.9953,
+      "step": 56980
+    },
+    {
+      "epoch": 17.05,
+      "grad_norm": 3.8292319774627686,
+      "learning_rate": 2.645372466298443e-06,
+      "loss": 0.9508,
+      "step": 56985
+    },
+    {
+      "epoch": 17.05,
+      "grad_norm": 2.0366878509521484,
+      "learning_rate": 2.64274276680995e-06,
+      "loss": 0.9986,
+      "step": 56990
+    },
+    {
+      "epoch": 17.05,
+      "grad_norm": 2.463479518890381,
+      "learning_rate": 2.6401143020895245e-06,
+      "loss": 1.0819,
+      "step": 56995
+    },
+    {
+      "epoch": 17.05,
+      "grad_norm": 1.6114779710769653,
+      "learning_rate": 2.637487072282355e-06,
+      "loss": 1.0996,
+      "step": 57000
+    },
+    {
+      "epoch": 17.06,
+      "grad_norm": 1.567771315574646,
+      "learning_rate": 2.6348610775335182e-06,
+      "loss": 1.001,
+      "step": 57005
+    },
+    {
+      "epoch": 17.06,
+      "grad_norm": 1.417826533317566,
+      "learning_rate": 2.632236317988068e-06,
+      "loss": 0.9204,
+      "step": 57010
+    },
+    {
+      "epoch": 17.06,
+      "grad_norm": 1.9809144735336304,
+      "learning_rate": 2.6296127937909504e-06,
+      "loss": 0.928,
+      "step": 57015
+    },
+    {
+      "epoch": 17.06,
+      "grad_norm": 2.4139678478240967,
+      "learning_rate": 2.626990505087068e-06,
+      "loss": 0.8989,
+      "step": 57020
+    },
+    {
+      "epoch": 17.06,
+      "grad_norm": 2.6247403621673584,
+      "learning_rate": 2.6243694520212414e-06,
+      "loss": 0.9612,
+      "step": 57025
+    },
+    {
+      "epoch": 17.06,
+      "grad_norm": 4.078159332275391,
+      "learning_rate": 2.6217496347382325e-06,
+      "loss": 1.0525,
+      "step": 57030
+    },
+    {
+      "epoch": 17.06,
+      "grad_norm": 3.896151065826416,
+      "learning_rate": 2.6191310533827313e-06,
+      "loss": 1.1528,
+      "step": 57035
+    },
+    {
+      "epoch": 17.07,
+      "grad_norm": 2.571683883666992,
+      "learning_rate": 2.616513708099358e-06,
+      "loss": 1.0413,
+      "step": 57040
+    },
+    {
+      "epoch": 17.07,
+      "grad_norm": 3.916944980621338,
+      "learning_rate": 2.613897599032666e-06,
+      "loss": 0.8858,
+      "step": 57045
+    },
+    {
+      "epoch": 17.07,
+      "grad_norm": 3.300248622894287,
+      "learning_rate": 2.611282726327141e-06,
+      "loss": 1.1224,
+      "step": 57050
+    },
+    {
+      "epoch": 17.07,
+      "grad_norm": 1.99065363407135,
+      "learning_rate": 2.6086690901272e-06,
+      "loss": 1.035,
+      "step": 57055
+    },
+    {
+      "epoch": 17.07,
+      "grad_norm": 2.4363794326782227,
+      "learning_rate": 2.606056690577191e-06,
+      "loss": 0.926,
+      "step": 57060
+    },
+    {
+      "epoch": 17.07,
+      "grad_norm": 1.8297630548477173,
+      "learning_rate": 2.603445527821391e-06,
+      "loss": 1.1107,
+      "step": 57065
+    },
+    {
+      "epoch": 17.07,
+      "grad_norm": 4.487589359283447,
+      "learning_rate": 2.6008356020040174e-06,
+      "loss": 1.0452,
+      "step": 57070
+    },
+    {
+      "epoch": 17.08,
+      "grad_norm": 4.226923942565918,
+      "learning_rate": 2.598226913269214e-06,
+      "loss": 0.8234,
+      "step": 57075
+    },
+    {
+      "epoch": 17.08,
+      "grad_norm": 2.6841580867767334,
+      "learning_rate": 2.595619461761045e-06,
+      "loss": 0.735,
+      "step": 57080
+    },
+    {
+      "epoch": 17.08,
+      "grad_norm": 4.622152805328369,
+      "learning_rate": 2.5930132476235326e-06,
+      "loss": 0.9079,
+      "step": 57085
+    },
+    {
+      "epoch": 17.08,
+      "grad_norm": 2.2728214263916016,
+      "learning_rate": 2.5904082710005995e-06,
+      "loss": 1.0287,
+      "step": 57090
+    },
+    {
+      "epoch": 17.08,
+      "grad_norm": 3.7428131103515625,
+      "learning_rate": 2.587804532036134e-06,
+      "loss": 0.8489,
+      "step": 57095
+    },
+    {
+      "epoch": 17.08,
+      "grad_norm": 5.1959099769592285,
+      "learning_rate": 2.585202030873915e-06,
+      "loss": 0.9871,
+      "step": 57100
+    },
+    {
+      "epoch": 17.09,
+      "grad_norm": 9.105627059936523,
+      "learning_rate": 2.582600767657703e-06,
+      "loss": 0.9979,
+      "step": 57105
+    },
+    {
+      "epoch": 17.09,
+      "grad_norm": 3.824483633041382,
+      "learning_rate": 2.580000742531141e-06,
+      "loss": 0.9031,
+      "step": 57110
+    },
+    {
+      "epoch": 17.09,
+      "grad_norm": 3.2056384086608887,
+      "learning_rate": 2.577401955637837e-06,
+      "loss": 0.9565,
+      "step": 57115
+    },
+    {
+      "epoch": 17.09,
+      "grad_norm": 1.918878197669983,
+      "learning_rate": 2.574804407121312e-06,
+      "loss": 1.0012,
+      "step": 57120
+    },
+    {
+      "epoch": 17.09,
+      "grad_norm": 2.389338731765747,
+      "learning_rate": 2.572208097125034e-06,
+      "loss": 1.0327,
+      "step": 57125
+    },
+    {
+      "epoch": 17.09,
+      "grad_norm": 1.8829810619354248,
+      "learning_rate": 2.5696130257923863e-06,
+      "loss": 0.9748,
+      "step": 57130
+    },
+    {
+      "epoch": 17.09,
+      "grad_norm": 2.50826358795166,
+      "learning_rate": 2.5670191932666983e-06,
+      "loss": 0.901,
+      "step": 57135
+    },
+    {
+      "epoch": 17.1,
+      "grad_norm": 3.3760509490966797,
+      "learning_rate": 2.5644265996912246e-06,
+      "loss": 1.0282,
+      "step": 57140
+    },
+    {
+      "epoch": 17.1,
+      "grad_norm": 2.7886831760406494,
+      "learning_rate": 2.561835245209146e-06,
+      "loss": 0.8485,
+      "step": 57145
+    },
+    {
+      "epoch": 17.1,
+      "grad_norm": 1.582168698310852,
+      "learning_rate": 2.559245129963586e-06,
+      "loss": 0.9798,
+      "step": 57150
+    },
+    {
+      "epoch": 17.1,
+      "grad_norm": 2.845203161239624,
+      "learning_rate": 2.556656254097589e-06,
+      "loss": 1.2068,
+      "step": 57155
+    },
+    {
+      "epoch": 17.1,
+      "grad_norm": 2.8991782665252686,
+      "learning_rate": 2.5540686177541408e-06,
+      "loss": 1.0069,
+      "step": 57160
+    },
+    {
+      "epoch": 17.1,
+      "grad_norm": 2.5524137020111084,
+      "learning_rate": 2.5514822210761485e-06,
+      "loss": 0.7097,
+      "step": 57165
+    },
+    {
+      "epoch": 17.1,
+      "grad_norm": 5.034450531005859,
+      "learning_rate": 2.5488970642064624e-06,
+      "loss": 0.9645,
+      "step": 57170
+    },
+    {
+      "epoch": 17.11,
+      "grad_norm": 3.3670802116394043,
+      "learning_rate": 2.5463131472878544e-06,
+      "loss": 1.0891,
+      "step": 57175
+    },
+    {
+      "epoch": 17.11,
+      "grad_norm": 5.955173015594482,
+      "learning_rate": 2.543730470463035e-06,
+      "loss": 0.9105,
+      "step": 57180
+    },
+    {
+      "epoch": 17.11,
+      "grad_norm": 2.0601420402526855,
+      "learning_rate": 2.541149033874632e-06,
+      "loss": 1.0384,
+      "step": 57185
+    },
+    {
+      "epoch": 17.11,
+      "grad_norm": 1.1754088401794434,
+      "learning_rate": 2.538568837665231e-06,
+      "loss": 0.8901,
+      "step": 57190
+    },
+    {
+      "epoch": 17.11,
+      "grad_norm": 6.220560550689697,
+      "learning_rate": 2.535989881977319e-06,
+      "loss": 0.9757,
+      "step": 57195
+    },
+    {
+      "epoch": 17.11,
+      "grad_norm": 2.9648501873016357,
+      "learning_rate": 2.5334121669533416e-06,
+      "loss": 0.9412,
+      "step": 57200
+    },
+    {
+      "epoch": 17.12,
+      "grad_norm": 2.4804067611694336,
+      "learning_rate": 2.5308356927356556e-06,
+      "loss": 1.1147,
+      "step": 57205
+    },
+    {
+      "epoch": 17.12,
+      "grad_norm": 2.2306246757507324,
+      "learning_rate": 2.5282604594665583e-06,
+      "loss": 0.9197,
+      "step": 57210
+    },
+    {
+      "epoch": 17.12,
+      "grad_norm": 1.3986696004867554,
+      "learning_rate": 2.525686467288277e-06,
+      "loss": 0.9156,
+      "step": 57215
+    },
+    {
+      "epoch": 17.12,
+      "grad_norm": 2.3900442123413086,
+      "learning_rate": 2.523113716342973e-06,
+      "loss": 0.9889,
+      "step": 57220
+    },
+    {
+      "epoch": 17.12,
+      "grad_norm": 6.6008100509643555,
+      "learning_rate": 2.520542206772733e-06,
+      "loss": 0.9103,
+      "step": 57225
+    },
+    {
+      "epoch": 17.12,
+      "grad_norm": 2.302074909210205,
+      "learning_rate": 2.517971938719582e-06,
+      "loss": 0.7901,
+      "step": 57230
+    },
+    {
+      "epoch": 17.12,
+      "grad_norm": 2.978907585144043,
+      "learning_rate": 2.5154029123254708e-06,
+      "loss": 1.1318,
+      "step": 57235
+    },
+    {
+      "epoch": 17.13,
+      "grad_norm": 3.265174388885498,
+      "learning_rate": 2.5128351277322853e-06,
+      "loss": 0.8781,
+      "step": 57240
+    },
+    {
+      "epoch": 17.13,
+      "grad_norm": 2.929891347885132,
+      "learning_rate": 2.510268585081843e-06,
+      "loss": 0.8406,
+      "step": 57245
+    },
+    {
+      "epoch": 17.13,
+      "grad_norm": 2.2917799949645996,
+      "learning_rate": 2.5077032845158886e-06,
+      "loss": 0.9925,
+      "step": 57250
+    },
+    {
+      "epoch": 17.13,
+      "grad_norm": 6.078389644622803,
+      "learning_rate": 2.505139226176104e-06,
+      "loss": 0.8077,
+      "step": 57255
+    },
+    {
+      "epoch": 17.13,
+      "grad_norm": 3.7742414474487305,
+      "learning_rate": 2.5025764102040966e-06,
+      "loss": 0.6981,
+      "step": 57260
+    },
+    {
+      "epoch": 17.13,
+      "grad_norm": 1.999995231628418,
+      "learning_rate": 2.50001483674141e-06,
+      "loss": 1.1849,
+      "step": 57265
+    },
+    {
+      "epoch": 17.13,
+      "grad_norm": 7.203548431396484,
+      "learning_rate": 2.497454505929517e-06,
+      "loss": 0.8368,
+      "step": 57270
+    },
+    {
+      "epoch": 17.14,
+      "grad_norm": 1.5824021100997925,
+      "learning_rate": 2.494895417909826e-06,
+      "loss": 1.0648,
+      "step": 57275
+    },
+    {
+      "epoch": 17.14,
+      "grad_norm": 2.918457269668579,
+      "learning_rate": 2.492337572823658e-06,
+      "loss": 0.8636,
+      "step": 57280
+    },
+    {
+      "epoch": 17.14,
+      "grad_norm": 4.666050434112549,
+      "learning_rate": 2.4897809708123e-06,
+      "loss": 0.8613,
+      "step": 57285
+    },
+    {
+      "epoch": 17.14,
+      "grad_norm": 3.866513967514038,
+      "learning_rate": 2.487225612016933e-06,
+      "loss": 1.0003,
+      "step": 57290
+    },
+    {
+      "epoch": 17.14,
+      "grad_norm": 3.5122618675231934,
+      "learning_rate": 2.4846714965787027e-06,
+      "loss": 1.0402,
+      "step": 57295
+    },
+    {
+      "epoch": 17.14,
+      "grad_norm": 3.2854020595550537,
+      "learning_rate": 2.4821186246386575e-06,
+      "loss": 0.8922,
+      "step": 57300
+    },
+    {
+      "epoch": 17.14,
+      "grad_norm": 3.5705018043518066,
+      "learning_rate": 2.4795669963377955e-06,
+      "loss": 1.0352,
+      "step": 57305
+    },
+    {
+      "epoch": 17.15,
+      "grad_norm": 2.31522798538208,
+      "learning_rate": 2.4770166118170374e-06,
+      "loss": 0.9692,
+      "step": 57310
+    },
+    {
+      "epoch": 17.15,
+      "grad_norm": 4.260332107543945,
+      "learning_rate": 2.474467471217243e-06,
+      "loss": 1.0586,
+      "step": 57315
+    },
+    {
+      "epoch": 17.15,
+      "grad_norm": 3.6906752586364746,
+      "learning_rate": 2.4719195746791964e-06,
+      "loss": 0.8967,
+      "step": 57320
+    },
+    {
+      "epoch": 17.15,
+      "grad_norm": 2.0712099075317383,
+      "learning_rate": 2.469372922343613e-06,
+      "loss": 1.1484,
+      "step": 57325
+    },
+    {
+      "epoch": 17.15,
+      "grad_norm": 2.5663881301879883,
+      "learning_rate": 2.466827514351144e-06,
+      "loss": 0.9263,
+      "step": 57330
+    },
+    {
+      "epoch": 17.15,
+      "grad_norm": 2.5558412075042725,
+      "learning_rate": 2.4642833508423724e-06,
+      "loss": 1.1596,
+      "step": 57335
+    },
+    {
+      "epoch": 17.16,
+      "grad_norm": 4.297339916229248,
+      "learning_rate": 2.461740431957804e-06,
+      "loss": 0.8887,
+      "step": 57340
+    },
+    {
+      "epoch": 17.16,
+      "grad_norm": 1.2254616022109985,
+      "learning_rate": 2.4591987578378856e-06,
+      "loss": 1.0521,
+      "step": 57345
+    },
+    {
+      "epoch": 17.16,
+      "grad_norm": 3.5310375690460205,
+      "learning_rate": 2.456658328622988e-06,
+      "loss": 0.7953,
+      "step": 57350
+    },
+    {
+      "epoch": 17.16,
+      "grad_norm": 2.0364389419555664,
+      "learning_rate": 2.4541191444534205e-06,
+      "loss": 0.9538,
+      "step": 57355
+    },
+    {
+      "epoch": 17.16,
+      "grad_norm": 2.6163721084594727,
+      "learning_rate": 2.4515812054694166e-06,
+      "loss": 0.9859,
+      "step": 57360
+    },
+    {
+      "epoch": 17.16,
+      "grad_norm": 4.850313663482666,
+      "learning_rate": 2.4490445118111437e-06,
+      "loss": 0.8275,
+      "step": 57365
+    },
+    {
+      "epoch": 17.16,
+      "grad_norm": 2.779263734817505,
+      "learning_rate": 2.44650906361871e-06,
+      "loss": 1.0238,
+      "step": 57370
+    },
+    {
+      "epoch": 17.17,
+      "grad_norm": 1.5706207752227783,
+      "learning_rate": 2.443974861032125e-06,
+      "loss": 1.1682,
+      "step": 57375
+    },
+    {
+      "epoch": 17.17,
+      "grad_norm": 2.980870246887207,
+      "learning_rate": 2.4414419041913744e-06,
+      "loss": 1.0511,
+      "step": 57380
+    },
+    {
+      "epoch": 17.17,
+      "grad_norm": 1.8307000398635864,
+      "learning_rate": 2.438910193236327e-06,
+      "loss": 1.1374,
+      "step": 57385
+    },
+    {
+      "epoch": 17.17,
+      "grad_norm": 4.98880672454834,
+      "learning_rate": 2.436379728306831e-06,
+      "loss": 0.8951,
+      "step": 57390
+    },
+    {
+      "epoch": 17.17,
+      "grad_norm": 3.5696189403533936,
+      "learning_rate": 2.433850509542618e-06,
+      "loss": 0.9665,
+      "step": 57395
+    },
+    {
+      "epoch": 17.17,
+      "grad_norm": 2.4421303272247314,
+      "learning_rate": 2.431322537083394e-06,
+      "loss": 1.0267,
+      "step": 57400
+    },
+    {
+      "epoch": 17.17,
+      "grad_norm": 4.078541278839111,
+      "learning_rate": 2.428795811068765e-06,
+      "loss": 0.8258,
+      "step": 57405
+    },
+    {
+      "epoch": 17.18,
+      "grad_norm": 2.0753166675567627,
+      "learning_rate": 2.426270331638281e-06,
+      "loss": 0.8415,
+      "step": 57410
+    },
+    {
+      "epoch": 17.18,
+      "grad_norm": 2.376988410949707,
+      "learning_rate": 2.423746098931423e-06,
+      "loss": 1.0177,
+      "step": 57415
+    },
+    {
+      "epoch": 17.18,
+      "grad_norm": 4.53126335144043,
+      "learning_rate": 2.4212231130876007e-06,
+      "loss": 0.9601,
+      "step": 57420
+    },
+    {
+      "epoch": 17.18,
+      "grad_norm": 2.623030662536621,
+      "learning_rate": 2.4187013742461545e-06,
+      "loss": 0.8718,
+      "step": 57425
+    },
+    {
+      "epoch": 17.18,
+      "grad_norm": 2.2002532482147217,
+      "learning_rate": 2.4161808825463622e-06,
+      "loss": 1.0355,
+      "step": 57430
+    },
+    {
+      "epoch": 17.18,
+      "grad_norm": 2.742001533508301,
+      "learning_rate": 2.4136616381274237e-06,
+      "loss": 0.8977,
+      "step": 57435
+    },
+    {
+      "epoch": 17.19,
+      "grad_norm": 2.284527540206909,
+      "learning_rate": 2.411143641128477e-06,
+      "loss": 0.9512,
+      "step": 57440
+    },
+    {
+      "epoch": 17.19,
+      "grad_norm": 4.969058036804199,
+      "learning_rate": 2.408626891688587e-06,
+      "loss": 0.8995,
+      "step": 57445
+    },
+    {
+      "epoch": 17.19,
+      "grad_norm": 2.935258150100708,
+      "learning_rate": 2.4061113899467497e-06,
+      "loss": 0.9235,
+      "step": 57450
+    },
+    {
+      "epoch": 17.19,
+      "grad_norm": 4.259613037109375,
+      "learning_rate": 2.4035971360418963e-06,
+      "loss": 1.0788,
+      "step": 57455
+    },
+    {
+      "epoch": 17.19,
+      "grad_norm": 2.900808095932007,
+      "learning_rate": 2.401084130112885e-06,
+      "loss": 1.1105,
+      "step": 57460
+    },
+    {
+      "epoch": 17.19,
+      "grad_norm": 2.809903860092163,
+      "learning_rate": 2.398572372298513e-06,
+      "loss": 0.986,
+      "step": 57465
+    },
+    {
+      "epoch": 17.19,
+      "grad_norm": 2.9427716732025146,
+      "learning_rate": 2.396061862737484e-06,
+      "loss": 1.0418,
+      "step": 57470
+    },
+    {
+      "epoch": 17.2,
+      "grad_norm": 4.519219398498535,
+      "learning_rate": 2.3935526015684745e-06,
+      "loss": 1.0045,
+      "step": 57475
+    },
+    {
+      "epoch": 17.2,
+      "grad_norm": 2.9605634212493896,
+      "learning_rate": 2.391044588930047e-06,
+      "loss": 0.9415,
+      "step": 57480
+    },
+    {
+      "epoch": 17.2,
+      "grad_norm": 2.0215342044830322,
+      "learning_rate": 2.388537824960735e-06,
+      "loss": 0.9759,
+      "step": 57485
+    },
+    {
+      "epoch": 17.2,
+      "grad_norm": 1.2563127279281616,
+      "learning_rate": 2.3860323097989663e-06,
+      "loss": 0.956,
+      "step": 57490
+    },
+    {
+      "epoch": 17.2,
+      "grad_norm": 1.9748398065567017,
+      "learning_rate": 2.3835280435831385e-06,
+      "loss": 0.9808,
+      "step": 57495
+    },
+    {
+      "epoch": 17.2,
+      "grad_norm": 2.0095794200897217,
+      "learning_rate": 2.3810250264515415e-06,
+      "loss": 1.0952,
+      "step": 57500
+    },
+    {
+      "epoch": 17.2,
+      "grad_norm": 2.950597047805786,
+      "learning_rate": 2.3785232585424226e-06,
+      "loss": 1.0353,
+      "step": 57505
+    },
+    {
+      "epoch": 17.21,
+      "grad_norm": 1.762516736984253,
+      "learning_rate": 2.3760227399939493e-06,
+      "loss": 1.081,
+      "step": 57510
+    },
+    {
+      "epoch": 17.21,
+      "grad_norm": 4.237101078033447,
+      "learning_rate": 2.3735234709442194e-06,
+      "loss": 1.1056,
+      "step": 57515
+    },
+    {
+      "epoch": 17.21,
+      "grad_norm": 2.0214059352874756,
+      "learning_rate": 2.371025451531278e-06,
+      "loss": 0.933,
+      "step": 57520
+    },
+    {
+      "epoch": 17.21,
+      "grad_norm": 2.3543801307678223,
+      "learning_rate": 2.36852868189307e-06,
+      "loss": 0.9907,
+      "step": 57525
+    },
+    {
+      "epoch": 17.21,
+      "grad_norm": 3.8169326782226562,
+      "learning_rate": 2.366033162167508e-06,
+      "loss": 1.045,
+      "step": 57530
+    },
+    {
+      "epoch": 17.21,
+      "grad_norm": 1.6182808876037598,
+      "learning_rate": 2.363538892492406e-06,
+      "loss": 0.9037,
+      "step": 57535
+    },
+    {
+      "epoch": 17.22,
+      "grad_norm": 2.3788492679595947,
+      "learning_rate": 2.3610458730055186e-06,
+      "loss": 1.0592,
+      "step": 57540
+    },
+    {
+      "epoch": 17.22,
+      "grad_norm": 1.563767910003662,
+      "learning_rate": 2.3585541038445374e-06,
+      "loss": 0.8683,
+      "step": 57545
+    },
+    {
+      "epoch": 17.22,
+      "grad_norm": 3.2709128856658936,
+      "learning_rate": 2.356063585147078e-06,
+      "loss": 0.8699,
+      "step": 57550
+    },
+    {
+      "epoch": 17.22,
+      "grad_norm": 2.178466320037842,
+      "learning_rate": 2.3535743170506887e-06,
+      "loss": 0.8278,
+      "step": 57555
+    },
+    {
+      "epoch": 17.22,
+      "grad_norm": 4.895992755889893,
+      "learning_rate": 2.351086299692856e-06,
+      "loss": 1.0167,
+      "step": 57560
+    },
+    {
+      "epoch": 17.22,
+      "grad_norm": 3.7449183464050293,
+      "learning_rate": 2.348599533210974e-06,
+      "loss": 1.1186,
+      "step": 57565
+    },
+    {
+      "epoch": 17.22,
+      "grad_norm": 4.224990367889404,
+      "learning_rate": 2.346114017742407e-06,
+      "loss": 0.9867,
+      "step": 57570
+    },
+    {
+      "epoch": 17.23,
+      "grad_norm": 1.4723498821258545,
+      "learning_rate": 2.3436297534244035e-06,
+      "loss": 1.0563,
+      "step": 57575
+    },
+    {
+      "epoch": 17.23,
+      "grad_norm": 2.0138280391693115,
+      "learning_rate": 2.34114674039419e-06,
+      "loss": 0.8924,
+      "step": 57580
+    },
+    {
+      "epoch": 17.23,
+      "grad_norm": 5.9495391845703125,
+      "learning_rate": 2.3386649787888794e-06,
+      "loss": 1.0926,
+      "step": 57585
+    },
+    {
+      "epoch": 17.23,
+      "grad_norm": 3.3847897052764893,
+      "learning_rate": 2.33618446874555e-06,
+      "loss": 0.9745,
+      "step": 57590
+    },
+    {
+      "epoch": 17.23,
+      "grad_norm": 8.030804634094238,
+      "learning_rate": 2.333705210401202e-06,
+      "loss": 0.9503,
+      "step": 57595
+    },
+    {
+      "epoch": 17.23,
+      "grad_norm": 1.3774622678756714,
+      "learning_rate": 2.3312272038927443e-06,
+      "loss": 0.8589,
+      "step": 57600
+    },
+    {
+      "epoch": 17.23,
+      "grad_norm": 4.3188652992248535,
+      "learning_rate": 2.3287504493570534e-06,
+      "loss": 0.8434,
+      "step": 57605
+    },
+    {
+      "epoch": 17.24,
+      "grad_norm": 2.444995880126953,
+      "learning_rate": 2.3262749469309007e-06,
+      "loss": 1.1008,
+      "step": 57610
+    },
+    {
+      "epoch": 17.24,
+      "grad_norm": 4.819691181182861,
+      "learning_rate": 2.3238006967510273e-06,
+      "loss": 0.9677,
+      "step": 57615
+    },
+    {
+      "epoch": 17.24,
+      "grad_norm": 3.0465807914733887,
+      "learning_rate": 2.321327698954057e-06,
+      "loss": 0.7771,
+      "step": 57620
+    },
+    {
+      "epoch": 17.24,
+      "grad_norm": 1.9639545679092407,
+      "learning_rate": 2.3188559536765965e-06,
+      "loss": 0.7874,
+      "step": 57625
+    },
+    {
+      "epoch": 17.24,
+      "grad_norm": 4.145541191101074,
+      "learning_rate": 2.3163854610551427e-06,
+      "loss": 0.9116,
+      "step": 57630
+    },
+    {
+      "epoch": 17.24,
+      "grad_norm": 5.50032377243042,
+      "learning_rate": 2.3139162212261394e-06,
+      "loss": 0.828,
+      "step": 57635
+    },
+    {
+      "epoch": 17.25,
+      "grad_norm": 3.153430223464966,
+      "learning_rate": 2.311448234325961e-06,
+      "loss": 1.1071,
+      "step": 57640
+    },
+    {
+      "epoch": 17.25,
+      "grad_norm": 4.366771697998047,
+      "learning_rate": 2.3089815004909145e-06,
+      "loss": 1.0254,
+      "step": 57645
+    },
+    {
+      "epoch": 17.25,
+      "grad_norm": 3.364361047744751,
+      "learning_rate": 2.3065160198572355e-06,
+      "loss": 0.8954,
+      "step": 57650
+    },
+    {
+      "epoch": 17.25,
+      "grad_norm": 2.316021680831909,
+      "learning_rate": 2.3040517925610904e-06,
+      "loss": 0.8784,
+      "step": 57655
+    },
+    {
+      "epoch": 17.25,
+      "grad_norm": 3.245994806289673,
+      "learning_rate": 2.3015888187385674e-06,
+      "loss": 1.0405,
+      "step": 57660
+    },
+    {
+      "epoch": 17.25,
+      "grad_norm": 2.167280435562134,
+      "learning_rate": 2.299127098525705e-06,
+      "loss": 0.9786,
+      "step": 57665
+    },
+    {
+      "epoch": 17.25,
+      "grad_norm": 3.1760175228118896,
+      "learning_rate": 2.29666663205845e-06,
+      "loss": 0.9913,
+      "step": 57670
+    },
+    {
+      "epoch": 17.26,
+      "grad_norm": 1.8148313760757446,
+      "learning_rate": 2.294207419472705e-06,
+      "loss": 1.0802,
+      "step": 57675
+    },
+    {
+      "epoch": 17.26,
+      "grad_norm": 5.893067836761475,
+      "learning_rate": 2.29174946090428e-06,
+      "loss": 0.751,
+      "step": 57680
+    },
+    {
+      "epoch": 17.26,
+      "grad_norm": 2.0050816535949707,
+      "learning_rate": 2.2892927564889284e-06,
+      "loss": 0.9561,
+      "step": 57685
+    },
+    {
+      "epoch": 17.26,
+      "grad_norm": 2.132786512374878,
+      "learning_rate": 2.286837306362338e-06,
+      "loss": 0.838,
+      "step": 57690
+    },
+    {
+      "epoch": 17.26,
+      "grad_norm": 3.9043054580688477,
+      "learning_rate": 2.2843831106601017e-06,
+      "loss": 0.9945,
+      "step": 57695
+    },
+    {
+      "epoch": 17.26,
+      "grad_norm": 2.53236722946167,
+      "learning_rate": 2.2819301695177876e-06,
+      "loss": 0.8508,
+      "step": 57700
+    },
+    {
+      "epoch": 17.26,
+      "grad_norm": 2.5461416244506836,
+      "learning_rate": 2.2794784830708432e-06,
+      "loss": 0.7963,
+      "step": 57705
+    },
+    {
+      "epoch": 17.27,
+      "grad_norm": 3.189567804336548,
+      "learning_rate": 2.2770280514546964e-06,
+      "loss": 0.9692,
+      "step": 57710
+    },
+    {
+      "epoch": 17.27,
+      "grad_norm": 2.352689027786255,
+      "learning_rate": 2.2745788748046614e-06,
+      "loss": 0.9831,
+      "step": 57715
+    },
+    {
+      "epoch": 17.27,
+      "grad_norm": 2.441016674041748,
+      "learning_rate": 2.2721309532560244e-06,
+      "loss": 0.962,
+      "step": 57720
+    },
+    {
+      "epoch": 17.27,
+      "grad_norm": 3.2766942977905273,
+      "learning_rate": 2.2696842869439657e-06,
+      "loss": 0.9884,
+      "step": 57725
+    },
+    {
+      "epoch": 17.27,
+      "grad_norm": 4.255434989929199,
+      "learning_rate": 2.267238876003616e-06,
+      "loss": 1.06,
+      "step": 57730
+    },
+    {
+      "epoch": 17.27,
+      "grad_norm": 4.779931545257568,
+      "learning_rate": 2.2647947205700322e-06,
+      "loss": 1.0216,
+      "step": 57735
+    },
+    {
+      "epoch": 17.28,
+      "grad_norm": 2.0698294639587402,
+      "learning_rate": 2.262351820778205e-06,
+      "loss": 0.8473,
+      "step": 57740
+    },
+    {
+      "epoch": 17.28,
+      "grad_norm": 2.572493553161621,
+      "learning_rate": 2.2599101767630524e-06,
+      "loss": 1.0734,
+      "step": 57745
+    },
+    {
+      "epoch": 17.28,
+      "grad_norm": 3.1653802394866943,
+      "learning_rate": 2.257469788659425e-06,
+      "loss": 0.8459,
+      "step": 57750
+    },
+    {
+      "epoch": 17.28,
+      "grad_norm": 2.538551092147827,
+      "learning_rate": 2.255030656602103e-06,
+      "loss": 0.9257,
+      "step": 57755
+    },
+    {
+      "epoch": 17.28,
+      "grad_norm": 1.3721609115600586,
+      "learning_rate": 2.2525927807257928e-06,
+      "loss": 0.984,
+      "step": 57760
+    },
+    {
+      "epoch": 17.28,
+      "grad_norm": 1.590498447418213,
+      "learning_rate": 2.25015616116514e-06,
+      "loss": 0.8294,
+      "step": 57765
+    },
+    {
+      "epoch": 17.28,
+      "grad_norm": 3.673794746398926,
+      "learning_rate": 2.2477207980547145e-06,
+      "loss": 0.9454,
+      "step": 57770
+    },
+    {
+      "epoch": 17.29,
+      "grad_norm": 3.491346836090088,
+      "learning_rate": 2.2452866915290195e-06,
+      "loss": 1.0912,
+      "step": 57775
+    },
+    {
+      "epoch": 17.29,
+      "grad_norm": 6.499100208282471,
+      "learning_rate": 2.2428538417224894e-06,
+      "loss": 0.9398,
+      "step": 57780
+    },
+    {
+      "epoch": 17.29,
+      "grad_norm": 2.021291971206665,
+      "learning_rate": 2.2404222487694913e-06,
+      "loss": 0.9325,
+      "step": 57785
+    },
+    {
+      "epoch": 17.29,
+      "grad_norm": 2.0946333408355713,
+      "learning_rate": 2.2379919128043047e-06,
+      "loss": 0.9485,
+      "step": 57790
+    },
+    {
+      "epoch": 17.29,
+      "grad_norm": 1.8785912990570068,
+      "learning_rate": 2.2355628339611744e-06,
+      "loss": 0.9824,
+      "step": 57795
+    },
+    {
+      "epoch": 17.29,
+      "grad_norm": 3.972280979156494,
+      "learning_rate": 2.233135012374238e-06,
+      "loss": 0.9679,
+      "step": 57800
+    },
+    {
+      "epoch": 17.29,
+      "grad_norm": 6.458027362823486,
+      "learning_rate": 2.2307084481775985e-06,
+      "loss": 0.9765,
+      "step": 57805
+    },
+    {
+      "epoch": 17.3,
+      "grad_norm": 8.290522575378418,
+      "learning_rate": 2.228283141505255e-06,
+      "loss": 1.0319,
+      "step": 57810
+    },
+    {
+      "epoch": 17.3,
+      "grad_norm": 3.988783121109009,
+      "learning_rate": 2.2258590924911753e-06,
+      "loss": 0.9265,
+      "step": 57815
+    },
+    {
+      "epoch": 17.3,
+      "grad_norm": 1.954484224319458,
+      "learning_rate": 2.223436301269219e-06,
+      "loss": 1.0169,
+      "step": 57820
+    },
+    {
+      "epoch": 17.3,
+      "grad_norm": 4.165136814117432,
+      "learning_rate": 2.221014767973201e-06,
+      "loss": 1.0911,
+      "step": 57825
+    },
+    {
+      "epoch": 17.3,
+      "grad_norm": 3.616419553756714,
+      "learning_rate": 2.2185944927368587e-06,
+      "loss": 0.9891,
+      "step": 57830
+    },
+    {
+      "epoch": 17.3,
+      "grad_norm": 2.0609145164489746,
+      "learning_rate": 2.216175475693863e-06,
+      "loss": 1.0497,
+      "step": 57835
+    },
+    {
+      "epoch": 17.31,
+      "grad_norm": 3.2805593013763428,
+      "learning_rate": 2.2137577169778155e-06,
+      "loss": 0.9588,
+      "step": 57840
+    },
+    {
+      "epoch": 17.31,
+      "grad_norm": 1.7305797338485718,
+      "learning_rate": 2.211341216722243e-06,
+      "loss": 0.9302,
+      "step": 57845
+    },
+    {
+      "epoch": 17.31,
+      "grad_norm": 3.898714303970337,
+      "learning_rate": 2.2089259750606052e-06,
+      "loss": 1.1782,
+      "step": 57850
+    },
+    {
+      "epoch": 17.31,
+      "grad_norm": 1.8012914657592773,
+      "learning_rate": 2.206511992126298e-06,
+      "loss": 0.9423,
+      "step": 57855
+    },
+    {
+      "epoch": 17.31,
+      "grad_norm": 1.263310432434082,
+      "learning_rate": 2.2040992680526423e-06,
+      "loss": 0.9829,
+      "step": 57860
+    },
+    {
+      "epoch": 17.31,
+      "grad_norm": 1.296061396598816,
+      "learning_rate": 2.2016878029728877e-06,
+      "loss": 0.9826,
+      "step": 57865
+    },
+    {
+      "epoch": 17.31,
+      "grad_norm": 3.5253007411956787,
+      "learning_rate": 2.199277597020219e-06,
+      "loss": 0.9559,
+      "step": 57870
+    },
+    {
+      "epoch": 17.32,
+      "grad_norm": 1.947613000869751,
+      "learning_rate": 2.1968686503277464e-06,
+      "loss": 0.9857,
+      "step": 57875
+    },
+    {
+      "epoch": 17.32,
+      "grad_norm": 1.1814169883728027,
+      "learning_rate": 2.194460963028516e-06,
+      "loss": 0.9992,
+      "step": 57880
+    },
+    {
+      "epoch": 17.32,
+      "grad_norm": 3.306941509246826,
+      "learning_rate": 2.192054535255503e-06,
+      "loss": 0.9873,
+      "step": 57885
+    },
+    {
+      "epoch": 17.32,
+      "grad_norm": 2.9445595741271973,
+      "learning_rate": 2.189649367141616e-06,
+      "loss": 0.8051,
+      "step": 57890
+    },
+    {
+      "epoch": 17.32,
+      "grad_norm": 2.4458541870117188,
+      "learning_rate": 2.187245458819673e-06,
+      "loss": 1.121,
+      "step": 57895
+    },
+    {
+      "epoch": 17.32,
+      "grad_norm": 4.300705432891846,
+      "learning_rate": 2.1848428104224605e-06,
+      "loss": 0.9273,
+      "step": 57900
+    },
+    {
+      "epoch": 17.32,
+      "grad_norm": 1.9480119943618774,
+      "learning_rate": 2.1824414220826566e-06,
+      "loss": 0.9085,
+      "step": 57905
+    },
+    {
+      "epoch": 17.33,
+      "grad_norm": 5.82395601272583,
+      "learning_rate": 2.180041293932905e-06,
+      "loss": 1.0309,
+      "step": 57910
+    },
+    {
+      "epoch": 17.33,
+      "grad_norm": 2.66611909866333,
+      "learning_rate": 2.177642426105747e-06,
+      "loss": 1.1571,
+      "step": 57915
+    },
+    {
+      "epoch": 17.33,
+      "grad_norm": 1.7344080209732056,
+      "learning_rate": 2.1752448187336738e-06,
+      "loss": 0.8145,
+      "step": 57920
+    },
+    {
+      "epoch": 17.33,
+      "grad_norm": 4.083990573883057,
+      "learning_rate": 2.172848471949107e-06,
+      "loss": 0.8786,
+      "step": 57925
+    },
+    {
+      "epoch": 17.33,
+      "grad_norm": 3.4210381507873535,
+      "learning_rate": 2.1704533858843884e-06,
+      "loss": 0.9086,
+      "step": 57930
+    },
+    {
+      "epoch": 17.33,
+      "grad_norm": 2.1340394020080566,
+      "learning_rate": 2.1680595606718008e-06,
+      "loss": 0.9275,
+      "step": 57935
+    },
+    {
+      "epoch": 17.33,
+      "grad_norm": 2.192603588104248,
+      "learning_rate": 2.165666996443552e-06,
+      "loss": 1.0049,
+      "step": 57940
+    },
+    {
+      "epoch": 17.34,
+      "grad_norm": 1.9342150688171387,
+      "learning_rate": 2.1632756933317815e-06,
+      "loss": 0.9161,
+      "step": 57945
+    },
+    {
+      "epoch": 17.34,
+      "grad_norm": 2.2969846725463867,
+      "learning_rate": 2.160885651468553e-06,
+      "loss": 1.0195,
+      "step": 57950
+    },
+    {
+      "epoch": 17.34,
+      "grad_norm": 2.6062443256378174,
+      "learning_rate": 2.158496870985874e-06,
+      "loss": 1.0117,
+      "step": 57955
+    },
+    {
+      "epoch": 17.34,
+      "grad_norm": 1.2157763242721558,
+      "learning_rate": 2.156109352015667e-06,
+      "loss": 0.9364,
+      "step": 57960
+    },
+    {
+      "epoch": 17.34,
+      "grad_norm": 3.201415777206421,
+      "learning_rate": 2.153723094689797e-06,
+      "loss": 1.1783,
+      "step": 57965
+    },
+    {
+      "epoch": 17.34,
+      "grad_norm": 2.736159324645996,
+      "learning_rate": 2.1513380991400547e-06,
+      "loss": 0.8852,
+      "step": 57970
+    },
+    {
+      "epoch": 17.35,
+      "grad_norm": 2.2211790084838867,
+      "learning_rate": 2.148954365498157e-06,
+      "loss": 0.8922,
+      "step": 57975
+    },
+    {
+      "epoch": 17.35,
+      "grad_norm": 0.9812418818473816,
+      "learning_rate": 2.1465718938957576e-06,
+      "loss": 1.049,
+      "step": 57980
+    },
+    {
+      "epoch": 17.35,
+      "grad_norm": 2.103342294692993,
+      "learning_rate": 2.144190684464442e-06,
+      "loss": 1.1637,
+      "step": 57985
+    },
+    {
+      "epoch": 17.35,
+      "grad_norm": 2.9801440238952637,
+      "learning_rate": 2.1418107373357116e-06,
+      "loss": 0.8715,
+      "step": 57990
+    },
+    {
+      "epoch": 17.35,
+      "grad_norm": 3.3744688034057617,
+      "learning_rate": 2.139432052641019e-06,
+      "loss": 1.0241,
+      "step": 57995
+    },
+    {
+      "epoch": 17.35,
+      "grad_norm": 3.7058229446411133,
+      "learning_rate": 2.1370546305117254e-06,
+      "loss": 1.0102,
+      "step": 58000
+    },
+    {
+      "epoch": 17.35,
+      "grad_norm": 4.445005416870117,
+      "learning_rate": 2.134678471079149e-06,
+      "loss": 1.1774,
+      "step": 58005
+    },
+    {
+      "epoch": 17.36,
+      "grad_norm": 2.2905967235565186,
+      "learning_rate": 2.132303574474509e-06,
+      "loss": 0.8403,
+      "step": 58010
+    },
+    {
+      "epoch": 17.36,
+      "grad_norm": 1.867017388343811,
+      "learning_rate": 2.129929940828973e-06,
+      "loss": 0.987,
+      "step": 58015
+    },
+    {
+      "epoch": 17.36,
+      "grad_norm": 2.154637575149536,
+      "learning_rate": 2.1275575702736334e-06,
+      "loss": 0.9626,
+      "step": 58020
+    },
+    {
+      "epoch": 17.36,
+      "grad_norm": 1.9806933403015137,
+      "learning_rate": 2.1251864629395156e-06,
+      "loss": 1.1247,
+      "step": 58025
+    },
+    {
+      "epoch": 17.36,
+      "grad_norm": 2.581702947616577,
+      "learning_rate": 2.122816618957571e-06,
+      "loss": 1.0137,
+      "step": 58030
+    },
+    {
+      "epoch": 17.36,
+      "grad_norm": 2.4726662635803223,
+      "learning_rate": 2.120448038458686e-06,
+      "loss": 0.9178,
+      "step": 58035
+    },
+    {
+      "epoch": 17.36,
+      "grad_norm": 1.8047939538955688,
+      "learning_rate": 2.118080721573673e-06,
+      "loss": 1.1279,
+      "step": 58040
+    },
+    {
+      "epoch": 17.37,
+      "grad_norm": 1.1406919956207275,
+      "learning_rate": 2.1157146684332774e-06,
+      "loss": 0.9604,
+      "step": 58045
+    },
+    {
+      "epoch": 17.37,
+      "grad_norm": 2.2998058795928955,
+      "learning_rate": 2.113349879168175e-06,
+      "loss": 1.1191,
+      "step": 58050
+    },
+    {
+      "epoch": 17.37,
+      "grad_norm": 2.92384672164917,
+      "learning_rate": 2.11098635390897e-06,
+      "loss": 1.0299,
+      "step": 58055
+    },
+    {
+      "epoch": 17.37,
+      "grad_norm": 2.4828195571899414,
+      "learning_rate": 2.1086240927861933e-06,
+      "loss": 1.1108,
+      "step": 58060
+    },
+    {
+      "epoch": 17.37,
+      "grad_norm": 2.7545273303985596,
+      "learning_rate": 2.1062630959303163e-06,
+      "loss": 1.0005,
+      "step": 58065
+    },
+    {
+      "epoch": 17.37,
+      "grad_norm": 2.8774526119232178,
+      "learning_rate": 2.103903363471732e-06,
+      "loss": 1.0559,
+      "step": 58070
+    },
+    {
+      "epoch": 17.38,
+      "grad_norm": 1.6022573709487915,
+      "learning_rate": 2.1015448955407637e-06,
+      "loss": 0.8921,
+      "step": 58075
+    },
+    {
+      "epoch": 17.38,
+      "grad_norm": 3.47226619720459,
+      "learning_rate": 2.0991876922676764e-06,
+      "loss": 1.0754,
+      "step": 58080
+    },
+    {
+      "epoch": 17.38,
+      "grad_norm": 1.9637651443481445,
+      "learning_rate": 2.096831753782638e-06,
+      "loss": 0.9987,
+      "step": 58085
+    },
+    {
+      "epoch": 17.38,
+      "grad_norm": 1.8076112270355225,
+      "learning_rate": 2.094477080215787e-06,
+      "loss": 0.9186,
+      "step": 58090
+    },
+    {
+      "epoch": 17.38,
+      "grad_norm": 3.499573230743408,
+      "learning_rate": 2.0921236716971465e-06,
+      "loss": 1.0806,
+      "step": 58095
+    },
+    {
+      "epoch": 17.38,
+      "grad_norm": 5.197295188903809,
+      "learning_rate": 2.0897715283567126e-06,
+      "loss": 1.0594,
+      "step": 58100
+    },
+    {
+      "epoch": 17.38,
+      "grad_norm": 4.264853477478027,
+      "learning_rate": 2.0874206503243758e-06,
+      "loss": 0.8798,
+      "step": 58105
+    },
+    {
+      "epoch": 17.39,
+      "grad_norm": 4.624975681304932,
+      "learning_rate": 2.0850710377299883e-06,
+      "loss": 0.9827,
+      "step": 58110
+    },
+    {
+      "epoch": 17.39,
+      "grad_norm": 2.4760243892669678,
+      "learning_rate": 2.0827226907033037e-06,
+      "loss": 0.9113,
+      "step": 58115
+    },
+    {
+      "epoch": 17.39,
+      "grad_norm": 1.9899121522903442,
+      "learning_rate": 2.0803756093740247e-06,
+      "loss": 0.9547,
+      "step": 58120
+    },
+    {
+      "epoch": 17.39,
+      "grad_norm": 2.200958251953125,
+      "learning_rate": 2.0780297938717776e-06,
+      "loss": 0.9839,
+      "step": 58125
+    },
+    {
+      "epoch": 17.39,
+      "grad_norm": 2.856152296066284,
+      "learning_rate": 2.075685244326117e-06,
+      "loss": 0.8033,
+      "step": 58130
+    },
+    {
+      "epoch": 17.39,
+      "grad_norm": 2.164071559906006,
+      "learning_rate": 2.0733419608665345e-06,
+      "loss": 0.8925,
+      "step": 58135
+    },
+    {
+      "epoch": 17.39,
+      "grad_norm": 1.5249561071395874,
+      "learning_rate": 2.0709999436224426e-06,
+      "loss": 0.958,
+      "step": 58140
+    },
+    {
+      "epoch": 17.4,
+      "grad_norm": 3.109286069869995,
+      "learning_rate": 2.0686591927231908e-06,
+      "loss": 0.8569,
+      "step": 58145
+    },
+    {
+      "epoch": 17.4,
+      "grad_norm": 3.086642265319824,
+      "learning_rate": 2.066319708298056e-06,
+      "loss": 1.0587,
+      "step": 58150
+    },
+    {
+      "epoch": 17.4,
+      "grad_norm": 3.6070919036865234,
+      "learning_rate": 2.063981490476247e-06,
+      "loss": 1.0945,
+      "step": 58155
+    },
+    {
+      "epoch": 17.4,
+      "grad_norm": 3.0385048389434814,
+      "learning_rate": 2.0616445393868977e-06,
+      "loss": 1.0345,
+      "step": 58160
+    },
+    {
+      "epoch": 17.4,
+      "grad_norm": 3.2083237171173096,
+      "learning_rate": 2.0593088551590783e-06,
+      "loss": 0.9921,
+      "step": 58165
+    },
+    {
+      "epoch": 17.4,
+      "grad_norm": 2.1536431312561035,
+      "learning_rate": 2.056974437921785e-06,
+      "loss": 1.1214,
+      "step": 58170
+    },
+    {
+      "epoch": 17.41,
+      "grad_norm": 4.799535274505615,
+      "learning_rate": 2.0546412878039517e-06,
+      "loss": 0.9159,
+      "step": 58175
+    },
+    {
+      "epoch": 17.41,
+      "grad_norm": 3.742593765258789,
+      "learning_rate": 2.052309404934419e-06,
+      "loss": 0.9264,
+      "step": 58180
+    },
+    {
+      "epoch": 17.41,
+      "grad_norm": 2.696674346923828,
+      "learning_rate": 2.049978789441992e-06,
+      "loss": 1.0157,
+      "step": 58185
+    },
+    {
+      "epoch": 17.41,
+      "grad_norm": 2.9800398349761963,
+      "learning_rate": 2.0476494414553765e-06,
+      "loss": 0.8998,
+      "step": 58190
+    },
+    {
+      "epoch": 17.41,
+      "grad_norm": 5.797310829162598,
+      "learning_rate": 2.045321361103231e-06,
+      "loss": 0.9081,
+      "step": 58195
+    },
+    {
+      "epoch": 17.41,
+      "grad_norm": 3.218557119369507,
+      "learning_rate": 2.0429945485141184e-06,
+      "loss": 0.8853,
+      "step": 58200
+    },
+    {
+      "epoch": 17.41,
+      "grad_norm": 4.598995208740234,
+      "learning_rate": 2.0411340113185823e-06,
+      "loss": 0.9268,
+      "step": 58205
+    },
+    {
+      "epoch": 17.42,
+      "grad_norm": 10.203455924987793,
+      "learning_rate": 2.0388094810267407e-06,
+      "loss": 0.9933,
+      "step": 58210
+    },
+    {
+      "epoch": 17.42,
+      "grad_norm": 1.8637256622314453,
+      "learning_rate": 2.0364862188575833e-06,
+      "loss": 1.0012,
+      "step": 58215
+    },
+    {
+      "epoch": 17.42,
+      "grad_norm": 4.718301296234131,
+      "learning_rate": 2.034164224939425e-06,
+      "loss": 1.0701,
+      "step": 58220
+    },
+    {
+      "epoch": 17.42,
+      "grad_norm": 3.8083252906799316,
+      "learning_rate": 2.0318434994005014e-06,
+      "loss": 0.8359,
+      "step": 58225
+    },
+    {
+      "epoch": 17.42,
+      "grad_norm": 3.1337733268737793,
+      "learning_rate": 2.0295240423689938e-06,
+      "loss": 0.968,
+      "step": 58230
+    },
+    {
+      "epoch": 17.42,
+      "grad_norm": 2.8172826766967773,
+      "learning_rate": 2.0272058539729882e-06,
+      "loss": 0.9344,
+      "step": 58235
+    },
+    {
+      "epoch": 17.42,
+      "grad_norm": 1.8639297485351562,
+      "learning_rate": 2.0248889343405324e-06,
+      "loss": 1.1076,
+      "step": 58240
+    },
+    {
+      "epoch": 17.43,
+      "grad_norm": 1.549034595489502,
+      "learning_rate": 2.0225732835995715e-06,
+      "loss": 1.0001,
+      "step": 58245
+    },
+    {
+      "epoch": 17.43,
+      "grad_norm": 1.5290988683700562,
+      "learning_rate": 2.020258901878014e-06,
+      "loss": 1.0315,
+      "step": 58250
+    },
+    {
+      "epoch": 17.43,
+      "grad_norm": 3.2922134399414062,
+      "learning_rate": 2.01794578930366e-06,
+      "loss": 0.976,
+      "step": 58255
+    },
+    {
+      "epoch": 17.43,
+      "grad_norm": 2.2940144538879395,
+      "learning_rate": 2.0156339460042804e-06,
+      "loss": 0.9909,
+      "step": 58260
+    },
+    {
+      "epoch": 17.43,
+      "grad_norm": 2.8492350578308105,
+      "learning_rate": 2.013323372107545e-06,
+      "loss": 1.0184,
+      "step": 58265
+    },
+    {
+      "epoch": 17.43,
+      "grad_norm": 7.312509059906006,
+      "learning_rate": 2.0110140677410655e-06,
+      "loss": 0.8777,
+      "step": 58270
+    },
+    {
+      "epoch": 17.44,
+      "grad_norm": 3.636141777038574,
+      "learning_rate": 2.0087060330323813e-06,
+      "loss": 0.8916,
+      "step": 58275
+    },
+    {
+      "epoch": 17.44,
+      "grad_norm": 4.8693366050720215,
+      "learning_rate": 2.006399268108969e-06,
+      "loss": 0.9538,
+      "step": 58280
+    },
+    {
+      "epoch": 17.44,
+      "grad_norm": 4.98232364654541,
+      "learning_rate": 2.00409377309822e-06,
+      "loss": 0.8613,
+      "step": 58285
+    },
+    {
+      "epoch": 17.44,
+      "grad_norm": 1.1212947368621826,
+      "learning_rate": 2.0017895481274724e-06,
+      "loss": 0.8985,
+      "step": 58290
+    },
+    {
+      "epoch": 17.44,
+      "grad_norm": 4.1995673179626465,
+      "learning_rate": 1.999486593323982e-06,
+      "loss": 0.9499,
+      "step": 58295
+    },
+    {
+      "epoch": 17.44,
+      "grad_norm": 5.486680507659912,
+      "learning_rate": 1.9971849088149393e-06,
+      "loss": 0.9938,
+      "step": 58300
+    },
+    {
+      "epoch": 17.44,
+      "grad_norm": 2.455383062362671,
+      "learning_rate": 1.9948844947274616e-06,
+      "loss": 1.2352,
+      "step": 58305
+    },
+    {
+      "epoch": 17.45,
+      "grad_norm": 2.302922010421753,
+      "learning_rate": 1.9925853511886026e-06,
+      "loss": 1.0595,
+      "step": 58310
+    },
+    {
+      "epoch": 17.45,
+      "grad_norm": 1.5419032573699951,
+      "learning_rate": 1.9902874783253393e-06,
+      "loss": 1.0541,
+      "step": 58315
+    },
+    {
+      "epoch": 17.45,
+      "grad_norm": 3.4702677726745605,
+      "learning_rate": 1.9879908762645837e-06,
+      "loss": 0.9092,
+      "step": 58320
+    },
+    {
+      "epoch": 17.45,
+      "grad_norm": 2.8185274600982666,
+      "learning_rate": 1.985695545133173e-06,
+      "loss": 0.9399,
+      "step": 58325
+    },
+    {
+      "epoch": 17.45,
+      "grad_norm": 2.852463960647583,
+      "learning_rate": 1.9834014850578703e-06,
+      "loss": 0.9442,
+      "step": 58330
+    },
+    {
+      "epoch": 17.45,
+      "grad_norm": 1.91469407081604,
+      "learning_rate": 1.9811086961653846e-06,
+      "loss": 1.1047,
+      "step": 58335
+    },
+    {
+      "epoch": 17.45,
+      "grad_norm": 3.1850225925445557,
+      "learning_rate": 1.9788171785823316e-06,
+      "loss": 0.9461,
+      "step": 58340
+    },
+    {
+      "epoch": 17.46,
+      "grad_norm": 2.731886386871338,
+      "learning_rate": 1.9765269324352855e-06,
+      "loss": 0.8916,
+      "step": 58345
+    },
+    {
+      "epoch": 17.46,
+      "grad_norm": 2.1498260498046875,
+      "learning_rate": 1.9742379578507163e-06,
+      "loss": 0.9886,
+      "step": 58350
+    },
+    {
+      "epoch": 17.46,
+      "grad_norm": 3.7887089252471924,
+      "learning_rate": 1.97195025495506e-06,
+      "loss": 0.9632,
+      "step": 58355
+    },
+    {
+      "epoch": 17.46,
+      "grad_norm": 2.147568702697754,
+      "learning_rate": 1.969663823874651e-06,
+      "loss": 0.911,
+      "step": 58360
+    },
+    {
+      "epoch": 17.46,
+      "grad_norm": 3.2035396099090576,
+      "learning_rate": 1.9673786647357693e-06,
+      "loss": 0.8811,
+      "step": 58365
+    },
+    {
+      "epoch": 17.46,
+      "grad_norm": 3.055304765701294,
+      "learning_rate": 1.9650947776646223e-06,
+      "loss": 1.1041,
+      "step": 58370
+    },
+    {
+      "epoch": 17.47,
+      "grad_norm": 3.3420493602752686,
+      "learning_rate": 1.9628121627873476e-06,
+      "loss": 0.9011,
+      "step": 58375
+    },
+    {
+      "epoch": 17.47,
+      "grad_norm": 1.865210771560669,
+      "learning_rate": 1.9605308202300136e-06,
+      "loss": 0.9144,
+      "step": 58380
+    },
+    {
+      "epoch": 17.47,
+      "grad_norm": 3.021851062774658,
+      "learning_rate": 1.958250750118612e-06,
+      "loss": 1.0783,
+      "step": 58385
+    },
+    {
+      "epoch": 17.47,
+      "grad_norm": 2.6499688625335693,
+      "learning_rate": 1.9559719525790693e-06,
+      "loss": 0.8827,
+      "step": 58390
+    },
+    {
+      "epoch": 17.47,
+      "grad_norm": 2.7535033226013184,
+      "learning_rate": 1.953694427737246e-06,
+      "loss": 0.9414,
+      "step": 58395
+    },
+    {
+      "epoch": 17.47,
+      "grad_norm": 1.7163090705871582,
+      "learning_rate": 1.9514181757189222e-06,
+      "loss": 0.8349,
+      "step": 58400
+    },
+    {
+      "epoch": 17.47,
+      "grad_norm": 1.809923529624939,
+      "learning_rate": 1.9491431966498162e-06,
+      "loss": 0.997,
+      "step": 58405
+    },
+    {
+      "epoch": 17.48,
+      "grad_norm": 2.4826242923736572,
+      "learning_rate": 1.94686949065557e-06,
+      "loss": 0.8481,
+      "step": 58410
+    },
+    {
+      "epoch": 17.48,
+      "grad_norm": 2.3572394847869873,
+      "learning_rate": 1.9445970578617605e-06,
+      "loss": 1.0958,
+      "step": 58415
+    },
+    {
+      "epoch": 17.48,
+      "grad_norm": 2.6370866298675537,
+      "learning_rate": 1.942325898393896e-06,
+      "loss": 0.8277,
+      "step": 58420
+    },
+    {
+      "epoch": 17.48,
+      "grad_norm": 2.8419244289398193,
+      "learning_rate": 1.9400560123773946e-06,
+      "loss": 0.9423,
+      "step": 58425
+    },
+    {
+      "epoch": 17.48,
+      "grad_norm": 0.7702686786651611,
+      "learning_rate": 1.937787399937638e-06,
+      "loss": 0.935,
+      "step": 58430
+    },
+    {
+      "epoch": 17.48,
+      "grad_norm": 2.7786147594451904,
+      "learning_rate": 1.9355200611999055e-06,
+      "loss": 0.778,
+      "step": 58435
+    },
+    {
+      "epoch": 17.48,
+      "grad_norm": 4.474719047546387,
+      "learning_rate": 1.933253996289433e-06,
+      "loss": 0.9768,
+      "step": 58440
+    },
+    {
+      "epoch": 17.49,
+      "grad_norm": 2.249441146850586,
+      "learning_rate": 1.930989205331357e-06,
+      "loss": 0.725,
+      "step": 58445
+    },
+    {
+      "epoch": 17.49,
+      "grad_norm": 3.9831950664520264,
+      "learning_rate": 1.9287256884507747e-06,
+      "loss": 0.8339,
+      "step": 58450
+    },
+    {
+      "epoch": 17.49,
+      "grad_norm": 3.666069269180298,
+      "learning_rate": 1.9264634457726914e-06,
+      "loss": 0.9266,
+      "step": 58455
+    },
+    {
+      "epoch": 17.49,
+      "grad_norm": 3.5630712509155273,
+      "learning_rate": 1.924202477422046e-06,
+      "loss": 0.8352,
+      "step": 58460
+    },
+    {
+      "epoch": 17.49,
+      "grad_norm": 2.659298896789551,
+      "learning_rate": 1.921942783523711e-06,
+      "loss": 0.942,
+      "step": 58465
+    },
+    {
+      "epoch": 17.49,
+      "grad_norm": 4.289942741394043,
+      "learning_rate": 1.9196843642024896e-06,
+      "loss": 0.9033,
+      "step": 58470
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 5.508975505828857,
+      "learning_rate": 1.9174272195831096e-06,
+      "loss": 1.015,
+      "step": 58475
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 3.3930578231811523,
+      "learning_rate": 1.915171349790232e-06,
+      "loss": 0.9336,
+      "step": 58480
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 2.193418264389038,
+      "learning_rate": 1.9129167549484435e-06,
+      "loss": 0.9792,
+      "step": 58485
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 1.5776050090789795,
+      "learning_rate": 1.9106634351822665e-06,
+      "loss": 1.0228,
+      "step": 58490
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 1.6216871738433838,
+      "learning_rate": 1.9084113906161464e-06,
+      "loss": 0.9904,
+      "step": 58495
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 2.6556360721588135,
+      "learning_rate": 1.9061606213744642e-06,
+      "loss": 1.0049,
+      "step": 58500
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 3.8837389945983887,
+      "learning_rate": 1.9039111275815257e-06,
+      "loss": 0.8822,
+      "step": 58505
+    },
+    {
+      "epoch": 17.51,
+      "grad_norm": 2.5499298572540283,
+      "learning_rate": 1.9016629093615706e-06,
+      "loss": 1.0682,
+      "step": 58510
+    },
+    {
+      "epoch": 17.51,
+      "grad_norm": 4.589527606964111,
+      "learning_rate": 1.8994159668387662e-06,
+      "loss": 0.8225,
+      "step": 58515
+    },
+    {
+      "epoch": 17.51,
+      "grad_norm": 2.39145827293396,
+      "learning_rate": 1.8971703001371965e-06,
+      "loss": 0.7831,
+      "step": 58520
+    },
+    {
+      "epoch": 17.51,
+      "grad_norm": 2.863334894180298,
+      "learning_rate": 1.8949259093809068e-06,
+      "loss": 0.8574,
+      "step": 58525
+    },
+    {
+      "epoch": 17.51,
+      "grad_norm": 1.6067497730255127,
+      "learning_rate": 1.892682794693834e-06,
+      "loss": 0.9947,
+      "step": 58530
+    },
+    {
+      "epoch": 17.51,
+      "grad_norm": 2.9973134994506836,
+      "learning_rate": 1.8904409561998793e-06,
+      "loss": 0.8852,
+      "step": 58535
+    },
+    {
+      "epoch": 17.51,
+      "grad_norm": 12.899645805358887,
+      "learning_rate": 1.8882003940228431e-06,
+      "loss": 1.015,
+      "step": 58540
+    },
+    {
+      "epoch": 17.52,
+      "grad_norm": 4.770505428314209,
+      "learning_rate": 1.8859611082864826e-06,
+      "loss": 1.1247,
+      "step": 58545
+    },
+    {
+      "epoch": 17.52,
+      "grad_norm": 1.5795246362686157,
+      "learning_rate": 1.8837230991144622e-06,
+      "loss": 1.0028,
+      "step": 58550
+    },
+    {
+      "epoch": 17.52,
+      "grad_norm": 3.8196096420288086,
+      "learning_rate": 1.8814863666303833e-06,
+      "loss": 0.9865,
+      "step": 58555
+    },
+    {
+      "epoch": 17.52,
+      "grad_norm": 1.8591629266738892,
+      "learning_rate": 1.879250910957786e-06,
+      "loss": 0.8811,
+      "step": 58560
+    },
+    {
+      "epoch": 17.52,
+      "grad_norm": 2.9709808826446533,
+      "learning_rate": 1.877016732220127e-06,
+      "loss": 0.9346,
+      "step": 58565
+    },
+    {
+      "epoch": 17.52,
+      "grad_norm": 4.240139961242676,
+      "learning_rate": 1.8747838305407967e-06,
+      "loss": 1.0878,
+      "step": 58570
+    },
+    {
+      "epoch": 17.52,
+      "grad_norm": 3.321091890335083,
+      "learning_rate": 1.872552206043121e-06,
+      "loss": 0.9013,
+      "step": 58575
+    },
+    {
+      "epoch": 17.53,
+      "grad_norm": 2.4288010597229004,
+      "learning_rate": 1.8703218588503463e-06,
+      "loss": 0.8679,
+      "step": 58580
+    },
+    {
+      "epoch": 17.53,
+      "grad_norm": 3.094264507293701,
+      "learning_rate": 1.8680927890856515e-06,
+      "loss": 1.1084,
+      "step": 58585
+    },
+    {
+      "epoch": 17.53,
+      "grad_norm": 2.697632312774658,
+      "learning_rate": 1.8658649968721492e-06,
+      "loss": 1.087,
+      "step": 58590
+    },
+    {
+      "epoch": 17.53,
+      "grad_norm": 1.5430232286453247,
+      "learning_rate": 1.8636384823328773e-06,
+      "loss": 0.7268,
+      "step": 58595
+    },
+    {
+      "epoch": 17.53,
+      "grad_norm": 3.243159532546997,
+      "learning_rate": 1.861413245590804e-06,
+      "loss": 0.9764,
+      "step": 58600
+    },
+    {
+      "epoch": 17.53,
+      "grad_norm": 2.758650541305542,
+      "learning_rate": 1.8591892867688226e-06,
+      "loss": 1.088,
+      "step": 58605
+    },
+    {
+      "epoch": 17.54,
+      "grad_norm": 3.265185594558716,
+      "learning_rate": 1.8569666059897655e-06,
+      "loss": 0.9399,
+      "step": 58610
+    },
+    {
+      "epoch": 17.54,
+      "grad_norm": 2.0844991207122803,
+      "learning_rate": 1.8547452033763873e-06,
+      "loss": 1.0002,
+      "step": 58615
+    },
+    {
+      "epoch": 17.54,
+      "grad_norm": 3.4934003353118896,
+      "learning_rate": 1.852525079051376e-06,
+      "loss": 0.8137,
+      "step": 58620
+    },
+    {
+      "epoch": 17.54,
+      "grad_norm": 3.0585551261901855,
+      "learning_rate": 1.8503062331373362e-06,
+      "loss": 0.9706,
+      "step": 58625
+    },
+    {
+      "epoch": 17.54,
+      "grad_norm": 2.177954912185669,
+      "learning_rate": 1.8480886657568307e-06,
+      "loss": 0.9228,
+      "step": 58630
+    },
+    {
+      "epoch": 17.54,
+      "grad_norm": 8.818868637084961,
+      "learning_rate": 1.8458723770323122e-06,
+      "loss": 0.8919,
+      "step": 58635
+    },
+    {
+      "epoch": 17.54,
+      "grad_norm": 4.750818252563477,
+      "learning_rate": 1.8436573670862045e-06,
+      "loss": 0.9735,
+      "step": 58640
+    },
+    {
+      "epoch": 17.55,
+      "grad_norm": 4.335287094116211,
+      "learning_rate": 1.8414436360408265e-06,
+      "loss": 0.9834,
+      "step": 58645
+    },
+    {
+      "epoch": 17.55,
+      "grad_norm": 3.9974465370178223,
+      "learning_rate": 1.8392311840184445e-06,
+      "loss": 0.9222,
+      "step": 58650
+    },
+    {
+      "epoch": 17.55,
+      "grad_norm": 2.57463002204895,
+      "learning_rate": 1.8370200111412495e-06,
+      "loss": 1.0656,
+      "step": 58655
+    },
+    {
+      "epoch": 17.55,
+      "grad_norm": 3.5472350120544434,
+      "learning_rate": 1.8348101175313632e-06,
+      "loss": 0.8107,
+      "step": 58660
+    },
+    {
+      "epoch": 17.55,
+      "grad_norm": 2.0845704078674316,
+      "learning_rate": 1.8326015033108351e-06,
+      "loss": 0.9723,
+      "step": 58665
+    },
+    {
+      "epoch": 17.55,
+      "grad_norm": 8.588332176208496,
+      "learning_rate": 1.8303941686016402e-06,
+      "loss": 1.0629,
+      "step": 58670
+    },
+    {
+      "epoch": 17.55,
+      "grad_norm": 2.690653085708618,
+      "learning_rate": 1.8281881135257e-06,
+      "loss": 0.9891,
+      "step": 58675
+    },
+    {
+      "epoch": 17.56,
+      "grad_norm": 1.5229490995407104,
+      "learning_rate": 1.8259833382048425e-06,
+      "loss": 0.9973,
+      "step": 58680
+    },
+    {
+      "epoch": 17.56,
+      "grad_norm": 1.4649319648742676,
+      "learning_rate": 1.8237798427608337e-06,
+      "loss": 0.9495,
+      "step": 58685
+    },
+    {
+      "epoch": 17.56,
+      "grad_norm": 4.790569305419922,
+      "learning_rate": 1.8215776273153767e-06,
+      "loss": 0.8056,
+      "step": 58690
+    },
+    {
+      "epoch": 17.56,
+      "grad_norm": 5.313488960266113,
+      "learning_rate": 1.8193766919900934e-06,
+      "loss": 0.9817,
+      "step": 58695
+    },
+    {
+      "epoch": 17.56,
+      "grad_norm": 1.8241521120071411,
+      "learning_rate": 1.8171770369065422e-06,
+      "loss": 0.9368,
+      "step": 58700
+    },
+    {
+      "epoch": 17.56,
+      "grad_norm": 2.6710360050201416,
+      "learning_rate": 1.8149786621862036e-06,
+      "loss": 0.9306,
+      "step": 58705
+    },
+    {
+      "epoch": 17.57,
+      "grad_norm": 3.22537899017334,
+      "learning_rate": 1.8127815679504972e-06,
+      "loss": 0.9387,
+      "step": 58710
+    },
+    {
+      "epoch": 17.57,
+      "grad_norm": 1.9393733739852905,
+      "learning_rate": 1.8105857543207649e-06,
+      "loss": 1.0874,
+      "step": 58715
+    },
+    {
+      "epoch": 17.57,
+      "grad_norm": 4.174792289733887,
+      "learning_rate": 1.808391221418268e-06,
+      "loss": 1.0662,
+      "step": 58720
+    },
+    {
+      "epoch": 17.57,
+      "grad_norm": 4.3584513664245605,
+      "learning_rate": 1.8061979693642262e-06,
+      "loss": 0.9845,
+      "step": 58725
+    },
+    {
+      "epoch": 17.57,
+      "grad_norm": 2.3744728565216064,
+      "learning_rate": 1.8040059982797563e-06,
+      "loss": 0.9638,
+      "step": 58730
+    },
+    {
+      "epoch": 17.57,
+      "grad_norm": 3.663601875305176,
+      "learning_rate": 1.8018153082859313e-06,
+      "loss": 0.9749,
+      "step": 58735
+    },
+    {
+      "epoch": 17.57,
+      "grad_norm": 2.476334810256958,
+      "learning_rate": 1.7996258995037291e-06,
+      "loss": 0.8683,
+      "step": 58740
+    },
+    {
+      "epoch": 17.58,
+      "grad_norm": 2.907877206802368,
+      "learning_rate": 1.797437772054067e-06,
+      "loss": 0.8443,
+      "step": 58745
+    },
+    {
+      "epoch": 17.58,
+      "grad_norm": 2.9625954627990723,
+      "learning_rate": 1.7952509260578092e-06,
+      "loss": 0.942,
+      "step": 58750
+    },
+    {
+      "epoch": 17.58,
+      "grad_norm": 3.8954741954803467,
+      "learning_rate": 1.7930653616357152e-06,
+      "loss": 0.956,
+      "step": 58755
+    },
+    {
+      "epoch": 17.58,
+      "grad_norm": 3.0243756771087646,
+      "learning_rate": 1.7908810789085073e-06,
+      "loss": 1.2546,
+      "step": 58760
+    },
+    {
+      "epoch": 17.58,
+      "grad_norm": 1.2879854440689087,
+      "learning_rate": 1.7886980779968032e-06,
+      "loss": 0.9019,
+      "step": 58765
+    },
+    {
+      "epoch": 17.58,
+      "grad_norm": 5.267662048339844,
+      "learning_rate": 1.786516359021187e-06,
+      "loss": 0.8835,
+      "step": 58770
+    },
+    {
+      "epoch": 17.58,
+      "grad_norm": 1.7943377494812012,
+      "learning_rate": 1.7843359221021372e-06,
+      "loss": 0.9029,
+      "step": 58775
+    },
+    {
+      "epoch": 17.59,
+      "grad_norm": 5.033475399017334,
+      "learning_rate": 1.7821567673600854e-06,
+      "loss": 0.8688,
+      "step": 58780
+    },
+    {
+      "epoch": 17.59,
+      "grad_norm": 3.698159694671631,
+      "learning_rate": 1.7799788949153796e-06,
+      "loss": 1.0448,
+      "step": 58785
+    },
+    {
+      "epoch": 17.59,
+      "grad_norm": 2.3559699058532715,
+      "learning_rate": 1.777802304888304e-06,
+      "loss": 0.7721,
+      "step": 58790
+    },
+    {
+      "epoch": 17.59,
+      "grad_norm": 3.055173873901367,
+      "learning_rate": 1.7756269973990713e-06,
+      "loss": 0.8932,
+      "step": 58795
+    },
+    {
+      "epoch": 17.59,
+      "grad_norm": 2.6839029788970947,
+      "learning_rate": 1.7734529725678184e-06,
+      "loss": 1.1173,
+      "step": 58800
+    },
+    {
+      "epoch": 17.59,
+      "grad_norm": 3.986969470977783,
+      "learning_rate": 1.7712802305146131e-06,
+      "loss": 0.8831,
+      "step": 58805
+    },
+    {
+      "epoch": 17.6,
+      "grad_norm": 3.205456018447876,
+      "learning_rate": 1.7691087713594595e-06,
+      "loss": 0.91,
+      "step": 58810
+    },
+    {
+      "epoch": 17.6,
+      "grad_norm": 3.0267038345336914,
+      "learning_rate": 1.7669385952222755e-06,
+      "loss": 0.9393,
+      "step": 58815
+    },
+    {
+      "epoch": 17.6,
+      "grad_norm": 3.005690336227417,
+      "learning_rate": 1.7647697022229321e-06,
+      "loss": 0.8653,
+      "step": 58820
+    },
+    {
+      "epoch": 17.6,
+      "grad_norm": 3.8134701251983643,
+      "learning_rate": 1.762602092481197e-06,
+      "loss": 1.0195,
+      "step": 58825
+    },
+    {
+      "epoch": 17.6,
+      "grad_norm": 5.009188175201416,
+      "learning_rate": 1.760435766116797e-06,
+      "loss": 0.9841,
+      "step": 58830
+    },
+    {
+      "epoch": 17.6,
+      "grad_norm": 2.0699079036712646,
+      "learning_rate": 1.758270723249375e-06,
+      "loss": 0.953,
+      "step": 58835
+    },
+    {
+      "epoch": 17.6,
+      "grad_norm": 2.2950937747955322,
+      "learning_rate": 1.7561069639985023e-06,
+      "loss": 1.0804,
+      "step": 58840
+    },
+    {
+      "epoch": 17.61,
+      "grad_norm": 1.1012648344039917,
+      "learning_rate": 1.753944488483686e-06,
+      "loss": 0.9962,
+      "step": 58845
+    },
+    {
+      "epoch": 17.61,
+      "grad_norm": 3.6749930381774902,
+      "learning_rate": 1.7517832968243442e-06,
+      "loss": 1.0799,
+      "step": 58850
+    },
+    {
+      "epoch": 17.61,
+      "grad_norm": 3.0792107582092285,
+      "learning_rate": 1.749623389139851e-06,
+      "loss": 0.8246,
+      "step": 58855
+    },
+    {
+      "epoch": 17.61,
+      "grad_norm": 5.0220046043396,
+      "learning_rate": 1.7474647655494863e-06,
+      "loss": 0.8415,
+      "step": 58860
+    },
+    {
+      "epoch": 17.61,
+      "grad_norm": 3.235442876815796,
+      "learning_rate": 1.7453074261724795e-06,
+      "loss": 0.944,
+      "step": 58865
+    },
+    {
+      "epoch": 17.61,
+      "grad_norm": 1.411855936050415,
+      "learning_rate": 1.7431513711279685e-06,
+      "loss": 1.0868,
+      "step": 58870
+    },
+    {
+      "epoch": 17.61,
+      "grad_norm": 1.5485836267471313,
+      "learning_rate": 1.7409966005350331e-06,
+      "loss": 0.8386,
+      "step": 58875
+    },
+    {
+      "epoch": 17.62,
+      "grad_norm": 2.2840311527252197,
+      "learning_rate": 1.7388431145126782e-06,
+      "loss": 1.0442,
+      "step": 58880
+    },
+    {
+      "epoch": 17.62,
+      "grad_norm": 2.2418808937072754,
+      "learning_rate": 1.7366909131798393e-06,
+      "loss": 0.9909,
+      "step": 58885
+    },
+    {
+      "epoch": 17.62,
+      "grad_norm": 2.9602952003479004,
+      "learning_rate": 1.7345399966553821e-06,
+      "loss": 1.0754,
+      "step": 58890
+    },
+    {
+      "epoch": 17.62,
+      "grad_norm": 3.5335211753845215,
+      "learning_rate": 1.7323903650580953e-06,
+      "loss": 0.9721,
+      "step": 58895
+    },
+    {
+      "epoch": 17.62,
+      "grad_norm": 3.265909194946289,
+      "learning_rate": 1.7302420185067058e-06,
+      "loss": 0.7851,
+      "step": 58900
+    },
+    {
+      "epoch": 17.62,
+      "grad_norm": 1.4548457860946655,
+      "learning_rate": 1.7280949571198607e-06,
+      "loss": 1.0308,
+      "step": 58905
+    },
+    {
+      "epoch": 17.63,
+      "grad_norm": 2.820035457611084,
+      "learning_rate": 1.7259491810161399e-06,
+      "loss": 1.0166,
+      "step": 58910
+    },
+    {
+      "epoch": 17.63,
+      "grad_norm": 3.4491255283355713,
+      "learning_rate": 1.7238046903140542e-06,
+      "loss": 0.9413,
+      "step": 58915
+    },
+    {
+      "epoch": 17.63,
+      "grad_norm": 1.849202275276184,
+      "learning_rate": 1.7216614851320451e-06,
+      "loss": 1.1075,
+      "step": 58920
+    },
+    {
+      "epoch": 17.63,
+      "grad_norm": 2.7424187660217285,
+      "learning_rate": 1.7195195655884706e-06,
+      "loss": 1.1669,
+      "step": 58925
+    },
+    {
+      "epoch": 17.63,
+      "grad_norm": 2.260991096496582,
+      "learning_rate": 1.717378931801636e-06,
+      "loss": 1.0069,
+      "step": 58930
+    },
+    {
+      "epoch": 17.63,
+      "grad_norm": 3.9417827129364014,
+      "learning_rate": 1.7152395838897605e-06,
+      "loss": 0.9019,
+      "step": 58935
+    },
+    {
+      "epoch": 17.63,
+      "grad_norm": 3.6205639839172363,
+      "learning_rate": 1.7131015219710029e-06,
+      "loss": 0.974,
+      "step": 58940
+    },
+    {
+      "epoch": 17.64,
+      "grad_norm": 2.234180450439453,
+      "learning_rate": 1.7109647461634348e-06,
+      "loss": 0.9879,
+      "step": 58945
+    },
+    {
+      "epoch": 17.64,
+      "grad_norm": 3.515759229660034,
+      "learning_rate": 1.7088292565850845e-06,
+      "loss": 1.0616,
+      "step": 58950
+    },
+    {
+      "epoch": 17.64,
+      "grad_norm": 1.790371060371399,
+      "learning_rate": 1.7066950533538767e-06,
+      "loss": 0.8907,
+      "step": 58955
+    },
+    {
+      "epoch": 17.64,
+      "grad_norm": 1.755277395248413,
+      "learning_rate": 1.7045621365876952e-06,
+      "loss": 0.9207,
+      "step": 58960
+    },
+    {
+      "epoch": 17.64,
+      "grad_norm": 1.3836179971694946,
+      "learning_rate": 1.702430506404329e-06,
+      "loss": 1.1258,
+      "step": 58965
+    },
+    {
+      "epoch": 17.64,
+      "grad_norm": 2.935575246810913,
+      "learning_rate": 1.7003001629215143e-06,
+      "loss": 0.9072,
+      "step": 58970
+    },
+    {
+      "epoch": 17.64,
+      "grad_norm": 2.9841513633728027,
+      "learning_rate": 1.6981711062568989e-06,
+      "loss": 0.9736,
+      "step": 58975
+    },
+    {
+      "epoch": 17.65,
+      "grad_norm": 8.933850288391113,
+      "learning_rate": 1.696043336528072e-06,
+      "loss": 0.9844,
+      "step": 58980
+    },
+    {
+      "epoch": 17.65,
+      "grad_norm": 1.3362852334976196,
+      "learning_rate": 1.6939168538525508e-06,
+      "loss": 1.0783,
+      "step": 58985
+    },
+    {
+      "epoch": 17.65,
+      "grad_norm": 3.199964761734009,
+      "learning_rate": 1.6917916583477744e-06,
+      "loss": 0.8654,
+      "step": 58990
+    },
+    {
+      "epoch": 17.65,
+      "grad_norm": 2.5024354457855225,
+      "learning_rate": 1.6896677501311186e-06,
+      "loss": 0.9169,
+      "step": 58995
+    },
+    {
+      "epoch": 17.65,
+      "grad_norm": 3.7208096981048584,
+      "learning_rate": 1.6875451293198812e-06,
+      "loss": 0.945,
+      "step": 59000
+    },
+    {
+      "epoch": 17.65,
+      "grad_norm": 6.412079811096191,
+      "learning_rate": 1.6854237960312963e-06,
+      "loss": 1.0159,
+      "step": 59005
+    },
+    {
+      "epoch": 17.66,
+      "grad_norm": 6.456012725830078,
+      "learning_rate": 1.68330375038252e-06,
+      "loss": 0.9998,
+      "step": 59010
+    },
+    {
+      "epoch": 17.66,
+      "grad_norm": 1.0914913415908813,
+      "learning_rate": 1.681184992490642e-06,
+      "loss": 0.9272,
+      "step": 59015
+    },
+    {
+      "epoch": 17.66,
+      "grad_norm": 1.7044358253479004,
+      "learning_rate": 1.6790675224726798e-06,
+      "loss": 1.0046,
+      "step": 59020
+    },
+    {
+      "epoch": 17.66,
+      "grad_norm": 3.6310768127441406,
+      "learning_rate": 1.6769513404455733e-06,
+      "loss": 1.0404,
+      "step": 59025
+    },
+    {
+      "epoch": 17.66,
+      "grad_norm": 1.748637080192566,
+      "learning_rate": 1.6748364465262039e-06,
+      "loss": 0.8425,
+      "step": 59030
+    },
+    {
+      "epoch": 17.66,
+      "grad_norm": 3.776822090148926,
+      "learning_rate": 1.6727228408313782e-06,
+      "loss": 1.1857,
+      "step": 59035
+    },
+    {
+      "epoch": 17.66,
+      "grad_norm": 2.349539279937744,
+      "learning_rate": 1.6706105234778114e-06,
+      "loss": 0.987,
+      "step": 59040
+    },
+    {
+      "epoch": 17.67,
+      "grad_norm": 8.603898048400879,
+      "learning_rate": 1.6684994945821847e-06,
+      "loss": 0.8815,
+      "step": 59045
+    },
+    {
+      "epoch": 17.67,
+      "grad_norm": 3.434723377227783,
+      "learning_rate": 1.6663897542610718e-06,
+      "loss": 0.9244,
+      "step": 59050
+    },
+    {
+      "epoch": 17.67,
+      "grad_norm": 3.597698211669922,
+      "learning_rate": 1.6642813026310072e-06,
+      "loss": 0.7415,
+      "step": 59055
+    },
+    {
+      "epoch": 17.67,
+      "grad_norm": 4.349695205688477,
+      "learning_rate": 1.6621741398084201e-06,
+      "loss": 0.9339,
+      "step": 59060
+    },
+    {
+      "epoch": 17.67,
+      "grad_norm": 1.6020125150680542,
+      "learning_rate": 1.6600682659097062e-06,
+      "loss": 0.9233,
+      "step": 59065
+    },
+    {
+      "epoch": 17.67,
+      "grad_norm": 3.7563843727111816,
+      "learning_rate": 1.6579636810511584e-06,
+      "loss": 1.053,
+      "step": 59070
+    },
+    {
+      "epoch": 17.67,
+      "grad_norm": 1.2841134071350098,
+      "learning_rate": 1.6558603853490145e-06,
+      "loss": 0.9631,
+      "step": 59075
+    },
+    {
+      "epoch": 17.68,
+      "grad_norm": 2.884526014328003,
+      "learning_rate": 1.6537583789194345e-06,
+      "loss": 0.7626,
+      "step": 59080
+    },
+    {
+      "epoch": 17.68,
+      "grad_norm": 1.4702849388122559,
+      "learning_rate": 1.651657661878514e-06,
+      "loss": 0.8606,
+      "step": 59085
+    },
+    {
+      "epoch": 17.68,
+      "grad_norm": 3.296031951904297,
+      "learning_rate": 1.6495582343422688e-06,
+      "loss": 0.9722,
+      "step": 59090
+    },
+    {
+      "epoch": 17.68,
+      "grad_norm": 2.626152753829956,
+      "learning_rate": 1.6474600964266534e-06,
+      "loss": 0.8891,
+      "step": 59095
+    },
+    {
+      "epoch": 17.68,
+      "grad_norm": 6.174269676208496,
+      "learning_rate": 1.6453632482475418e-06,
+      "loss": 1.2104,
+      "step": 59100
+    },
+    {
+      "epoch": 17.68,
+      "grad_norm": 2.552041530609131,
+      "learning_rate": 1.6432676899207439e-06,
+      "loss": 1.1366,
+      "step": 59105
+    },
+    {
+      "epoch": 17.69,
+      "grad_norm": 2.2161214351654053,
+      "learning_rate": 1.6411734215619923e-06,
+      "loss": 0.8592,
+      "step": 59110
+    },
+    {
+      "epoch": 17.69,
+      "grad_norm": 3.290187120437622,
+      "learning_rate": 1.6390804432869527e-06,
+      "loss": 1.1547,
+      "step": 59115
+    },
+    {
+      "epoch": 17.69,
+      "grad_norm": 4.858699798583984,
+      "learning_rate": 1.6369887552112163e-06,
+      "loss": 0.7281,
+      "step": 59120
+    },
+    {
+      "epoch": 17.69,
+      "grad_norm": 3.530651569366455,
+      "learning_rate": 1.6348983574503097e-06,
+      "loss": 0.983,
+      "step": 59125
+    },
+    {
+      "epoch": 17.69,
+      "grad_norm": 3.629481077194214,
+      "learning_rate": 1.6328092501196823e-06,
+      "loss": 0.8718,
+      "step": 59130
+    },
+    {
+      "epoch": 17.69,
+      "grad_norm": 1.404110074043274,
+      "learning_rate": 1.6307214333347032e-06,
+      "loss": 0.9193,
+      "step": 59135
+    },
+    {
+      "epoch": 17.69,
+      "grad_norm": 2.185800313949585,
+      "learning_rate": 1.6286349072106965e-06,
+      "loss": 0.8672,
+      "step": 59140
+    },
+    {
+      "epoch": 17.7,
+      "grad_norm": 3.7186684608459473,
+      "learning_rate": 1.6265496718628815e-06,
+      "loss": 0.9765,
+      "step": 59145
+    },
+    {
+      "epoch": 17.7,
+      "grad_norm": 4.324354648590088,
+      "learning_rate": 1.6244657274064407e-06,
+      "loss": 0.9801,
+      "step": 59150
+    },
+    {
+      "epoch": 17.7,
+      "grad_norm": 3.0012760162353516,
+      "learning_rate": 1.6223830739564544e-06,
+      "loss": 0.9174,
+      "step": 59155
+    },
+    {
+      "epoch": 17.7,
+      "grad_norm": 2.157304286956787,
+      "learning_rate": 1.6203017116279584e-06,
+      "loss": 0.9751,
+      "step": 59160
+    },
+    {
+      "epoch": 17.7,
+      "grad_norm": 1.090956449508667,
+      "learning_rate": 1.618221640535894e-06,
+      "loss": 0.891,
+      "step": 59165
+    },
+    {
+      "epoch": 17.7,
+      "grad_norm": 3.6914825439453125,
+      "learning_rate": 1.6161428607951417e-06,
+      "loss": 1.0343,
+      "step": 59170
+    },
+    {
+      "epoch": 17.7,
+      "grad_norm": 3.7854952812194824,
+      "learning_rate": 1.6140653725205152e-06,
+      "loss": 1.0016,
+      "step": 59175
+    },
+    {
+      "epoch": 17.71,
+      "grad_norm": 4.192695617675781,
+      "learning_rate": 1.6119891758267503e-06,
+      "loss": 1.0018,
+      "step": 59180
+    },
+    {
+      "epoch": 17.71,
+      "grad_norm": 2.511425256729126,
+      "learning_rate": 1.6099142708285108e-06,
+      "loss": 0.9797,
+      "step": 59185
+    },
+    {
+      "epoch": 17.71,
+      "grad_norm": 2.2904865741729736,
+      "learning_rate": 1.6078406576403943e-06,
+      "loss": 0.8575,
+      "step": 59190
+    },
+    {
+      "epoch": 17.71,
+      "grad_norm": 2.1356563568115234,
+      "learning_rate": 1.6057683363769227e-06,
+      "loss": 1.042,
+      "step": 59195
+    },
+    {
+      "epoch": 17.71,
+      "grad_norm": 2.6813127994537354,
+      "learning_rate": 1.6036973071525518e-06,
+      "loss": 0.6984,
+      "step": 59200
+    },
+    {
+      "epoch": 17.71,
+      "grad_norm": 2.049669027328491,
+      "learning_rate": 1.6016275700816568e-06,
+      "loss": 1.1138,
+      "step": 59205
+    },
+    {
+      "epoch": 17.71,
+      "grad_norm": 1.8621686697006226,
+      "learning_rate": 1.599559125278549e-06,
+      "loss": 1.0503,
+      "step": 59210
+    },
+    {
+      "epoch": 17.72,
+      "grad_norm": 2.028589963912964,
+      "learning_rate": 1.5974919728574706e-06,
+      "loss": 0.7726,
+      "step": 59215
+    },
+    {
+      "epoch": 17.72,
+      "grad_norm": 6.469555377960205,
+      "learning_rate": 1.5954261129325825e-06,
+      "loss": 1.0135,
+      "step": 59220
+    },
+    {
+      "epoch": 17.72,
+      "grad_norm": 3.396925449371338,
+      "learning_rate": 1.5933615456179885e-06,
+      "loss": 1.0272,
+      "step": 59225
+    },
+    {
+      "epoch": 17.72,
+      "grad_norm": 3.9737088680267334,
+      "learning_rate": 1.5912982710276968e-06,
+      "loss": 0.975,
+      "step": 59230
+    },
+    {
+      "epoch": 17.72,
+      "grad_norm": 3.608037233352661,
+      "learning_rate": 1.5892362892756779e-06,
+      "loss": 0.9554,
+      "step": 59235
+    },
+    {
+      "epoch": 17.72,
+      "grad_norm": 1.3548368215560913,
+      "learning_rate": 1.587175600475796e-06,
+      "loss": 0.9917,
+      "step": 59240
+    },
+    {
+      "epoch": 17.73,
+      "grad_norm": 5.358870983123779,
+      "learning_rate": 1.5851162047418793e-06,
+      "loss": 1.0518,
+      "step": 59245
+    },
+    {
+      "epoch": 17.73,
+      "grad_norm": 1.2182296514511108,
+      "learning_rate": 1.5830581021876455e-06,
+      "loss": 1.0005,
+      "step": 59250
+    },
+    {
+      "epoch": 17.73,
+      "grad_norm": 2.458919048309326,
+      "learning_rate": 1.5810012929267815e-06,
+      "loss": 1.116,
+      "step": 59255
+    },
+    {
+      "epoch": 17.73,
+      "grad_norm": 2.0615477561950684,
+      "learning_rate": 1.578945777072871e-06,
+      "loss": 0.8852,
+      "step": 59260
+    },
+    {
+      "epoch": 17.73,
+      "grad_norm": 3.5121617317199707,
+      "learning_rate": 1.5768915547394375e-06,
+      "loss": 1.0029,
+      "step": 59265
+    },
+    {
+      "epoch": 17.73,
+      "grad_norm": 2.439354658126831,
+      "learning_rate": 1.57483862603994e-06,
+      "loss": 1.0474,
+      "step": 59270
+    },
+    {
+      "epoch": 17.73,
+      "grad_norm": 1.8715204000473022,
+      "learning_rate": 1.5727869910877547e-06,
+      "loss": 1.1325,
+      "step": 59275
+    },
+    {
+      "epoch": 17.74,
+      "grad_norm": 1.4641121625900269,
+      "learning_rate": 1.5707366499961907e-06,
+      "loss": 0.8829,
+      "step": 59280
+    },
+    {
+      "epoch": 17.74,
+      "grad_norm": 2.544196367263794,
+      "learning_rate": 1.5686876028784907e-06,
+      "loss": 1.0373,
+      "step": 59285
+    },
+    {
+      "epoch": 17.74,
+      "grad_norm": 2.298362970352173,
+      "learning_rate": 1.5666398498478175e-06,
+      "loss": 1.0759,
+      "step": 59290
+    },
+    {
+      "epoch": 17.74,
+      "grad_norm": 3.0841598510742188,
+      "learning_rate": 1.564593391017266e-06,
+      "loss": 0.9647,
+      "step": 59295
+    },
+    {
+      "epoch": 17.74,
+      "grad_norm": 4.1196722984313965,
+      "learning_rate": 1.5625482264998658e-06,
+      "loss": 0.8123,
+      "step": 59300
+    },
+    {
+      "epoch": 17.74,
+      "grad_norm": 1.699238657951355,
+      "learning_rate": 1.5605043564085626e-06,
+      "loss": 1.1025,
+      "step": 59305
+    },
+    {
+      "epoch": 17.74,
+      "grad_norm": 3.654175043106079,
+      "learning_rate": 1.5584617808562407e-06,
+      "loss": 0.9208,
+      "step": 59310
+    },
+    {
+      "epoch": 17.75,
+      "grad_norm": 1.3798152208328247,
+      "learning_rate": 1.5564204999557075e-06,
+      "loss": 0.9044,
+      "step": 59315
+    },
+    {
+      "epoch": 17.75,
+      "grad_norm": 3.55618953704834,
+      "learning_rate": 1.554380513819706e-06,
+      "loss": 1.0186,
+      "step": 59320
+    },
+    {
+      "epoch": 17.75,
+      "grad_norm": 3.153846263885498,
+      "learning_rate": 1.5523418225608904e-06,
+      "loss": 0.9396,
+      "step": 59325
+    },
+    {
+      "epoch": 17.75,
+      "grad_norm": 2.9710476398468018,
+      "learning_rate": 1.5503044262918738e-06,
+      "loss": 0.8835,
+      "step": 59330
+    },
+    {
+      "epoch": 17.75,
+      "grad_norm": 3.084639072418213,
+      "learning_rate": 1.5482683251251572e-06,
+      "loss": 0.8919,
+      "step": 59335
+    },
+    {
+      "epoch": 17.75,
+      "grad_norm": 2.0667994022369385,
+      "learning_rate": 1.5462335191732153e-06,
+      "loss": 1.0935,
+      "step": 59340
+    },
+    {
+      "epoch": 17.76,
+      "grad_norm": 2.048433303833008,
+      "learning_rate": 1.5442000085484077e-06,
+      "loss": 1.0158,
+      "step": 59345
+    },
+    {
+      "epoch": 17.76,
+      "grad_norm": 3.5328614711761475,
+      "learning_rate": 1.5421677933630612e-06,
+      "loss": 0.923,
+      "step": 59350
+    },
+    {
+      "epoch": 17.76,
+      "grad_norm": 1.7786338329315186,
+      "learning_rate": 1.5401368737294003e-06,
+      "loss": 0.9076,
+      "step": 59355
+    },
+    {
+      "epoch": 17.76,
+      "grad_norm": 2.726869583129883,
+      "learning_rate": 1.5381072497595961e-06,
+      "loss": 0.9152,
+      "step": 59360
+    },
+    {
+      "epoch": 17.76,
+      "grad_norm": 3.708962917327881,
+      "learning_rate": 1.5360789215657395e-06,
+      "loss": 0.9758,
+      "step": 59365
+    },
+    {
+      "epoch": 17.76,
+      "grad_norm": 4.803073406219482,
+      "learning_rate": 1.5340518892598548e-06,
+      "loss": 1.1745,
+      "step": 59370
+    },
+    {
+      "epoch": 17.76,
+      "grad_norm": 2.451441764831543,
+      "learning_rate": 1.5320261529538943e-06,
+      "loss": 0.9447,
+      "step": 59375
+    },
+    {
+      "epoch": 17.77,
+      "grad_norm": 6.812480449676514,
+      "learning_rate": 1.530001712759735e-06,
+      "loss": 0.9028,
+      "step": 59380
+    },
+    {
+      "epoch": 17.77,
+      "grad_norm": 6.00867223739624,
+      "learning_rate": 1.5279785687891846e-06,
+      "loss": 0.8863,
+      "step": 59385
+    },
+    {
+      "epoch": 17.77,
+      "grad_norm": 3.013056516647339,
+      "learning_rate": 1.5259567211539788e-06,
+      "loss": 0.9852,
+      "step": 59390
+    },
+    {
+      "epoch": 17.77,
+      "grad_norm": 1.796088457107544,
+      "learning_rate": 1.5239361699657867e-06,
+      "loss": 0.9819,
+      "step": 59395
+    },
+    {
+      "epoch": 17.77,
+      "grad_norm": 4.895589351654053,
+      "learning_rate": 1.5219169153361967e-06,
+      "loss": 0.9859,
+      "step": 59400
+    },
+    {
+      "epoch": 17.77,
+      "grad_norm": 1.5124635696411133,
+      "learning_rate": 1.5198989573767308e-06,
+      "loss": 0.9221,
+      "step": 59405
+    },
+    {
+      "epoch": 17.77,
+      "grad_norm": 2.5901122093200684,
+      "learning_rate": 1.5178822961988387e-06,
+      "loss": 0.8347,
+      "step": 59410
+    },
+    {
+      "epoch": 17.78,
+      "grad_norm": 5.9025397300720215,
+      "learning_rate": 1.5158669319139007e-06,
+      "loss": 0.9824,
+      "step": 59415
+    },
+    {
+      "epoch": 17.78,
+      "grad_norm": 2.827467918395996,
+      "learning_rate": 1.5138528646332195e-06,
+      "loss": 1.1154,
+      "step": 59420
+    },
+    {
+      "epoch": 17.78,
+      "grad_norm": 2.0498061180114746,
+      "learning_rate": 1.5118400944680366e-06,
+      "loss": 0.9927,
+      "step": 59425
+    },
+    {
+      "epoch": 17.78,
+      "grad_norm": 1.4182449579238892,
+      "learning_rate": 1.509828621529505e-06,
+      "loss": 0.8675,
+      "step": 59430
+    },
+    {
+      "epoch": 17.78,
+      "grad_norm": 4.468570232391357,
+      "learning_rate": 1.507818445928727e-06,
+      "loss": 0.9225,
+      "step": 59435
+    },
+    {
+      "epoch": 17.78,
+      "grad_norm": 2.804314374923706,
+      "learning_rate": 1.5058095677767113e-06,
+      "loss": 1.0115,
+      "step": 59440
+    },
+    {
+      "epoch": 17.79,
+      "grad_norm": 3.139026641845703,
+      "learning_rate": 1.5038019871844222e-06,
+      "loss": 0.9551,
+      "step": 59445
+    },
+    {
+      "epoch": 17.79,
+      "grad_norm": 2.992780923843384,
+      "learning_rate": 1.501795704262718e-06,
+      "loss": 1.0013,
+      "step": 59450
+    },
+    {
+      "epoch": 17.79,
+      "grad_norm": 4.901608943939209,
+      "learning_rate": 1.4997907191224152e-06,
+      "loss": 1.04,
+      "step": 59455
+    },
+    {
+      "epoch": 17.79,
+      "grad_norm": 2.3052425384521484,
+      "learning_rate": 1.4977870318742426e-06,
+      "loss": 0.9111,
+      "step": 59460
+    },
+    {
+      "epoch": 17.79,
+      "grad_norm": 3.878119707107544,
+      "learning_rate": 1.495784642628864e-06,
+      "loss": 1.0182,
+      "step": 59465
+    },
+    {
+      "epoch": 17.79,
+      "grad_norm": 3.6295268535614014,
+      "learning_rate": 1.4937835514968663e-06,
+      "loss": 0.9679,
+      "step": 59470
+    },
+    {
+      "epoch": 17.79,
+      "grad_norm": 3.5464370250701904,
+      "learning_rate": 1.491783758588769e-06,
+      "loss": 1.0276,
+      "step": 59475
+    },
+    {
+      "epoch": 17.8,
+      "grad_norm": 5.657863140106201,
+      "learning_rate": 1.4897852640150228e-06,
+      "loss": 1.0774,
+      "step": 59480
+    },
+    {
+      "epoch": 17.8,
+      "grad_norm": 1.0737026929855347,
+      "learning_rate": 1.487788067885998e-06,
+      "loss": 0.9233,
+      "step": 59485
+    },
+    {
+      "epoch": 17.8,
+      "grad_norm": 10.904875755310059,
+      "learning_rate": 1.4857921703119976e-06,
+      "loss": 0.7518,
+      "step": 59490
+    },
+    {
+      "epoch": 17.8,
+      "grad_norm": 2.129427909851074,
+      "learning_rate": 1.483797571403256e-06,
+      "loss": 0.9003,
+      "step": 59495
+    },
+    {
+      "epoch": 17.8,
+      "grad_norm": 2.353217601776123,
+      "learning_rate": 1.4818042712699294e-06,
+      "loss": 1.0227,
+      "step": 59500
+    },
+    {
+      "epoch": 17.8,
+      "grad_norm": 2.1998493671417236,
+      "learning_rate": 1.4798122700221074e-06,
+      "loss": 1.0736,
+      "step": 59505
+    },
+    {
+      "epoch": 17.8,
+      "grad_norm": 5.145125865936279,
+      "learning_rate": 1.4778215677698048e-06,
+      "loss": 0.916,
+      "step": 59510
+    },
+    {
+      "epoch": 17.81,
+      "grad_norm": 1.4517490863800049,
+      "learning_rate": 1.4758321646229672e-06,
+      "loss": 0.8587,
+      "step": 59515
+    },
+    {
+      "epoch": 17.81,
+      "grad_norm": 1.5610160827636719,
+      "learning_rate": 1.4738440606914706e-06,
+      "loss": 1.056,
+      "step": 59520
+    },
+    {
+      "epoch": 17.81,
+      "grad_norm": 7.770084857940674,
+      "learning_rate": 1.4718572560851073e-06,
+      "loss": 0.8253,
+      "step": 59525
+    },
+    {
+      "epoch": 17.81,
+      "grad_norm": 2.7066264152526855,
+      "learning_rate": 1.469871750913618e-06,
+      "loss": 1.0866,
+      "step": 59530
+    },
+    {
+      "epoch": 17.81,
+      "grad_norm": 4.087286949157715,
+      "learning_rate": 1.4678875452866448e-06,
+      "loss": 1.0194,
+      "step": 59535
+    },
+    {
+      "epoch": 17.81,
+      "grad_norm": 1.208440899848938,
+      "learning_rate": 1.4659046393137893e-06,
+      "loss": 0.9872,
+      "step": 59540
+    },
+    {
+      "epoch": 17.82,
+      "grad_norm": 3.085111141204834,
+      "learning_rate": 1.4639230331045528e-06,
+      "loss": 0.8561,
+      "step": 59545
+    },
+    {
+      "epoch": 17.82,
+      "grad_norm": 1.8525588512420654,
+      "learning_rate": 1.4619427267683894e-06,
+      "loss": 0.8675,
+      "step": 59550
+    },
+    {
+      "epoch": 17.82,
+      "grad_norm": 3.9756102561950684,
+      "learning_rate": 1.4599637204146587e-06,
+      "loss": 0.8729,
+      "step": 59555
+    },
+    {
+      "epoch": 17.82,
+      "grad_norm": 3.060441017150879,
+      "learning_rate": 1.4579860141526624e-06,
+      "loss": 0.9836,
+      "step": 59560
+    },
+    {
+      "epoch": 17.82,
+      "grad_norm": 2.611800193786621,
+      "learning_rate": 1.4560096080916325e-06,
+      "loss": 0.9945,
+      "step": 59565
+    },
+    {
+      "epoch": 17.82,
+      "grad_norm": 3.2961905002593994,
+      "learning_rate": 1.454034502340712e-06,
+      "loss": 0.9673,
+      "step": 59570
+    },
+    {
+      "epoch": 17.82,
+      "grad_norm": 3.6914517879486084,
+      "learning_rate": 1.452060697009e-06,
+      "loss": 0.9933,
+      "step": 59575
+    },
+    {
+      "epoch": 17.83,
+      "grad_norm": 1.3202792406082153,
+      "learning_rate": 1.4500881922054926e-06,
+      "loss": 1.0733,
+      "step": 59580
+    },
+    {
+      "epoch": 17.83,
+      "grad_norm": 3.6315410137176514,
+      "learning_rate": 1.4481169880391388e-06,
+      "loss": 1.0638,
+      "step": 59585
+    },
+    {
+      "epoch": 17.83,
+      "grad_norm": 3.1422181129455566,
+      "learning_rate": 1.4461470846188013e-06,
+      "loss": 1.0285,
+      "step": 59590
+    },
+    {
+      "epoch": 17.83,
+      "grad_norm": 2.8239080905914307,
+      "learning_rate": 1.4441784820532766e-06,
+      "loss": 0.9607,
+      "step": 59595
+    },
+    {
+      "epoch": 17.83,
+      "grad_norm": 2.704333782196045,
+      "learning_rate": 1.4422111804512916e-06,
+      "loss": 0.9963,
+      "step": 59600
+    },
+    {
+      "epoch": 17.83,
+      "grad_norm": 2.927246332168579,
+      "learning_rate": 1.4402451799214927e-06,
+      "loss": 1.0692,
+      "step": 59605
+    },
+    {
+      "epoch": 17.83,
+      "grad_norm": 3.1765265464782715,
+      "learning_rate": 1.438280480572468e-06,
+      "loss": 0.8938,
+      "step": 59610
+    },
+    {
+      "epoch": 17.84,
+      "grad_norm": 4.268390655517578,
+      "learning_rate": 1.4363170825127226e-06,
+      "loss": 1.0705,
+      "step": 59615
+    },
+    {
+      "epoch": 17.84,
+      "grad_norm": 16.1699275970459,
+      "learning_rate": 1.4343549858506833e-06,
+      "loss": 0.9742,
+      "step": 59620
+    },
+    {
+      "epoch": 17.84,
+      "grad_norm": 2.6375677585601807,
+      "learning_rate": 1.432394190694733e-06,
+      "loss": 0.9267,
+      "step": 59625
+    },
+    {
+      "epoch": 17.84,
+      "grad_norm": 1.9000338315963745,
+      "learning_rate": 1.4304346971531434e-06,
+      "loss": 0.994,
+      "step": 59630
+    },
+    {
+      "epoch": 17.84,
+      "grad_norm": 3.127405881881714,
+      "learning_rate": 1.4284765053341532e-06,
+      "loss": 1.0222,
+      "step": 59635
+    },
+    {
+      "epoch": 17.84,
+      "grad_norm": 2.3889312744140625,
+      "learning_rate": 1.4265196153459032e-06,
+      "loss": 0.7955,
+      "step": 59640
+    },
+    {
+      "epoch": 17.85,
+      "grad_norm": 2.0697555541992188,
+      "learning_rate": 1.424564027296471e-06,
+      "loss": 1.0505,
+      "step": 59645
+    },
+    {
+      "epoch": 17.85,
+      "grad_norm": 3.0797669887542725,
+      "learning_rate": 1.4226097412938677e-06,
+      "loss": 0.9022,
+      "step": 59650
+    },
+    {
+      "epoch": 17.85,
+      "grad_norm": 3.074493885040283,
+      "learning_rate": 1.4206567574460121e-06,
+      "loss": 0.9206,
+      "step": 59655
+    },
+    {
+      "epoch": 17.85,
+      "grad_norm": 2.3382043838500977,
+      "learning_rate": 1.4187050758607824e-06,
+      "loss": 1.0961,
+      "step": 59660
+    },
+    {
+      "epoch": 17.85,
+      "grad_norm": 4.023479461669922,
+      "learning_rate": 1.4167546966459527e-06,
+      "loss": 0.9317,
+      "step": 59665
+    },
+    {
+      "epoch": 17.85,
+      "grad_norm": 1.3876618146896362,
+      "learning_rate": 1.415195331053179e-06,
+      "loss": 1.0919,
+      "step": 59670
+    },
+    {
+      "epoch": 17.85,
+      "grad_norm": 2.543372869491577,
+      "learning_rate": 1.4132472963764826e-06,
+      "loss": 0.8302,
+      "step": 59675
+    },
+    {
+      "epoch": 17.86,
+      "grad_norm": 3.604238271713257,
+      "learning_rate": 1.411300564371626e-06,
+      "loss": 1.0208,
+      "step": 59680
+    },
+    {
+      "epoch": 17.86,
+      "grad_norm": 3.2941768169403076,
+      "learning_rate": 1.4093551351461149e-06,
+      "loss": 1.0286,
+      "step": 59685
+    },
+    {
+      "epoch": 17.86,
+      "grad_norm": 3.470616579055786,
+      "learning_rate": 1.407411008807405e-06,
+      "loss": 1.003,
+      "step": 59690
+    },
+    {
+      "epoch": 17.86,
+      "grad_norm": 4.124062538146973,
+      "learning_rate": 1.4054681854628548e-06,
+      "loss": 0.9581,
+      "step": 59695
+    },
+    {
+      "epoch": 17.86,
+      "grad_norm": 2.6687521934509277,
+      "learning_rate": 1.4035266652197786e-06,
+      "loss": 0.9555,
+      "step": 59700
+    },
+    {
+      "epoch": 17.86,
+      "grad_norm": 6.374357223510742,
+      "learning_rate": 1.4015864481853963e-06,
+      "loss": 0.9077,
+      "step": 59705
+    },
+    {
+      "epoch": 17.86,
+      "grad_norm": 5.034076690673828,
+      "learning_rate": 1.3996475344668636e-06,
+      "loss": 0.9172,
+      "step": 59710
+    },
+    {
+      "epoch": 17.87,
+      "grad_norm": 5.536380767822266,
+      "learning_rate": 1.3977099241712643e-06,
+      "loss": 1.0173,
+      "step": 59715
+    },
+    {
+      "epoch": 17.87,
+      "grad_norm": 2.3674838542938232,
+      "learning_rate": 1.3957736174056157e-06,
+      "loss": 1.0644,
+      "step": 59720
+    },
+    {
+      "epoch": 17.87,
+      "grad_norm": 6.831422805786133,
+      "learning_rate": 1.3938386142768517e-06,
+      "loss": 1.0125,
+      "step": 59725
+    },
+    {
+      "epoch": 17.87,
+      "grad_norm": 2.7988386154174805,
+      "learning_rate": 1.3919049148918478e-06,
+      "loss": 0.9512,
+      "step": 59730
+    },
+    {
+      "epoch": 17.87,
+      "grad_norm": 1.8349326848983765,
+      "learning_rate": 1.3899725193573937e-06,
+      "loss": 0.8805,
+      "step": 59735
+    },
+    {
+      "epoch": 17.87,
+      "grad_norm": 4.807056903839111,
+      "learning_rate": 1.3880414277802178e-06,
+      "loss": 0.9358,
+      "step": 59740
+    },
+    {
+      "epoch": 17.88,
+      "grad_norm": 1.473981261253357,
+      "learning_rate": 1.3861116402669683e-06,
+      "loss": 1.0964,
+      "step": 59745
+    },
+    {
+      "epoch": 17.88,
+      "grad_norm": 1.3169115781784058,
+      "learning_rate": 1.384183156924229e-06,
+      "loss": 1.0681,
+      "step": 59750
+    },
+    {
+      "epoch": 17.88,
+      "grad_norm": 3.194916009902954,
+      "learning_rate": 1.382255977858507e-06,
+      "loss": 1.1128,
+      "step": 59755
+    },
+    {
+      "epoch": 17.88,
+      "grad_norm": 3.641432285308838,
+      "learning_rate": 1.3803301031762362e-06,
+      "loss": 0.9769,
+      "step": 59760
+    },
+    {
+      "epoch": 17.88,
+      "grad_norm": 1.8899619579315186,
+      "learning_rate": 1.378405532983787e-06,
+      "loss": 1.1635,
+      "step": 59765
+    },
+    {
+      "epoch": 17.88,
+      "grad_norm": 2.5514962673187256,
+      "learning_rate": 1.3764822673874411e-06,
+      "loss": 1.0953,
+      "step": 59770
+    },
+    {
+      "epoch": 17.88,
+      "grad_norm": 2.5154950618743896,
+      "learning_rate": 1.3745603064934303e-06,
+      "loss": 0.8913,
+      "step": 59775
+    },
+    {
+      "epoch": 17.89,
+      "grad_norm": 2.303077220916748,
+      "learning_rate": 1.3726396504078892e-06,
+      "loss": 0.865,
+      "step": 59780
+    },
+    {
+      "epoch": 17.89,
+      "grad_norm": 4.337673664093018,
+      "learning_rate": 1.3707202992369078e-06,
+      "loss": 0.8839,
+      "step": 59785
+    },
+    {
+      "epoch": 17.89,
+      "grad_norm": 3.780333995819092,
+      "learning_rate": 1.3688022530864763e-06,
+      "loss": 0.8552,
+      "step": 59790
+    },
+    {
+      "epoch": 17.89,
+      "grad_norm": 1.5258413553237915,
+      "learning_rate": 1.3668855120625406e-06,
+      "loss": 0.8518,
+      "step": 59795
+    },
+    {
+      "epoch": 17.89,
+      "grad_norm": 2.358565330505371,
+      "learning_rate": 1.3649700762709495e-06,
+      "loss": 0.954,
+      "step": 59800
+    },
+    {
+      "epoch": 17.89,
+      "grad_norm": 1.6710665225982666,
+      "learning_rate": 1.363055945817493e-06,
+      "loss": 1.056,
+      "step": 59805
+    },
+    {
+      "epoch": 17.89,
+      "grad_norm": 2.9221675395965576,
+      "learning_rate": 1.3611431208078896e-06,
+      "loss": 1.0636,
+      "step": 59810
+    },
+    {
+      "epoch": 17.9,
+      "grad_norm": 2.2900736331939697,
+      "learning_rate": 1.359231601347777e-06,
+      "loss": 0.9176,
+      "step": 59815
+    },
+    {
+      "epoch": 17.9,
+      "grad_norm": 2.580756187438965,
+      "learning_rate": 1.3573213875427316e-06,
+      "loss": 0.9621,
+      "step": 59820
+    },
+    {
+      "epoch": 17.9,
+      "grad_norm": 3.829488515853882,
+      "learning_rate": 1.3554124794982498e-06,
+      "loss": 0.919,
+      "step": 59825
+    },
+    {
+      "epoch": 17.9,
+      "grad_norm": 2.805250883102417,
+      "learning_rate": 1.3535048773197611e-06,
+      "loss": 0.9749,
+      "step": 59830
+    },
+    {
+      "epoch": 17.9,
+      "grad_norm": 1.3262838125228882,
+      "learning_rate": 1.3515985811126174e-06,
+      "loss": 1.0604,
+      "step": 59835
+    },
+    {
+      "epoch": 17.9,
+      "grad_norm": 3.3138718605041504,
+      "learning_rate": 1.349693590982104e-06,
+      "loss": 1.0974,
+      "step": 59840
+    },
+    {
+      "epoch": 17.9,
+      "grad_norm": 1.5498145818710327,
+      "learning_rate": 1.347789907033431e-06,
+      "loss": 0.9018,
+      "step": 59845
+    },
+    {
+      "epoch": 17.91,
+      "grad_norm": 2.336205244064331,
+      "learning_rate": 1.3458875293717366e-06,
+      "loss": 1.0362,
+      "step": 59850
+    },
+    {
+      "epoch": 17.91,
+      "grad_norm": 2.3710501194000244,
+      "learning_rate": 1.343986458102084e-06,
+      "loss": 0.9496,
+      "step": 59855
+    },
+    {
+      "epoch": 17.91,
+      "grad_norm": 5.016022205352783,
+      "learning_rate": 1.3420866933294752e-06,
+      "loss": 1.1354,
+      "step": 59860
+    },
+    {
+      "epoch": 17.91,
+      "grad_norm": 5.362601280212402,
+      "learning_rate": 1.3401882351588207e-06,
+      "loss": 0.7814,
+      "step": 59865
+    },
+    {
+      "epoch": 17.91,
+      "grad_norm": 1.2273238897323608,
+      "learning_rate": 1.338291083694984e-06,
+      "loss": 1.0238,
+      "step": 59870
+    },
+    {
+      "epoch": 17.91,
+      "grad_norm": 2.9525585174560547,
+      "learning_rate": 1.3363952390427286e-06,
+      "loss": 0.8545,
+      "step": 59875
+    },
+    {
+      "epoch": 17.92,
+      "grad_norm": 2.0316078662872314,
+      "learning_rate": 1.3345007013067763e-06,
+      "loss": 1.1211,
+      "step": 59880
+    },
+    {
+      "epoch": 17.92,
+      "grad_norm": 3.018373727798462,
+      "learning_rate": 1.3326074705917401e-06,
+      "loss": 0.8393,
+      "step": 59885
+    },
+    {
+      "epoch": 17.92,
+      "grad_norm": 1.7973732948303223,
+      "learning_rate": 1.3307155470022038e-06,
+      "loss": 0.9589,
+      "step": 59890
+    },
+    {
+      "epoch": 17.92,
+      "grad_norm": 1.8492597341537476,
+      "learning_rate": 1.3288249306426387e-06,
+      "loss": 1.0118,
+      "step": 59895
+    },
+    {
+      "epoch": 17.92,
+      "grad_norm": 2.7371528148651123,
+      "learning_rate": 1.3269356216174645e-06,
+      "loss": 1.0925,
+      "step": 59900
+    },
+    {
+      "epoch": 17.92,
+      "grad_norm": 3.346501350402832,
+      "learning_rate": 1.3250476200310363e-06,
+      "loss": 1.0044,
+      "step": 59905
+    },
+    {
+      "epoch": 17.92,
+      "grad_norm": 2.5383148193359375,
+      "learning_rate": 1.3231609259876126e-06,
+      "loss": 0.7532,
+      "step": 59910
+    },
+    {
+      "epoch": 17.93,
+      "grad_norm": 2.1317331790924072,
+      "learning_rate": 1.3212755395914074e-06,
+      "loss": 1.1127,
+      "step": 59915
+    },
+    {
+      "epoch": 17.93,
+      "grad_norm": 1.8375227451324463,
+      "learning_rate": 1.319391460946534e-06,
+      "loss": 1.0262,
+      "step": 59920
+    },
+    {
+      "epoch": 17.93,
+      "grad_norm": 1.335551142692566,
+      "learning_rate": 1.3175086901570626e-06,
+      "loss": 0.9814,
+      "step": 59925
+    },
+    {
+      "epoch": 17.93,
+      "grad_norm": 1.4829717874526978,
+      "learning_rate": 1.3156272273269682e-06,
+      "loss": 0.8828,
+      "step": 59930
+    },
+    {
+      "epoch": 17.93,
+      "grad_norm": 1.677444577217102,
+      "learning_rate": 1.313747072560162e-06,
+      "loss": 0.8892,
+      "step": 59935
+    },
+    {
+      "epoch": 17.93,
+      "grad_norm": 5.412178993225098,
+      "learning_rate": 1.3118682259604832e-06,
+      "loss": 0.9235,
+      "step": 59940
+    },
+    {
+      "epoch": 17.93,
+      "grad_norm": 2.3385939598083496,
+      "learning_rate": 1.3099906876317014e-06,
+      "loss": 1.1063,
+      "step": 59945
+    },
+    {
+      "epoch": 17.94,
+      "grad_norm": 10.752957344055176,
+      "learning_rate": 1.3081144576775089e-06,
+      "loss": 0.9414,
+      "step": 59950
+    },
+    {
+      "epoch": 17.94,
+      "grad_norm": 5.151787281036377,
+      "learning_rate": 1.306239536201531e-06,
+      "loss": 0.8772,
+      "step": 59955
+    },
+    {
+      "epoch": 17.94,
+      "grad_norm": 1.5891977548599243,
+      "learning_rate": 1.3043659233073102e-06,
+      "loss": 0.951,
+      "step": 59960
+    },
+    {
+      "epoch": 17.94,
+      "grad_norm": 3.044320583343506,
+      "learning_rate": 1.3024936190983355e-06,
+      "loss": 1.0105,
+      "step": 59965
+    },
+    {
+      "epoch": 17.94,
+      "grad_norm": 4.976190090179443,
+      "learning_rate": 1.3006226236779996e-06,
+      "loss": 0.9063,
+      "step": 59970
+    },
+    {
+      "epoch": 17.94,
+      "grad_norm": 3.8189163208007812,
+      "learning_rate": 1.2987529371496498e-06,
+      "loss": 0.9465,
+      "step": 59975
+    },
+    {
+      "epoch": 17.95,
+      "grad_norm": 1.8560311794281006,
+      "learning_rate": 1.296884559616529e-06,
+      "loss": 1.015,
+      "step": 59980
+    },
+    {
+      "epoch": 17.95,
+      "grad_norm": 4.682796478271484,
+      "learning_rate": 1.2950174911818407e-06,
+      "loss": 0.8412,
+      "step": 59985
+    },
+    {
+      "epoch": 17.95,
+      "grad_norm": 1.3414361476898193,
+      "learning_rate": 1.2931517319487024e-06,
+      "loss": 0.9081,
+      "step": 59990
+    },
+    {
+      "epoch": 17.95,
+      "grad_norm": 3.5960850715637207,
+      "learning_rate": 1.2912872820201427e-06,
+      "loss": 1.1309,
+      "step": 59995
+    },
+    {
+      "epoch": 17.95,
+      "grad_norm": 2.8373448848724365,
+      "learning_rate": 1.2894241414991488e-06,
+      "loss": 0.9147,
+      "step": 60000
+    },
+    {
+      "epoch": 17.95,
+      "grad_norm": 2.71930193901062,
+      "learning_rate": 1.2875623104886104e-06,
+      "loss": 0.9289,
+      "step": 60005
+    },
+    {
+      "epoch": 17.95,
+      "grad_norm": 2.6316206455230713,
+      "learning_rate": 1.2857017890913619e-06,
+      "loss": 0.8153,
+      "step": 60010
+    },
+    {
+      "epoch": 17.96,
+      "grad_norm": 5.35076379776001,
+      "learning_rate": 1.2838425774101466e-06,
+      "loss": 0.8993,
+      "step": 60015
+    },
+    {
+      "epoch": 17.96,
+      "grad_norm": 2.4414889812469482,
+      "learning_rate": 1.2819846755476622e-06,
+      "loss": 0.993,
+      "step": 60020
+    },
+    {
+      "epoch": 17.96,
+      "grad_norm": 2.973344087600708,
+      "learning_rate": 1.2801280836065077e-06,
+      "loss": 1.1067,
+      "step": 60025
+    },
+    {
+      "epoch": 17.96,
+      "grad_norm": 3.7870774269104004,
+      "learning_rate": 1.2782728016892231e-06,
+      "loss": 1.0017,
+      "step": 60030
+    },
+    {
+      "epoch": 17.96,
+      "grad_norm": 1.356503963470459,
+      "learning_rate": 1.2764188298982738e-06,
+      "loss": 0.9037,
+      "step": 60035
+    },
+    {
+      "epoch": 17.96,
+      "grad_norm": 1.4454678297042847,
+      "learning_rate": 1.2745661683360554e-06,
+      "loss": 1.0378,
+      "step": 60040
+    },
+    {
+      "epoch": 17.96,
+      "grad_norm": 1.8113714456558228,
+      "learning_rate": 1.2727148171048836e-06,
+      "loss": 0.8887,
+      "step": 60045
+    },
+    {
+      "epoch": 17.97,
+      "grad_norm": 4.557557582855225,
+      "learning_rate": 1.2708647763070152e-06,
+      "loss": 0.8867,
+      "step": 60050
+    },
+    {
+      "epoch": 17.97,
+      "grad_norm": 2.7099125385284424,
+      "learning_rate": 1.2690160460446105e-06,
+      "loss": 0.9419,
+      "step": 60055
+    },
+    {
+      "epoch": 17.97,
+      "grad_norm": 4.08752965927124,
+      "learning_rate": 1.2671686264197874e-06,
+      "loss": 0.8125,
+      "step": 60060
+    },
+    {
+      "epoch": 17.97,
+      "grad_norm": 1.8687279224395752,
+      "learning_rate": 1.265322517534573e-06,
+      "loss": 0.7801,
+      "step": 60065
+    },
+    {
+      "epoch": 17.97,
+      "grad_norm": 4.718087673187256,
+      "learning_rate": 1.2634777194909242e-06,
+      "loss": 0.9194,
+      "step": 60070
+    },
+    {
+      "epoch": 17.97,
+      "grad_norm": 2.097789764404297,
+      "learning_rate": 1.2616342323907293e-06,
+      "loss": 0.9047,
+      "step": 60075
+    },
+    {
+      "epoch": 17.98,
+      "grad_norm": 2.0857861042022705,
+      "learning_rate": 1.2597920563357984e-06,
+      "loss": 0.9454,
+      "step": 60080
+    },
+    {
+      "epoch": 17.98,
+      "grad_norm": 1.1246609687805176,
+      "learning_rate": 1.2579511914278807e-06,
+      "loss": 1.0526,
+      "step": 60085
+    },
+    {
+      "epoch": 17.98,
+      "grad_norm": 3.467363119125366,
+      "learning_rate": 1.2561116377686338e-06,
+      "loss": 0.8471,
+      "step": 60090
+    },
+    {
+      "epoch": 17.98,
+      "grad_norm": 1.099082350730896,
+      "learning_rate": 1.254273395459668e-06,
+      "loss": 1.0249,
+      "step": 60095
+    },
+    {
+      "epoch": 17.98,
+      "grad_norm": 3.2694308757781982,
+      "learning_rate": 1.252436464602494e-06,
+      "loss": 1.0626,
+      "step": 60100
+    },
+    {
+      "epoch": 17.98,
+      "grad_norm": 5.15191125869751,
+      "learning_rate": 1.2506008452985746e-06,
+      "loss": 0.9525,
+      "step": 60105
+    },
+    {
+      "epoch": 17.98,
+      "grad_norm": 4.068530082702637,
+      "learning_rate": 1.248766537649279e-06,
+      "loss": 1.0582,
+      "step": 60110
+    },
+    {
+      "epoch": 17.99,
+      "grad_norm": 1.9388269186019897,
+      "learning_rate": 1.2469335417559236e-06,
+      "loss": 0.8859,
+      "step": 60115
+    },
+    {
+      "epoch": 17.99,
+      "grad_norm": 3.2138922214508057,
+      "learning_rate": 1.245101857719738e-06,
+      "loss": 0.856,
+      "step": 60120
+    },
+    {
+      "epoch": 17.99,
+      "grad_norm": 2.7871532440185547,
+      "learning_rate": 1.2432714856418837e-06,
+      "loss": 1.0063,
+      "step": 60125
+    },
+    {
+      "epoch": 17.99,
+      "grad_norm": 4.555139541625977,
+      "learning_rate": 1.241442425623454e-06,
+      "loss": 0.7705,
+      "step": 60130
+    },
+    {
+      "epoch": 17.99,
+      "grad_norm": 4.78867244720459,
+      "learning_rate": 1.2396146777654604e-06,
+      "loss": 1.1051,
+      "step": 60135
+    },
+    {
+      "epoch": 17.99,
+      "grad_norm": 2.8767669200897217,
+      "learning_rate": 1.2377882421688526e-06,
+      "loss": 0.8788,
+      "step": 60140
+    },
+    {
+      "epoch": 17.99,
+      "grad_norm": 5.313528060913086,
+      "learning_rate": 1.2359631189344994e-06,
+      "loss": 1.0437,
+      "step": 60145
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 3.7219114303588867,
+      "learning_rate": 1.234139308163204e-06,
+      "loss": 0.8678,
+      "step": 60150
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 3.0899860858917236,
+      "learning_rate": 1.2323168099556886e-06,
+      "loss": 0.7888,
+      "step": 60155
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 1.9980682134628296,
+      "learning_rate": 1.2304956244126115e-06,
+      "loss": 1.1034,
+      "step": 60160
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 2.616227626800537,
+      "learning_rate": 1.228675751634556e-06,
+      "loss": 1.0001,
+      "step": 60165
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 8.774314880371094,
+      "learning_rate": 1.226857191722028e-06,
+      "loss": 0.9797,
+      "step": 60170
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 1.6300761699676514,
+      "learning_rate": 1.2250399447754663e-06,
+      "loss": 1.0554,
+      "step": 60175
+    },
+    {
+      "epoch": 18.01,
+      "grad_norm": 2.3184926509857178,
+      "learning_rate": 1.223224010895238e-06,
+      "loss": 0.9453,
+      "step": 60180
+    },
+    {
+      "epoch": 18.01,
+      "grad_norm": 5.169106483459473,
+      "learning_rate": 1.2214093901816297e-06,
+      "loss": 0.7603,
+      "step": 60185
+    },
+    {
+      "epoch": 18.01,
+      "grad_norm": 2.690866231918335,
+      "learning_rate": 1.2195960827348696e-06,
+      "loss": 0.9451,
+      "step": 60190
+    },
+    {
+      "epoch": 18.01,
+      "grad_norm": 3.5127060413360596,
+      "learning_rate": 1.2177840886550911e-06,
+      "loss": 1.0737,
+      "step": 60195
+    },
+    {
+      "epoch": 18.01,
+      "grad_norm": 2.8311686515808105,
+      "learning_rate": 1.215973408042384e-06,
+      "loss": 0.7565,
+      "step": 60200
+    },
+    {
+      "epoch": 18.01,
+      "grad_norm": 3.0393755435943604,
+      "learning_rate": 1.2141640409967403e-06,
+      "loss": 0.9511,
+      "step": 60205
+    },
+    {
+      "epoch": 18.01,
+      "grad_norm": 1.882336139678955,
+      "learning_rate": 1.2123559876180968e-06,
+      "loss": 1.1201,
+      "step": 60210
+    },
+    {
+      "epoch": 18.02,
+      "grad_norm": 3.4361493587493896,
+      "learning_rate": 1.2105492480063013e-06,
+      "loss": 0.5961,
+      "step": 60215
+    },
+    {
+      "epoch": 18.02,
+      "grad_norm": 2.833055019378662,
+      "learning_rate": 1.2087438222611463e-06,
+      "loss": 1.1693,
+      "step": 60220
+    },
+    {
+      "epoch": 18.02,
+      "grad_norm": 3.044955015182495,
+      "learning_rate": 1.2069397104823381e-06,
+      "loss": 1.0318,
+      "step": 60225
+    },
+    {
+      "epoch": 18.02,
+      "grad_norm": 3.623894691467285,
+      "learning_rate": 1.2051369127695195e-06,
+      "loss": 1.0285,
+      "step": 60230
+    },
+    {
+      "epoch": 18.02,
+      "grad_norm": 1.35172700881958,
+      "learning_rate": 1.2033354292222547e-06,
+      "loss": 0.9498,
+      "step": 60235
+    },
+    {
+      "epoch": 18.02,
+      "grad_norm": 1.692931056022644,
+      "learning_rate": 1.2015352599400399e-06,
+      "loss": 0.8844,
+      "step": 60240
+    },
+    {
+      "epoch": 18.02,
+      "grad_norm": 2.4801394939422607,
+      "learning_rate": 1.1997364050222947e-06,
+      "loss": 0.8478,
+      "step": 60245
+    },
+    {
+      "epoch": 18.03,
+      "grad_norm": 2.9214184284210205,
+      "learning_rate": 1.1979388645683682e-06,
+      "loss": 0.8282,
+      "step": 60250
+    },
+    {
+      "epoch": 18.03,
+      "grad_norm": 1.6741719245910645,
+      "learning_rate": 1.1961426386775386e-06,
+      "loss": 1.0689,
+      "step": 60255
+    },
+    {
+      "epoch": 18.03,
+      "grad_norm": 2.074756383895874,
+      "learning_rate": 1.1943477274490079e-06,
+      "loss": 0.9789,
+      "step": 60260
+    },
+    {
+      "epoch": 18.03,
+      "grad_norm": 4.593778610229492,
+      "learning_rate": 1.1925541309819071e-06,
+      "loss": 1.0066,
+      "step": 60265
+    },
+    {
+      "epoch": 18.03,
+      "grad_norm": 4.825080394744873,
+      "learning_rate": 1.1907618493752965e-06,
+      "loss": 1.0119,
+      "step": 60270
+    },
+    {
+      "epoch": 18.03,
+      "grad_norm": 2.745469093322754,
+      "learning_rate": 1.1889708827281604e-06,
+      "loss": 1.0391,
+      "step": 60275
+    },
+    {
+      "epoch": 18.04,
+      "grad_norm": 3.292245864868164,
+      "learning_rate": 1.1871812311394115e-06,
+      "loss": 0.8295,
+      "step": 60280
+    },
+    {
+      "epoch": 18.04,
+      "grad_norm": 2.7830917835235596,
+      "learning_rate": 1.1853928947078957e-06,
+      "loss": 1.0809,
+      "step": 60285
+    },
+    {
+      "epoch": 18.04,
+      "grad_norm": 3.622986316680908,
+      "learning_rate": 1.1836058735323674e-06,
+      "loss": 1.0924,
+      "step": 60290
+    },
+    {
+      "epoch": 18.04,
+      "grad_norm": 2.329106569290161,
+      "learning_rate": 1.1818201677115393e-06,
+      "loss": 0.9369,
+      "step": 60295
+    },
+    {
+      "epoch": 18.04,
+      "grad_norm": 2.0072081089019775,
+      "learning_rate": 1.1800357773440212e-06,
+      "loss": 0.8325,
+      "step": 60300
+    },
+    {
+      "epoch": 18.04,
+      "grad_norm": 2.3265275955200195,
+      "learning_rate": 1.1782527025283706e-06,
+      "loss": 0.827,
+      "step": 60305
+    },
+    {
+      "epoch": 18.04,
+      "grad_norm": 4.844279766082764,
+      "learning_rate": 1.1764709433630589e-06,
+      "loss": 0.88,
+      "step": 60310
+    },
+    {
+      "epoch": 18.05,
+      "grad_norm": 2.5498387813568115,
+      "learning_rate": 1.174690499946496e-06,
+      "loss": 0.9118,
+      "step": 60315
+    },
+    {
+      "epoch": 18.05,
+      "grad_norm": 2.94402813911438,
+      "learning_rate": 1.1729113723770114e-06,
+      "loss": 1.0359,
+      "step": 60320
+    },
+    {
+      "epoch": 18.05,
+      "grad_norm": 8.016414642333984,
+      "learning_rate": 1.1711335607528629e-06,
+      "loss": 0.8906,
+      "step": 60325
+    },
+    {
+      "epoch": 18.05,
+      "grad_norm": 1.8241318464279175,
+      "learning_rate": 1.1693570651722414e-06,
+      "loss": 0.9315,
+      "step": 60330
+    },
+    {
+      "epoch": 18.05,
+      "grad_norm": 3.5544328689575195,
+      "learning_rate": 1.167581885733257e-06,
+      "loss": 0.8926,
+      "step": 60335
+    },
+    {
+      "epoch": 18.05,
+      "grad_norm": 4.292209148406982,
+      "learning_rate": 1.1658080225339541e-06,
+      "loss": 0.9165,
+      "step": 60340
+    },
+    {
+      "epoch": 18.05,
+      "grad_norm": 2.8831968307495117,
+      "learning_rate": 1.1640354756722981e-06,
+      "loss": 1.1162,
+      "step": 60345
+    },
+    {
+      "epoch": 18.06,
+      "grad_norm": 3.58367657661438,
+      "learning_rate": 1.1622642452461862e-06,
+      "loss": 0.8821,
+      "step": 60350
+    },
+    {
+      "epoch": 18.06,
+      "grad_norm": 2.5088016986846924,
+      "learning_rate": 1.1604943313534455e-06,
+      "loss": 1.0714,
+      "step": 60355
+    },
+    {
+      "epoch": 18.06,
+      "grad_norm": 4.4307780265808105,
+      "learning_rate": 1.15872573409182e-06,
+      "loss": 1.0852,
+      "step": 60360
+    },
+    {
+      "epoch": 18.06,
+      "grad_norm": 3.551413059234619,
+      "learning_rate": 1.1569584535589929e-06,
+      "loss": 0.9419,
+      "step": 60365
+    },
+    {
+      "epoch": 18.06,
+      "grad_norm": 3.166933059692383,
+      "learning_rate": 1.1551924898525634e-06,
+      "loss": 0.9873,
+      "step": 60370
+    },
+    {
+      "epoch": 18.06,
+      "grad_norm": 4.157015800476074,
+      "learning_rate": 1.1534278430700707e-06,
+      "loss": 0.9274,
+      "step": 60375
+    },
+    {
+      "epoch": 18.06,
+      "grad_norm": 3.1233348846435547,
+      "learning_rate": 1.1516645133089727e-06,
+      "loss": 0.9412,
+      "step": 60380
+    },
+    {
+      "epoch": 18.07,
+      "grad_norm": 1.6911544799804688,
+      "learning_rate": 1.1499025006666498e-06,
+      "loss": 0.8977,
+      "step": 60385
+    },
+    {
+      "epoch": 18.07,
+      "grad_norm": 2.505157470703125,
+      "learning_rate": 1.148141805240427e-06,
+      "loss": 1.0199,
+      "step": 60390
+    },
+    {
+      "epoch": 18.07,
+      "grad_norm": 3.155808925628662,
+      "learning_rate": 1.1463824271275319e-06,
+      "loss": 1.0877,
+      "step": 60395
+    },
+    {
+      "epoch": 18.07,
+      "grad_norm": 2.7817635536193848,
+      "learning_rate": 1.1446243664251482e-06,
+      "loss": 0.8983,
+      "step": 60400
+    },
+    {
+      "epoch": 18.07,
+      "grad_norm": 3.2476534843444824,
+      "learning_rate": 1.1428676232303592e-06,
+      "loss": 1.1019,
+      "step": 60405
+    },
+    {
+      "epoch": 18.07,
+      "grad_norm": 2.456259250640869,
+      "learning_rate": 1.1411121976401956e-06,
+      "loss": 1.1568,
+      "step": 60410
+    },
+    {
+      "epoch": 18.08,
+      "grad_norm": 3.9711551666259766,
+      "learning_rate": 1.139358089751605e-06,
+      "loss": 0.9174,
+      "step": 60415
+    },
+    {
+      "epoch": 18.08,
+      "grad_norm": 4.814849376678467,
+      "learning_rate": 1.137605299661465e-06,
+      "loss": 0.8198,
+      "step": 60420
+    },
+    {
+      "epoch": 18.08,
+      "grad_norm": 2.4193475246429443,
+      "learning_rate": 1.135853827466582e-06,
+      "loss": 0.9368,
+      "step": 60425
+    },
+    {
+      "epoch": 18.08,
+      "grad_norm": 4.294716835021973,
+      "learning_rate": 1.1341036732636868e-06,
+      "loss": 0.932,
+      "step": 60430
+    },
+    {
+      "epoch": 18.08,
+      "grad_norm": 2.2720484733581543,
+      "learning_rate": 1.1323548371494352e-06,
+      "loss": 1.0362,
+      "step": 60435
+    },
+    {
+      "epoch": 18.08,
+      "grad_norm": 2.023895740509033,
+      "learning_rate": 1.1306073192204198e-06,
+      "loss": 0.9637,
+      "step": 60440
+    },
+    {
+      "epoch": 18.08,
+      "grad_norm": 1.7245546579360962,
+      "learning_rate": 1.1288611195731518e-06,
+      "loss": 1.1067,
+      "step": 60445
+    },
+    {
+      "epoch": 18.09,
+      "grad_norm": 3.7294325828552246,
+      "learning_rate": 1.1271162383040685e-06,
+      "loss": 0.9301,
+      "step": 60450
+    },
+    {
+      "epoch": 18.09,
+      "grad_norm": 3.503115177154541,
+      "learning_rate": 1.125372675509545e-06,
+      "loss": 1.0133,
+      "step": 60455
+    },
+    {
+      "epoch": 18.09,
+      "grad_norm": 3.752464532852173,
+      "learning_rate": 1.1236304312858686e-06,
+      "loss": 1.0687,
+      "step": 60460
+    },
+    {
+      "epoch": 18.09,
+      "grad_norm": 1.6993839740753174,
+      "learning_rate": 1.1218895057292678e-06,
+      "loss": 0.8668,
+      "step": 60465
+    },
+    {
+      "epoch": 18.09,
+      "grad_norm": 1.8143806457519531,
+      "learning_rate": 1.1201498989358878e-06,
+      "loss": 0.925,
+      "step": 60470
+    },
+    {
+      "epoch": 18.09,
+      "grad_norm": 1.6880947351455688,
+      "learning_rate": 1.118411611001813e-06,
+      "loss": 1.0283,
+      "step": 60475
+    },
+    {
+      "epoch": 18.09,
+      "grad_norm": 3.8792147636413574,
+      "learning_rate": 1.116674642023033e-06,
+      "loss": 0.9818,
+      "step": 60480
+    },
+    {
+      "epoch": 18.1,
+      "grad_norm": 3.6388609409332275,
+      "learning_rate": 1.1149389920954933e-06,
+      "loss": 0.9869,
+      "step": 60485
+    },
+    {
+      "epoch": 18.1,
+      "grad_norm": 2.7923548221588135,
+      "learning_rate": 1.1132046613150398e-06,
+      "loss": 1.0708,
+      "step": 60490
+    },
+    {
+      "epoch": 18.1,
+      "grad_norm": 3.55141282081604,
+      "learning_rate": 1.1114716497774708e-06,
+      "loss": 0.918,
+      "step": 60495
+    },
+    {
+      "epoch": 18.1,
+      "grad_norm": 5.645336151123047,
+      "learning_rate": 1.1097399575784845e-06,
+      "loss": 0.8854,
+      "step": 60500
+    },
+    {
+      "epoch": 18.1,
+      "grad_norm": 4.349913120269775,
+      "learning_rate": 1.108009584813735e-06,
+      "loss": 0.9067,
+      "step": 60505
+    },
+    {
+      "epoch": 18.1,
+      "grad_norm": 3.4893198013305664,
+      "learning_rate": 1.1062805315787794e-06,
+      "loss": 0.8483,
+      "step": 60510
+    },
+    {
+      "epoch": 18.11,
+      "grad_norm": 4.366122245788574,
+      "learning_rate": 1.1045527979691134e-06,
+      "loss": 0.9011,
+      "step": 60515
+    },
+    {
+      "epoch": 18.11,
+      "grad_norm": 4.734853267669678,
+      "learning_rate": 1.1028263840801577e-06,
+      "loss": 0.9712,
+      "step": 60520
+    },
+    {
+      "epoch": 18.11,
+      "grad_norm": 3.5593693256378174,
+      "learning_rate": 1.1011012900072616e-06,
+      "loss": 0.9062,
+      "step": 60525
+    },
+    {
+      "epoch": 18.11,
+      "grad_norm": 3.4199602603912354,
+      "learning_rate": 1.099377515845698e-06,
+      "loss": 0.8207,
+      "step": 60530
+    },
+    {
+      "epoch": 18.11,
+      "grad_norm": 1.6151660680770874,
+      "learning_rate": 1.097655061690675e-06,
+      "loss": 0.8984,
+      "step": 60535
+    },
+    {
+      "epoch": 18.11,
+      "grad_norm": 8.06431770324707,
+      "learning_rate": 1.0959339276373132e-06,
+      "loss": 1.0251,
+      "step": 60540
+    },
+    {
+      "epoch": 18.11,
+      "grad_norm": 1.1797876358032227,
+      "learning_rate": 1.0942141137806782e-06,
+      "loss": 1.0986,
+      "step": 60545
+    },
+    {
+      "epoch": 18.12,
+      "grad_norm": 2.656918525695801,
+      "learning_rate": 1.0924956202157443e-06,
+      "loss": 1.0583,
+      "step": 60550
+    },
+    {
+      "epoch": 18.12,
+      "grad_norm": 4.957900524139404,
+      "learning_rate": 1.09077844703743e-06,
+      "loss": 0.981,
+      "step": 60555
+    },
+    {
+      "epoch": 18.12,
+      "grad_norm": 2.982581853866577,
+      "learning_rate": 1.0890625943405703e-06,
+      "loss": 1.0173,
+      "step": 60560
+    },
+    {
+      "epoch": 18.12,
+      "grad_norm": 3.549740791320801,
+      "learning_rate": 1.0873480622199283e-06,
+      "loss": 0.8502,
+      "step": 60565
+    },
+    {
+      "epoch": 18.12,
+      "grad_norm": 2.2221314907073975,
+      "learning_rate": 1.085634850770198e-06,
+      "loss": 1.0024,
+      "step": 60570
+    },
+    {
+      "epoch": 18.12,
+      "grad_norm": 4.6512675285339355,
+      "learning_rate": 1.0839229600859952e-06,
+      "loss": 1.1475,
+      "step": 60575
+    },
+    {
+      "epoch": 18.12,
+      "grad_norm": 1.9820324182510376,
+      "learning_rate": 1.082212390261872e-06,
+      "loss": 1.0201,
+      "step": 60580
+    },
+    {
+      "epoch": 18.13,
+      "grad_norm": 1.954908013343811,
+      "learning_rate": 1.0805031413922888e-06,
+      "loss": 0.9867,
+      "step": 60585
+    },
+    {
+      "epoch": 18.13,
+      "grad_norm": 2.869485855102539,
+      "learning_rate": 1.078795213571665e-06,
+      "loss": 1.0383,
+      "step": 60590
+    },
+    {
+      "epoch": 18.13,
+      "grad_norm": 5.065496444702148,
+      "learning_rate": 1.0770886068943082e-06,
+      "loss": 0.8963,
+      "step": 60595
+    },
+    {
+      "epoch": 18.13,
+      "grad_norm": 2.9921464920043945,
+      "learning_rate": 1.0753833214544872e-06,
+      "loss": 0.9474,
+      "step": 60600
+    },
+    {
+      "epoch": 18.13,
+      "grad_norm": 1.8409907817840576,
+      "learning_rate": 1.0736793573463744e-06,
+      "loss": 1.0217,
+      "step": 60605
+    },
+    {
+      "epoch": 18.13,
+      "grad_norm": 4.104241371154785,
+      "learning_rate": 1.0719767146640803e-06,
+      "loss": 0.884,
+      "step": 60610
+    },
+    {
+      "epoch": 18.14,
+      "grad_norm": 3.4805054664611816,
+      "learning_rate": 1.0702753935016408e-06,
+      "loss": 0.8448,
+      "step": 60615
+    },
+    {
+      "epoch": 18.14,
+      "grad_norm": 1.8355894088745117,
+      "learning_rate": 1.068575393953017e-06,
+      "loss": 0.9421,
+      "step": 60620
+    },
+    {
+      "epoch": 18.14,
+      "grad_norm": 2.950094699859619,
+      "learning_rate": 1.0668767161121e-06,
+      "loss": 0.9723,
+      "step": 60625
+    },
+    {
+      "epoch": 18.14,
+      "grad_norm": 4.175784587860107,
+      "learning_rate": 1.0651793600727018e-06,
+      "loss": 0.7554,
+      "step": 60630
+    },
+    {
+      "epoch": 18.14,
+      "grad_norm": 1.5484530925750732,
+      "learning_rate": 1.0634833259285743e-06,
+      "loss": 0.9147,
+      "step": 60635
+    },
+    {
+      "epoch": 18.14,
+      "grad_norm": 2.7090961933135986,
+      "learning_rate": 1.061788613773379e-06,
+      "loss": 0.9327,
+      "step": 60640
+    },
+    {
+      "epoch": 18.14,
+      "grad_norm": 2.208608627319336,
+      "learning_rate": 1.0600952237007162e-06,
+      "loss": 1.1507,
+      "step": 60645
+    },
+    {
+      "epoch": 18.15,
+      "grad_norm": 3.7526814937591553,
+      "learning_rate": 1.0584031558041108e-06,
+      "loss": 1.0072,
+      "step": 60650
+    },
+    {
+      "epoch": 18.15,
+      "grad_norm": 4.413444519042969,
+      "learning_rate": 1.0567124101770103e-06,
+      "loss": 0.9917,
+      "step": 60655
+    },
+    {
+      "epoch": 18.15,
+      "grad_norm": 5.183726787567139,
+      "learning_rate": 1.0550229869127987e-06,
+      "loss": 1.0578,
+      "step": 60660
+    },
+    {
+      "epoch": 18.15,
+      "grad_norm": 1.6198534965515137,
+      "learning_rate": 1.0533348861047814e-06,
+      "loss": 0.9569,
+      "step": 60665
+    },
+    {
+      "epoch": 18.15,
+      "grad_norm": 3.9471218585968018,
+      "learning_rate": 1.0516481078461786e-06,
+      "loss": 0.9391,
+      "step": 60670
+    },
+    {
+      "epoch": 18.15,
+      "grad_norm": 2.755523204803467,
+      "learning_rate": 1.049962652230166e-06,
+      "loss": 0.8292,
+      "step": 60675
+    },
+    {
+      "epoch": 18.15,
+      "grad_norm": 3.8161582946777344,
+      "learning_rate": 1.0482785193498156e-06,
+      "loss": 0.8619,
+      "step": 60680
+    },
+    {
+      "epoch": 18.16,
+      "grad_norm": 2.392561435699463,
+      "learning_rate": 1.046595709298151e-06,
+      "loss": 1.0699,
+      "step": 60685
+    },
+    {
+      "epoch": 18.16,
+      "grad_norm": 2.5720813274383545,
+      "learning_rate": 1.044914222168103e-06,
+      "loss": 0.9893,
+      "step": 60690
+    },
+    {
+      "epoch": 18.16,
+      "grad_norm": 2.1463463306427,
+      "learning_rate": 1.0432340580525473e-06,
+      "loss": 0.9201,
+      "step": 60695
+    },
+    {
+      "epoch": 18.16,
+      "grad_norm": 2.0431580543518066,
+      "learning_rate": 1.0415552170442682e-06,
+      "loss": 0.8918,
+      "step": 60700
+    },
+    {
+      "epoch": 18.16,
+      "grad_norm": 3.739319086074829,
+      "learning_rate": 1.0398776992359943e-06,
+      "loss": 0.9097,
+      "step": 60705
+    },
+    {
+      "epoch": 18.16,
+      "grad_norm": 3.897839069366455,
+      "learning_rate": 1.038201504720368e-06,
+      "loss": 0.8448,
+      "step": 60710
+    },
+    {
+      "epoch": 18.17,
+      "grad_norm": 2.4379286766052246,
+      "learning_rate": 1.0365266335899626e-06,
+      "loss": 1.0746,
+      "step": 60715
+    },
+    {
+      "epoch": 18.17,
+      "grad_norm": 5.488431453704834,
+      "learning_rate": 1.0348530859372902e-06,
+      "loss": 0.9621,
+      "step": 60720
+    },
+    {
+      "epoch": 18.17,
+      "grad_norm": 2.122537136077881,
+      "learning_rate": 1.0331808618547628e-06,
+      "loss": 1.044,
+      "step": 60725
+    },
+    {
+      "epoch": 18.17,
+      "grad_norm": 1.5338541269302368,
+      "learning_rate": 1.0315099614347511e-06,
+      "loss": 0.8884,
+      "step": 60730
+    },
+    {
+      "epoch": 18.17,
+      "grad_norm": 2.4747703075408936,
+      "learning_rate": 1.0298403847695286e-06,
+      "loss": 0.818,
+      "step": 60735
+    },
+    {
+      "epoch": 18.17,
+      "grad_norm": 5.456902980804443,
+      "learning_rate": 1.028172131951305e-06,
+      "loss": 0.985,
+      "step": 60740
+    },
+    {
+      "epoch": 18.17,
+      "grad_norm": 3.0986452102661133,
+      "learning_rate": 1.0265052030722171e-06,
+      "loss": 1.0442,
+      "step": 60745
+    },
+    {
+      "epoch": 18.18,
+      "grad_norm": 8.805809020996094,
+      "learning_rate": 1.024839598224331e-06,
+      "loss": 0.7961,
+      "step": 60750
+    },
+    {
+      "epoch": 18.18,
+      "grad_norm": 3.979262351989746,
+      "learning_rate": 1.0231753174996278e-06,
+      "loss": 1.0433,
+      "step": 60755
+    },
+    {
+      "epoch": 18.18,
+      "grad_norm": 3.9714319705963135,
+      "learning_rate": 1.0215123609900373e-06,
+      "loss": 1.0825,
+      "step": 60760
+    },
+    {
+      "epoch": 18.18,
+      "grad_norm": 2.5382189750671387,
+      "learning_rate": 1.019850728787386e-06,
+      "loss": 0.6957,
+      "step": 60765
+    },
+    {
+      "epoch": 18.18,
+      "grad_norm": 2.689236640930176,
+      "learning_rate": 1.0181904209834586e-06,
+      "loss": 0.9993,
+      "step": 60770
+    },
+    {
+      "epoch": 18.18,
+      "grad_norm": 2.2523913383483887,
+      "learning_rate": 1.016531437669943e-06,
+      "loss": 1.0457,
+      "step": 60775
+    },
+    {
+      "epoch": 18.18,
+      "grad_norm": 2.1695034503936768,
+      "learning_rate": 1.0148737789384689e-06,
+      "loss": 0.945,
+      "step": 60780
+    },
+    {
+      "epoch": 18.19,
+      "grad_norm": 4.191242694854736,
+      "learning_rate": 1.013217444880582e-06,
+      "loss": 1.0195,
+      "step": 60785
+    },
+    {
+      "epoch": 18.19,
+      "grad_norm": 2.4551188945770264,
+      "learning_rate": 1.011562435587765e-06,
+      "loss": 1.0352,
+      "step": 60790
+    },
+    {
+      "epoch": 18.19,
+      "grad_norm": 5.271094799041748,
+      "learning_rate": 1.0099087511514227e-06,
+      "loss": 1.0489,
+      "step": 60795
+    },
+    {
+      "epoch": 18.19,
+      "grad_norm": 3.048891067504883,
+      "learning_rate": 1.008256391662879e-06,
+      "loss": 1.0531,
+      "step": 60800
+    },
+    {
+      "epoch": 18.19,
+      "grad_norm": 3.0242154598236084,
+      "learning_rate": 1.0066053572133999e-06,
+      "loss": 1.1194,
+      "step": 60805
+    },
+    {
+      "epoch": 18.19,
+      "grad_norm": 3.903019905090332,
+      "learning_rate": 1.004955647894165e-06,
+      "loss": 0.9434,
+      "step": 60810
+    },
+    {
+      "epoch": 18.2,
+      "grad_norm": 2.762580394744873,
+      "learning_rate": 1.003307263796291e-06,
+      "loss": 0.8196,
+      "step": 60815
+    },
+    {
+      "epoch": 18.2,
+      "grad_norm": 22.09214210510254,
+      "learning_rate": 1.0016602050108099e-06,
+      "loss": 0.8785,
+      "step": 60820
+    },
+    {
+      "epoch": 18.2,
+      "grad_norm": 4.526973247528076,
+      "learning_rate": 1.0000144716286935e-06,
+      "loss": 0.7305,
+      "step": 60825
+    },
+    {
+      "epoch": 18.2,
+      "grad_norm": 4.591505527496338,
+      "learning_rate": 9.983700637408305e-07,
+      "loss": 1.1613,
+      "step": 60830
+    },
+    {
+      "epoch": 18.2,
+      "grad_norm": 6.078658580780029,
+      "learning_rate": 9.967269814380425e-07,
+      "loss": 0.9575,
+      "step": 60835
+    },
+    {
+      "epoch": 18.2,
+      "grad_norm": 3.495609998703003,
+      "learning_rate": 9.950852248110709e-07,
+      "loss": 1.0076,
+      "step": 60840
+    },
+    {
+      "epoch": 18.2,
+      "grad_norm": 1.9880722761154175,
+      "learning_rate": 9.934447939505904e-07,
+      "loss": 0.89,
+      "step": 60845
+    },
+    {
+      "epoch": 18.21,
+      "grad_norm": 4.335904121398926,
+      "learning_rate": 9.918056889472006e-07,
+      "loss": 1.0215,
+      "step": 60850
+    },
+    {
+      "epoch": 18.21,
+      "grad_norm": 1.3856970071792603,
+      "learning_rate": 9.901679098914291e-07,
+      "loss": 1.0306,
+      "step": 60855
+    },
+    {
+      "epoch": 18.21,
+      "grad_norm": 2.3149681091308594,
+      "learning_rate": 9.885314568737258e-07,
+      "loss": 0.9417,
+      "step": 60860
+    },
+    {
+      "epoch": 18.21,
+      "grad_norm": 2.511589527130127,
+      "learning_rate": 9.868963299844742e-07,
+      "loss": 0.9716,
+      "step": 60865
+    },
+    {
+      "epoch": 18.21,
+      "grad_norm": 1.1993108987808228,
+      "learning_rate": 9.85262529313974e-07,
+      "loss": 1.0695,
+      "step": 60870
+    },
+    {
+      "epoch": 18.21,
+      "grad_norm": 3.9712471961975098,
+      "learning_rate": 9.836300549524642e-07,
+      "loss": 1.0194,
+      "step": 60875
+    },
+    {
+      "epoch": 18.21,
+      "grad_norm": 1.5070483684539795,
+      "learning_rate": 9.81998906990103e-07,
+      "loss": 0.9899,
+      "step": 60880
+    },
+    {
+      "epoch": 18.22,
+      "grad_norm": 1.7017158269882202,
+      "learning_rate": 9.803690855169772e-07,
+      "loss": 1.0185,
+      "step": 60885
+    },
+    {
+      "epoch": 18.22,
+      "grad_norm": 2.1888837814331055,
+      "learning_rate": 9.787405906231006e-07,
+      "loss": 1.0552,
+      "step": 60890
+    },
+    {
+      "epoch": 18.22,
+      "grad_norm": 2.574962615966797,
+      "learning_rate": 9.771134223984097e-07,
+      "loss": 0.8768,
+      "step": 60895
+    },
+    {
+      "epoch": 18.22,
+      "grad_norm": 2.792041301727295,
+      "learning_rate": 9.754875809327769e-07,
+      "loss": 0.8991,
+      "step": 60900
+    },
+    {
+      "epoch": 18.22,
+      "grad_norm": 2.1841979026794434,
+      "learning_rate": 9.738630663159886e-07,
+      "loss": 1.1429,
+      "step": 60905
+    },
+    {
+      "epoch": 18.22,
+      "grad_norm": 3.5888936519622803,
+      "learning_rate": 9.722398786377762e-07,
+      "loss": 0.8153,
+      "step": 60910
+    },
+    {
+      "epoch": 18.23,
+      "grad_norm": 5.765038967132568,
+      "learning_rate": 9.706180179877733e-07,
+      "loss": 0.9749,
+      "step": 60915
+    },
+    {
+      "epoch": 18.23,
+      "grad_norm": 4.348241329193115,
+      "learning_rate": 9.68997484455564e-07,
+      "loss": 1.0241,
+      "step": 60920
+    },
+    {
+      "epoch": 18.23,
+      "grad_norm": 1.9215142726898193,
+      "learning_rate": 9.673782781306428e-07,
+      "loss": 1.1308,
+      "step": 60925
+    },
+    {
+      "epoch": 18.23,
+      "grad_norm": 2.1733293533325195,
+      "learning_rate": 9.657603991024416e-07,
+      "loss": 0.8968,
+      "step": 60930
+    },
+    {
+      "epoch": 18.23,
+      "grad_norm": 3.1698646545410156,
+      "learning_rate": 9.641438474603081e-07,
+      "loss": 0.9575,
+      "step": 60935
+    },
+    {
+      "epoch": 18.23,
+      "grad_norm": 4.437852382659912,
+      "learning_rate": 9.625286232935266e-07,
+      "loss": 1.0785,
+      "step": 60940
+    },
+    {
+      "epoch": 18.23,
+      "grad_norm": 2.4240994453430176,
+      "learning_rate": 9.60914726691306e-07,
+      "loss": 0.9774,
+      "step": 60945
+    },
+    {
+      "epoch": 18.24,
+      "grad_norm": 4.216778755187988,
+      "learning_rate": 9.593021577427752e-07,
+      "loss": 1.0754,
+      "step": 60950
+    },
+    {
+      "epoch": 18.24,
+      "grad_norm": 5.071732044219971,
+      "learning_rate": 9.57690916536999e-07,
+      "loss": 1.0031,
+      "step": 60955
+    },
+    {
+      "epoch": 18.24,
+      "grad_norm": 5.060432434082031,
+      "learning_rate": 9.56081003162962e-07,
+      "loss": 0.9962,
+      "step": 60960
+    },
+    {
+      "epoch": 18.24,
+      "grad_norm": 4.500984191894531,
+      "learning_rate": 9.544724177095787e-07,
+      "loss": 0.9923,
+      "step": 60965
+    },
+    {
+      "epoch": 18.24,
+      "grad_norm": 3.332535743713379,
+      "learning_rate": 9.528651602656896e-07,
+      "loss": 1.1458,
+      "step": 60970
+    },
+    {
+      "epoch": 18.24,
+      "grad_norm": 2.8214833736419678,
+      "learning_rate": 9.512592309200625e-07,
+      "loss": 1.0929,
+      "step": 60975
+    },
+    {
+      "epoch": 18.24,
+      "grad_norm": 3.44423770904541,
+      "learning_rate": 9.496546297613901e-07,
+      "loss": 1.0643,
+      "step": 60980
+    },
+    {
+      "epoch": 18.25,
+      "grad_norm": 4.527592182159424,
+      "learning_rate": 9.480513568782962e-07,
+      "loss": 0.9452,
+      "step": 60985
+    },
+    {
+      "epoch": 18.25,
+      "grad_norm": 1.5139596462249756,
+      "learning_rate": 9.46449412359321e-07,
+      "loss": 0.951,
+      "step": 60990
+    },
+    {
+      "epoch": 18.25,
+      "grad_norm": 3.112560749053955,
+      "learning_rate": 9.448487962929492e-07,
+      "loss": 1.0462,
+      "step": 60995
+    },
+    {
+      "epoch": 18.25,
+      "grad_norm": 6.812036991119385,
+      "learning_rate": 9.432495087675657e-07,
+      "loss": 1.0087,
+      "step": 61000
+    },
+    {
+      "epoch": 18.25,
+      "grad_norm": 1.6199243068695068,
+      "learning_rate": 9.416515498715139e-07,
+      "loss": 0.8079,
+      "step": 61005
+    },
+    {
+      "epoch": 18.25,
+      "grad_norm": 2.11490797996521,
+      "learning_rate": 9.400549196930342e-07,
+      "loss": 0.9901,
+      "step": 61010
+    },
+    {
+      "epoch": 18.25,
+      "grad_norm": 1.8027985095977783,
+      "learning_rate": 9.38459618320317e-07,
+      "loss": 1.1343,
+      "step": 61015
+    },
+    {
+      "epoch": 18.26,
+      "grad_norm": 2.305330514907837,
+      "learning_rate": 9.368656458414643e-07,
+      "loss": 1.1905,
+      "step": 61020
+    },
+    {
+      "epoch": 18.26,
+      "grad_norm": 1.3872077465057373,
+      "learning_rate": 9.352730023445055e-07,
+      "loss": 0.9287,
+      "step": 61025
+    },
+    {
+      "epoch": 18.26,
+      "grad_norm": 3.1503524780273438,
+      "learning_rate": 9.336816879174093e-07,
+      "loss": 0.9731,
+      "step": 61030
+    },
+    {
+      "epoch": 18.26,
+      "grad_norm": 1.6931301355361938,
+      "learning_rate": 9.320917026480553e-07,
+      "loss": 0.9644,
+      "step": 61035
+    },
+    {
+      "epoch": 18.26,
+      "grad_norm": 1.539621353149414,
+      "learning_rate": 9.305030466242592e-07,
+      "loss": 1.0219,
+      "step": 61040
+    },
+    {
+      "epoch": 18.26,
+      "grad_norm": 2.573394298553467,
+      "learning_rate": 9.289157199337622e-07,
+      "loss": 1.0027,
+      "step": 61045
+    },
+    {
+      "epoch": 18.27,
+      "grad_norm": 2.412062168121338,
+      "learning_rate": 9.273297226642275e-07,
+      "loss": 1.0636,
+      "step": 61050
+    },
+    {
+      "epoch": 18.27,
+      "grad_norm": 1.9419618844985962,
+      "learning_rate": 9.257450549032515e-07,
+      "loss": 0.9127,
+      "step": 61055
+    },
+    {
+      "epoch": 18.27,
+      "grad_norm": 1.3534716367721558,
+      "learning_rate": 9.241617167383531e-07,
+      "loss": 1.0985,
+      "step": 61060
+    },
+    {
+      "epoch": 18.27,
+      "grad_norm": 3.1224892139434814,
+      "learning_rate": 9.225797082569765e-07,
+      "loss": 1.0405,
+      "step": 61065
+    },
+    {
+      "epoch": 18.27,
+      "grad_norm": 2.6927273273468018,
+      "learning_rate": 9.209990295464959e-07,
+      "loss": 0.8734,
+      "step": 61070
+    },
+    {
+      "epoch": 18.27,
+      "grad_norm": 1.9963269233703613,
+      "learning_rate": 9.194196806942112e-07,
+      "loss": 0.8057,
+      "step": 61075
+    },
+    {
+      "epoch": 18.27,
+      "grad_norm": 3.602238178253174,
+      "learning_rate": 9.178416617873442e-07,
+      "loss": 0.9009,
+      "step": 61080
+    },
+    {
+      "epoch": 18.28,
+      "grad_norm": 1.87447988986969,
+      "learning_rate": 9.162649729130529e-07,
+      "loss": 0.8498,
+      "step": 61085
+    },
+    {
+      "epoch": 18.28,
+      "grad_norm": 2.5276732444763184,
+      "learning_rate": 9.146896141584149e-07,
+      "loss": 0.9557,
+      "step": 61090
+    },
+    {
+      "epoch": 18.28,
+      "grad_norm": 2.9176580905914307,
+      "learning_rate": 9.131155856104301e-07,
+      "loss": 0.9328,
+      "step": 61095
+    },
+    {
+      "epoch": 18.28,
+      "grad_norm": 1.1620190143585205,
+      "learning_rate": 9.115428873560372e-07,
+      "loss": 1.0674,
+      "step": 61100
+    },
+    {
+      "epoch": 18.28,
+      "grad_norm": 3.642578601837158,
+      "learning_rate": 9.09971519482089e-07,
+      "loss": 0.8607,
+      "step": 61105
+    },
+    {
+      "epoch": 18.28,
+      "grad_norm": 3.2962098121643066,
+      "learning_rate": 9.0840148207538e-07,
+      "loss": 0.8209,
+      "step": 61110
+    },
+    {
+      "epoch": 18.28,
+      "grad_norm": 2.739952325820923,
+      "learning_rate": 9.068327752226102e-07,
+      "loss": 0.9166,
+      "step": 61115
+    },
+    {
+      "epoch": 18.29,
+      "grad_norm": 2.815523862838745,
+      "learning_rate": 9.052653990104243e-07,
+      "loss": 1.0532,
+      "step": 61120
+    },
+    {
+      "epoch": 18.29,
+      "grad_norm": 1.6121025085449219,
+      "learning_rate": 9.036993535253863e-07,
+      "loss": 1.0011,
+      "step": 61125
+    },
+    {
+      "epoch": 18.29,
+      "grad_norm": 1.8355814218521118,
+      "learning_rate": 9.021346388539825e-07,
+      "loss": 0.9553,
+      "step": 61130
+    },
+    {
+      "epoch": 18.29,
+      "grad_norm": 2.7154858112335205,
+      "learning_rate": 9.005712550826384e-07,
+      "loss": 1.089,
+      "step": 61135
+    },
+    {
+      "epoch": 18.29,
+      "grad_norm": 2.364108085632324,
+      "learning_rate": 8.990092022976904e-07,
+      "loss": 0.9518,
+      "step": 61140
+    },
+    {
+      "epoch": 18.29,
+      "grad_norm": 1.0809826850891113,
+      "learning_rate": 8.974484805854166e-07,
+      "loss": 0.9845,
+      "step": 61145
+    },
+    {
+      "epoch": 18.3,
+      "grad_norm": 1.4934520721435547,
+      "learning_rate": 8.958890900320066e-07,
+      "loss": 0.9068,
+      "step": 61150
+    },
+    {
+      "epoch": 18.3,
+      "grad_norm": 2.0490729808807373,
+      "learning_rate": 8.943310307235886e-07,
+      "loss": 1.0463,
+      "step": 61155
+    },
+    {
+      "epoch": 18.3,
+      "grad_norm": 2.349944591522217,
+      "learning_rate": 8.927743027462104e-07,
+      "loss": 1.065,
+      "step": 61160
+    },
+    {
+      "epoch": 18.3,
+      "grad_norm": 2.097916841506958,
+      "learning_rate": 8.912189061858506e-07,
+      "loss": 0.9555,
+      "step": 61165
+    },
+    {
+      "epoch": 18.3,
+      "grad_norm": 1.0531599521636963,
+      "learning_rate": 8.896648411284097e-07,
+      "loss": 0.9135,
+      "step": 61170
+    },
+    {
+      "epoch": 18.3,
+      "grad_norm": 3.910612106323242,
+      "learning_rate": 8.881121076597193e-07,
+      "loss": 0.7968,
+      "step": 61175
+    },
+    {
+      "epoch": 18.3,
+      "grad_norm": 2.345278739929199,
+      "learning_rate": 8.865607058655356e-07,
+      "loss": 0.9632,
+      "step": 61180
+    },
+    {
+      "epoch": 18.31,
+      "grad_norm": 2.56709885597229,
+      "learning_rate": 8.850106358315402e-07,
+      "loss": 1.1304,
+      "step": 61185
+    },
+    {
+      "epoch": 18.31,
+      "grad_norm": 3.583867073059082,
+      "learning_rate": 8.834618976433367e-07,
+      "loss": 0.7337,
+      "step": 61190
+    },
+    {
+      "epoch": 18.31,
+      "grad_norm": 2.285977840423584,
+      "learning_rate": 8.819144913864708e-07,
+      "loss": 1.0113,
+      "step": 61195
+    },
+    {
+      "epoch": 18.31,
+      "grad_norm": 3.424396514892578,
+      "learning_rate": 8.803684171463905e-07,
+      "loss": 0.91,
+      "step": 61200
+    },
+    {
+      "epoch": 18.31,
+      "grad_norm": 3.49349045753479,
+      "learning_rate": 8.788236750084999e-07,
+      "loss": 1.0395,
+      "step": 61205
+    },
+    {
+      "epoch": 18.31,
+      "grad_norm": 1.7467879056930542,
+      "learning_rate": 8.772802650580975e-07,
+      "loss": 0.8866,
+      "step": 61210
+    },
+    {
+      "epoch": 18.31,
+      "grad_norm": 5.124495506286621,
+      "learning_rate": 8.757381873804371e-07,
+      "loss": 0.9363,
+      "step": 61215
+    },
+    {
+      "epoch": 18.32,
+      "grad_norm": 1.684017539024353,
+      "learning_rate": 8.741974420606813e-07,
+      "loss": 0.8885,
+      "step": 61220
+    },
+    {
+      "epoch": 18.32,
+      "grad_norm": 3.0965070724487305,
+      "learning_rate": 8.726580291839203e-07,
+      "loss": 0.9159,
+      "step": 61225
+    },
+    {
+      "epoch": 18.32,
+      "grad_norm": 4.807112693786621,
+      "learning_rate": 8.711199488351779e-07,
+      "loss": 0.9592,
+      "step": 61230
+    },
+    {
+      "epoch": 18.32,
+      "grad_norm": 2.7470273971557617,
+      "learning_rate": 8.695832010993998e-07,
+      "loss": 0.8318,
+      "step": 61235
+    },
+    {
+      "epoch": 18.32,
+      "grad_norm": 5.211430072784424,
+      "learning_rate": 8.6804778606146e-07,
+      "loss": 0.9829,
+      "step": 61240
+    },
+    {
+      "epoch": 18.32,
+      "grad_norm": 2.6941945552825928,
+      "learning_rate": 8.665137038061572e-07,
+      "loss": 0.9167,
+      "step": 61245
+    },
+    {
+      "epoch": 18.33,
+      "grad_norm": 2.2288553714752197,
+      "learning_rate": 8.64980954418218e-07,
+      "loss": 0.9119,
+      "step": 61250
+    },
+    {
+      "epoch": 18.33,
+      "grad_norm": 2.5863308906555176,
+      "learning_rate": 8.634495379822943e-07,
+      "loss": 1.03,
+      "step": 61255
+    },
+    {
+      "epoch": 18.33,
+      "grad_norm": 1.7381858825683594,
+      "learning_rate": 8.619194545829628e-07,
+      "loss": 1.1305,
+      "step": 61260
+    },
+    {
+      "epoch": 18.33,
+      "grad_norm": 2.788691997528076,
+      "learning_rate": 8.603907043047283e-07,
+      "loss": 1.1239,
+      "step": 61265
+    },
+    {
+      "epoch": 18.33,
+      "grad_norm": 3.6779868602752686,
+      "learning_rate": 8.588632872320257e-07,
+      "loss": 1.1608,
+      "step": 61270
+    },
+    {
+      "epoch": 18.33,
+      "grad_norm": 3.579209327697754,
+      "learning_rate": 8.573372034492099e-07,
+      "loss": 0.9267,
+      "step": 61275
+    },
+    {
+      "epoch": 18.33,
+      "grad_norm": 4.529468059539795,
+      "learning_rate": 8.558124530405664e-07,
+      "loss": 0.9345,
+      "step": 61280
+    },
+    {
+      "epoch": 18.34,
+      "grad_norm": 1.8296302556991577,
+      "learning_rate": 8.542890360903e-07,
+      "loss": 0.796,
+      "step": 61285
+    },
+    {
+      "epoch": 18.34,
+      "grad_norm": 2.3215250968933105,
+      "learning_rate": 8.527669526825599e-07,
+      "loss": 1.0763,
+      "step": 61290
+    },
+    {
+      "epoch": 18.34,
+      "grad_norm": 3.3448548316955566,
+      "learning_rate": 8.512462029013929e-07,
+      "loss": 0.9275,
+      "step": 61295
+    },
+    {
+      "epoch": 18.34,
+      "grad_norm": 2.267402172088623,
+      "learning_rate": 8.49726786830804e-07,
+      "loss": 1.107,
+      "step": 61300
+    },
+    {
+      "epoch": 18.34,
+      "grad_norm": 3.513765335083008,
+      "learning_rate": 8.482087045546955e-07,
+      "loss": 1.0165,
+      "step": 61305
+    },
+    {
+      "epoch": 18.34,
+      "grad_norm": 2.1769518852233887,
+      "learning_rate": 8.466919561569225e-07,
+      "loss": 0.9577,
+      "step": 61310
+    },
+    {
+      "epoch": 18.34,
+      "grad_norm": 2.1250414848327637,
+      "learning_rate": 8.451765417212432e-07,
+      "loss": 0.8942,
+      "step": 61315
+    },
+    {
+      "epoch": 18.35,
+      "grad_norm": 1.5039304494857788,
+      "learning_rate": 8.43662461331357e-07,
+      "loss": 0.9996,
+      "step": 61320
+    },
+    {
+      "epoch": 18.35,
+      "grad_norm": 1.338394045829773,
+      "learning_rate": 8.421497150708835e-07,
+      "loss": 0.8906,
+      "step": 61325
+    },
+    {
+      "epoch": 18.35,
+      "grad_norm": 2.3154618740081787,
+      "learning_rate": 8.406383030233694e-07,
+      "loss": 1.0178,
+      "step": 61330
+    },
+    {
+      "epoch": 18.35,
+      "grad_norm": 1.515351414680481,
+      "learning_rate": 8.391282252722898e-07,
+      "loss": 0.9011,
+      "step": 61335
+    },
+    {
+      "epoch": 18.35,
+      "grad_norm": 4.342297077178955,
+      "learning_rate": 8.376194819010446e-07,
+      "loss": 0.9005,
+      "step": 61340
+    },
+    {
+      "epoch": 18.35,
+      "grad_norm": 3.609494209289551,
+      "learning_rate": 8.361120729929617e-07,
+      "loss": 0.9243,
+      "step": 61345
+    },
+    {
+      "epoch": 18.36,
+      "grad_norm": 2.8200643062591553,
+      "learning_rate": 8.34605998631291e-07,
+      "loss": 1.0519,
+      "step": 61350
+    },
+    {
+      "epoch": 18.36,
+      "grad_norm": 2.881962537765503,
+      "learning_rate": 8.331012588992132e-07,
+      "loss": 0.8966,
+      "step": 61355
+    },
+    {
+      "epoch": 18.36,
+      "grad_norm": 2.507383108139038,
+      "learning_rate": 8.315978538798314e-07,
+      "loss": 0.8928,
+      "step": 61360
+    },
+    {
+      "epoch": 18.36,
+      "grad_norm": 1.8377445936203003,
+      "learning_rate": 8.30095783656179e-07,
+      "loss": 0.9829,
+      "step": 61365
+    },
+    {
+      "epoch": 18.36,
+      "grad_norm": 1.3998005390167236,
+      "learning_rate": 8.28595048311212e-07,
+      "loss": 0.9379,
+      "step": 61370
+    },
+    {
+      "epoch": 18.36,
+      "grad_norm": 1.7559727430343628,
+      "learning_rate": 8.270956479278196e-07,
+      "loss": 1.1797,
+      "step": 61375
+    },
+    {
+      "epoch": 18.36,
+      "grad_norm": 1.639439582824707,
+      "learning_rate": 8.255975825888024e-07,
+      "loss": 0.8938,
+      "step": 61380
+    },
+    {
+      "epoch": 18.37,
+      "grad_norm": 2.903548002243042,
+      "learning_rate": 8.241008523769106e-07,
+      "loss": 0.8785,
+      "step": 61385
+    },
+    {
+      "epoch": 18.37,
+      "grad_norm": 2.10689377784729,
+      "learning_rate": 8.226054573747894e-07,
+      "loss": 1.0115,
+      "step": 61390
+    },
+    {
+      "epoch": 18.37,
+      "grad_norm": 3.6648929119110107,
+      "learning_rate": 8.211113976650475e-07,
+      "loss": 0.9137,
+      "step": 61395
+    },
+    {
+      "epoch": 18.37,
+      "grad_norm": 3.99898624420166,
+      "learning_rate": 8.19618673330183e-07,
+      "loss": 1.1167,
+      "step": 61400
+    },
+    {
+      "epoch": 18.37,
+      "grad_norm": 3.173650026321411,
+      "learning_rate": 8.181272844526495e-07,
+      "loss": 1.1092,
+      "step": 61405
+    },
+    {
+      "epoch": 18.37,
+      "grad_norm": 3.0180552005767822,
+      "learning_rate": 8.166372311148113e-07,
+      "loss": 0.8644,
+      "step": 61410
+    },
+    {
+      "epoch": 18.37,
+      "grad_norm": 3.1037330627441406,
+      "learning_rate": 8.151485133989584e-07,
+      "loss": 0.8741,
+      "step": 61415
+    },
+    {
+      "epoch": 18.38,
+      "grad_norm": 3.032813310623169,
+      "learning_rate": 8.136611313873166e-07,
+      "loss": 0.9209,
+      "step": 61420
+    },
+    {
+      "epoch": 18.38,
+      "grad_norm": 2.8166263103485107,
+      "learning_rate": 8.121750851620286e-07,
+      "loss": 0.8434,
+      "step": 61425
+    },
+    {
+      "epoch": 18.38,
+      "grad_norm": 2.1553432941436768,
+      "learning_rate": 8.106903748051675e-07,
+      "loss": 0.8497,
+      "step": 61430
+    },
+    {
+      "epoch": 18.38,
+      "grad_norm": 2.0251331329345703,
+      "learning_rate": 8.092070003987373e-07,
+      "loss": 1.1232,
+      "step": 61435
+    },
+    {
+      "epoch": 18.38,
+      "grad_norm": 4.4784345626831055,
+      "learning_rate": 8.077249620246558e-07,
+      "loss": 0.9472,
+      "step": 61440
+    },
+    {
+      "epoch": 18.38,
+      "grad_norm": 3.576678514480591,
+      "learning_rate": 8.062442597647795e-07,
+      "loss": 0.8144,
+      "step": 61445
+    },
+    {
+      "epoch": 18.39,
+      "grad_norm": 1.705344796180725,
+      "learning_rate": 8.047648937008851e-07,
+      "loss": 1.1875,
+      "step": 61450
+    },
+    {
+      "epoch": 18.39,
+      "grad_norm": 1.4665272235870361,
+      "learning_rate": 8.032868639146762e-07,
+      "loss": 1.0611,
+      "step": 61455
+    },
+    {
+      "epoch": 18.39,
+      "grad_norm": 2.712339162826538,
+      "learning_rate": 8.018101704877823e-07,
+      "loss": 1.0206,
+      "step": 61460
+    },
+    {
+      "epoch": 18.39,
+      "grad_norm": 3.855246067047119,
+      "learning_rate": 8.003348135017603e-07,
+      "loss": 1.0525,
+      "step": 61465
+    },
+    {
+      "epoch": 18.39,
+      "grad_norm": 2.246304988861084,
+      "learning_rate": 7.988607930380948e-07,
+      "loss": 0.8379,
+      "step": 61470
+    },
+    {
+      "epoch": 18.39,
+      "grad_norm": 3.892084836959839,
+      "learning_rate": 7.973881091781876e-07,
+      "loss": 1.0106,
+      "step": 61475
+    },
+    {
+      "epoch": 18.39,
+      "grad_norm": 1.8931845426559448,
+      "learning_rate": 7.959167620033847e-07,
+      "loss": 0.9843,
+      "step": 61480
+    },
+    {
+      "epoch": 18.4,
+      "grad_norm": 2.768260955810547,
+      "learning_rate": 7.944467515949322e-07,
+      "loss": 0.8572,
+      "step": 61485
+    },
+    {
+      "epoch": 18.4,
+      "grad_norm": 11.066926002502441,
+      "learning_rate": 7.929780780340318e-07,
+      "loss": 0.8194,
+      "step": 61490
+    },
+    {
+      "epoch": 18.4,
+      "grad_norm": 4.7817912101745605,
+      "learning_rate": 7.915107414017825e-07,
+      "loss": 1.0095,
+      "step": 61495
+    },
+    {
+      "epoch": 18.4,
+      "grad_norm": 1.809881329536438,
+      "learning_rate": 7.900447417792389e-07,
+      "loss": 0.8781,
+      "step": 61500
+    },
+    {
+      "epoch": 18.4,
+      "grad_norm": 3.047032356262207,
+      "learning_rate": 7.885800792473585e-07,
+      "loss": 1.0429,
+      "step": 61505
+    },
+    {
+      "epoch": 18.4,
+      "grad_norm": 1.490786075592041,
+      "learning_rate": 7.871167538870322e-07,
+      "loss": 0.8668,
+      "step": 61510
+    },
+    {
+      "epoch": 18.4,
+      "grad_norm": 1.9980204105377197,
+      "learning_rate": 7.856547657790786e-07,
+      "loss": 1.0007,
+      "step": 61515
+    },
+    {
+      "epoch": 18.41,
+      "grad_norm": 3.4392709732055664,
+      "learning_rate": 7.841941150042415e-07,
+      "loss": 0.9901,
+      "step": 61520
+    },
+    {
+      "epoch": 18.41,
+      "grad_norm": 4.489561557769775,
+      "learning_rate": 7.827348016431979e-07,
+      "loss": 1.0117,
+      "step": 61525
+    },
+    {
+      "epoch": 18.41,
+      "grad_norm": 4.4032979011535645,
+      "learning_rate": 7.812768257765335e-07,
+      "loss": 1.0882,
+      "step": 61530
+    },
+    {
+      "epoch": 18.41,
+      "grad_norm": 3.137603282928467,
+      "learning_rate": 7.79820187484781e-07,
+      "loss": 0.921,
+      "step": 61535
+    },
+    {
+      "epoch": 18.41,
+      "grad_norm": 4.174556732177734,
+      "learning_rate": 7.783648868483817e-07,
+      "loss": 0.982,
+      "step": 61540
+    },
+    {
+      "epoch": 18.41,
+      "grad_norm": 2.2925820350646973,
+      "learning_rate": 7.769109239477129e-07,
+      "loss": 0.9188,
+      "step": 61545
+    },
+    {
+      "epoch": 18.42,
+      "grad_norm": 2.8582160472869873,
+      "learning_rate": 7.754582988630743e-07,
+      "loss": 1.0092,
+      "step": 61550
+    },
+    {
+      "epoch": 18.42,
+      "grad_norm": 3.7304418087005615,
+      "learning_rate": 7.740070116746961e-07,
+      "loss": 0.7871,
+      "step": 61555
+    },
+    {
+      "epoch": 18.42,
+      "grad_norm": 2.1194798946380615,
+      "learning_rate": 7.725570624627282e-07,
+      "loss": 0.9669,
+      "step": 61560
+    },
+    {
+      "epoch": 18.42,
+      "grad_norm": 2.3362677097320557,
+      "learning_rate": 7.71108451307248e-07,
+      "loss": 0.9318,
+      "step": 61565
+    },
+    {
+      "epoch": 18.42,
+      "grad_norm": 1.7674001455307007,
+      "learning_rate": 7.696611782882668e-07,
+      "loss": 0.9901,
+      "step": 61570
+    },
+    {
+      "epoch": 18.42,
+      "grad_norm": 1.8596383333206177,
+      "learning_rate": 7.682152434857149e-07,
+      "loss": 0.8932,
+      "step": 61575
+    },
+    {
+      "epoch": 18.42,
+      "grad_norm": 3.049741744995117,
+      "learning_rate": 7.667706469794395e-07,
+      "loss": 0.8584,
+      "step": 61580
+    },
+    {
+      "epoch": 18.43,
+      "grad_norm": 2.637298107147217,
+      "learning_rate": 7.653273888492407e-07,
+      "loss": 1.1926,
+      "step": 61585
+    },
+    {
+      "epoch": 18.43,
+      "grad_norm": 2.7286012172698975,
+      "learning_rate": 7.638854691748132e-07,
+      "loss": 1.0714,
+      "step": 61590
+    },
+    {
+      "epoch": 18.43,
+      "grad_norm": 4.056117057800293,
+      "learning_rate": 7.624448880358043e-07,
+      "loss": 0.9455,
+      "step": 61595
+    },
+    {
+      "epoch": 18.43,
+      "grad_norm": 2.3087191581726074,
+      "learning_rate": 7.61005645511767e-07,
+      "loss": 1.0218,
+      "step": 61600
+    },
+    {
+      "epoch": 18.43,
+      "grad_norm": 2.0572636127471924,
+      "learning_rate": 7.595677416821933e-07,
+      "loss": 0.9051,
+      "step": 61605
+    },
+    {
+      "epoch": 18.43,
+      "grad_norm": 6.119582176208496,
+      "learning_rate": 7.581311766265003e-07,
+      "loss": 0.8552,
+      "step": 61610
+    },
+    {
+      "epoch": 18.43,
+      "grad_norm": 2.6109657287597656,
+      "learning_rate": 7.56695950424019e-07,
+      "loss": 0.9766,
+      "step": 61615
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 3.126173734664917,
+      "learning_rate": 7.552620631540247e-07,
+      "loss": 0.9172,
+      "step": 61620
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 3.1183481216430664,
+      "learning_rate": 7.538295148957015e-07,
+      "loss": 0.7652,
+      "step": 61625
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 1.6673765182495117,
+      "learning_rate": 7.523983057281775e-07,
+      "loss": 0.884,
+      "step": 61630
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 3.5469305515289307,
+      "learning_rate": 7.509684357304897e-07,
+      "loss": 0.9741,
+      "step": 61635
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 2.485867977142334,
+      "learning_rate": 7.495399049816082e-07,
+      "loss": 0.7726,
+      "step": 61640
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 1.444970965385437,
+      "learning_rate": 7.48112713560431e-07,
+      "loss": 0.9309,
+      "step": 61645
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 3.333101987838745,
+      "learning_rate": 7.466868615457783e-07,
+      "loss": 1.1428,
+      "step": 61650
+    },
+    {
+      "epoch": 18.45,
+      "grad_norm": 2.304952383041382,
+      "learning_rate": 7.45262349016404e-07,
+      "loss": 0.9161,
+      "step": 61655
+    },
+    {
+      "epoch": 18.45,
+      "grad_norm": 2.9560463428497314,
+      "learning_rate": 7.438391760509755e-07,
+      "loss": 0.9267,
+      "step": 61660
+    },
+    {
+      "epoch": 18.45,
+      "grad_norm": 4.144384384155273,
+      "learning_rate": 7.427016022174993e-07,
+      "loss": 1.1577,
+      "step": 61665
+    },
+    {
+      "epoch": 18.45,
+      "grad_norm": 5.686766624450684,
+      "learning_rate": 7.412808406652038e-07,
+      "loss": 0.9974,
+      "step": 61670
+    },
+    {
+      "epoch": 18.45,
+      "grad_norm": 3.3661046028137207,
+      "learning_rate": 7.398614188967534e-07,
+      "loss": 0.8438,
+      "step": 61675
+    },
+    {
+      "epoch": 18.45,
+      "grad_norm": 5.448729515075684,
+      "learning_rate": 7.38443336990538e-07,
+      "loss": 0.8848,
+      "step": 61680
+    },
+    {
+      "epoch": 18.46,
+      "grad_norm": 2.479353189468384,
+      "learning_rate": 7.370265950248783e-07,
+      "loss": 1.0998,
+      "step": 61685
+    },
+    {
+      "epoch": 18.46,
+      "grad_norm": 2.4149727821350098,
+      "learning_rate": 7.356111930780201e-07,
+      "loss": 0.9238,
+      "step": 61690
+    },
+    {
+      "epoch": 18.46,
+      "grad_norm": 2.909254312515259,
+      "learning_rate": 7.341971312281343e-07,
+      "loss": 0.9805,
+      "step": 61695
+    },
+    {
+      "epoch": 18.46,
+      "grad_norm": 2.9561984539031982,
+      "learning_rate": 7.327844095533193e-07,
+      "loss": 1.1177,
+      "step": 61700
+    },
+    {
+      "epoch": 18.46,
+      "grad_norm": 1.3195778131484985,
+      "learning_rate": 7.313730281315961e-07,
+      "loss": 1.0986,
+      "step": 61705
+    },
+    {
+      "epoch": 18.46,
+      "grad_norm": 1.9130079746246338,
+      "learning_rate": 7.299629870409136e-07,
+      "loss": 0.8861,
+      "step": 61710
+    },
+    {
+      "epoch": 18.46,
+      "grad_norm": 4.564031600952148,
+      "learning_rate": 7.285542863591482e-07,
+      "loss": 0.8437,
+      "step": 61715
+    },
+    {
+      "epoch": 18.47,
+      "grad_norm": 2.09362530708313,
+      "learning_rate": 7.271469261641017e-07,
+      "loss": 0.8984,
+      "step": 61720
+    },
+    {
+      "epoch": 18.47,
+      "grad_norm": 3.5240318775177,
+      "learning_rate": 7.257409065335035e-07,
+      "loss": 0.7457,
+      "step": 61725
+    },
+    {
+      "epoch": 18.47,
+      "grad_norm": 4.574277400970459,
+      "learning_rate": 7.243362275449972e-07,
+      "loss": 0.9621,
+      "step": 61730
+    },
+    {
+      "epoch": 18.47,
+      "grad_norm": 4.2122578620910645,
+      "learning_rate": 7.229328892761733e-07,
+      "loss": 0.9186,
+      "step": 61735
+    },
+    {
+      "epoch": 18.47,
+      "grad_norm": 2.056427001953125,
+      "learning_rate": 7.215308918045255e-07,
+      "loss": 1.0321,
+      "step": 61740
+    },
+    {
+      "epoch": 18.47,
+      "grad_norm": 3.035954236984253,
+      "learning_rate": 7.201302352074973e-07,
+      "loss": 0.9903,
+      "step": 61745
+    },
+    {
+      "epoch": 18.47,
+      "grad_norm": 3.6890830993652344,
+      "learning_rate": 7.187309195624353e-07,
+      "loss": 0.9409,
+      "step": 61750
+    },
+    {
+      "epoch": 18.48,
+      "grad_norm": 2.126034736633301,
+      "learning_rate": 7.173329449466248e-07,
+      "loss": 1.0002,
+      "step": 61755
+    },
+    {
+      "epoch": 18.48,
+      "grad_norm": 12.646512031555176,
+      "learning_rate": 7.159363114372763e-07,
+      "loss": 0.8951,
+      "step": 61760
+    },
+    {
+      "epoch": 18.48,
+      "grad_norm": 4.24096155166626,
+      "learning_rate": 7.145410191115226e-07,
+      "loss": 0.8652,
+      "step": 61765
+    },
+    {
+      "epoch": 18.48,
+      "grad_norm": 1.79692804813385,
+      "learning_rate": 7.13147068046427e-07,
+      "loss": 1.0179,
+      "step": 61770
+    },
+    {
+      "epoch": 18.48,
+      "grad_norm": 2.1023857593536377,
+      "learning_rate": 7.117544583189723e-07,
+      "loss": 1.0399,
+      "step": 61775
+    },
+    {
+      "epoch": 18.48,
+      "grad_norm": 3.325761079788208,
+      "learning_rate": 7.10363190006072e-07,
+      "loss": 0.9114,
+      "step": 61780
+    },
+    {
+      "epoch": 18.49,
+      "grad_norm": 1.3993957042694092,
+      "learning_rate": 7.089732631845674e-07,
+      "loss": 0.9271,
+      "step": 61785
+    },
+    {
+      "epoch": 18.49,
+      "grad_norm": 2.963554620742798,
+      "learning_rate": 7.075846779312196e-07,
+      "loss": 1.0317,
+      "step": 61790
+    },
+    {
+      "epoch": 18.49,
+      "grad_norm": 4.754477500915527,
+      "learning_rate": 7.061974343227168e-07,
+      "loss": 0.87,
+      "step": 61795
+    },
+    {
+      "epoch": 18.49,
+      "grad_norm": 2.58296799659729,
+      "learning_rate": 7.048115324356814e-07,
+      "loss": 1.0729,
+      "step": 61800
+    },
+    {
+      "epoch": 18.49,
+      "grad_norm": 1.766918659210205,
+      "learning_rate": 7.034269723466491e-07,
+      "loss": 1.0143,
+      "step": 61805
+    },
+    {
+      "epoch": 18.49,
+      "grad_norm": 6.538331985473633,
+      "learning_rate": 7.020437541320924e-07,
+      "loss": 0.9332,
+      "step": 61810
+    },
+    {
+      "epoch": 18.49,
+      "grad_norm": 1.1120764017105103,
+      "learning_rate": 7.006618778683999e-07,
+      "loss": 1.0328,
+      "step": 61815
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 3.360107898712158,
+      "learning_rate": 6.992813436318996e-07,
+      "loss": 0.9667,
+      "step": 61820
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 1.8557356595993042,
+      "learning_rate": 6.979021514988249e-07,
+      "loss": 1.0017,
+      "step": 61825
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 2.17095947265625,
+      "learning_rate": 6.965243015453593e-07,
+      "loss": 0.9728,
+      "step": 61830
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 2.826488971710205,
+      "learning_rate": 6.951477938475892e-07,
+      "loss": 0.9713,
+      "step": 61835
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 2.857255220413208,
+      "learning_rate": 6.937726284815482e-07,
+      "loss": 0.9573,
+      "step": 61840
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 6.233930587768555,
+      "learning_rate": 6.923988055231784e-07,
+      "loss": 1.2165,
+      "step": 61845
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 4.246452808380127,
+      "learning_rate": 6.910263250483551e-07,
+      "loss": 1.1228,
+      "step": 61850
+    },
+    {
+      "epoch": 18.51,
+      "grad_norm": 2.3864681720733643,
+      "learning_rate": 6.896551871328788e-07,
+      "loss": 0.9279,
+      "step": 61855
+    },
+    {
+      "epoch": 18.51,
+      "grad_norm": 3.7490153312683105,
+      "learning_rate": 6.882853918524779e-07,
+      "loss": 0.9047,
+      "step": 61860
+    },
+    {
+      "epoch": 18.51,
+      "grad_norm": 3.5015833377838135,
+      "learning_rate": 6.869169392828056e-07,
+      "loss": 1.0886,
+      "step": 61865
+    },
+    {
+      "epoch": 18.51,
+      "grad_norm": 4.527100563049316,
+      "learning_rate": 6.85549829499435e-07,
+      "loss": 0.8575,
+      "step": 61870
+    },
+    {
+      "epoch": 18.51,
+      "grad_norm": 4.221701145172119,
+      "learning_rate": 6.841840625778805e-07,
+      "loss": 0.9333,
+      "step": 61875
+    },
+    {
+      "epoch": 18.51,
+      "grad_norm": 2.010347843170166,
+      "learning_rate": 6.828196385935626e-07,
+      "loss": 0.8162,
+      "step": 61880
+    },
+    {
+      "epoch": 18.52,
+      "grad_norm": 2.520059585571289,
+      "learning_rate": 6.814565576218374e-07,
+      "loss": 1.0076,
+      "step": 61885
+    },
+    {
+      "epoch": 18.52,
+      "grad_norm": 2.8664329051971436,
+      "learning_rate": 6.80094819737992e-07,
+      "loss": 1.0359,
+      "step": 61890
+    },
+    {
+      "epoch": 18.52,
+      "grad_norm": 3.0318400859832764,
+      "learning_rate": 6.787344250172273e-07,
+      "loss": 0.9725,
+      "step": 61895
+    },
+    {
+      "epoch": 18.52,
+      "grad_norm": 4.079439640045166,
+      "learning_rate": 6.773753735346805e-07,
+      "loss": 0.9406,
+      "step": 61900
+    },
+    {
+      "epoch": 18.52,
+      "grad_norm": 2.608128309249878,
+      "learning_rate": 6.76017665365411e-07,
+      "loss": 0.8831,
+      "step": 61905
+    },
+    {
+      "epoch": 18.52,
+      "grad_norm": 2.588148355484009,
+      "learning_rate": 6.746613005844033e-07,
+      "loss": 1.0509,
+      "step": 61910
+    },
+    {
+      "epoch": 18.52,
+      "grad_norm": 1.884019374847412,
+      "learning_rate": 6.733062792665695e-07,
+      "loss": 1.0,
+      "step": 61915
+    },
+    {
+      "epoch": 18.53,
+      "grad_norm": 3.63097882270813,
+      "learning_rate": 6.719526014867361e-07,
+      "loss": 1.114,
+      "step": 61920
+    },
+    {
+      "epoch": 18.53,
+      "grad_norm": 1.6369963884353638,
+      "learning_rate": 6.706002673196793e-07,
+      "loss": 1.0093,
+      "step": 61925
+    },
+    {
+      "epoch": 18.53,
+      "grad_norm": 3.430427312850952,
+      "learning_rate": 6.692492768400782e-07,
+      "loss": 1.1051,
+      "step": 61930
+    },
+    {
+      "epoch": 18.53,
+      "grad_norm": 1.9971526861190796,
+      "learning_rate": 6.67899630122551e-07,
+      "loss": 0.9439,
+      "step": 61935
+    },
+    {
+      "epoch": 18.53,
+      "grad_norm": 1.4598491191864014,
+      "learning_rate": 6.665513272416324e-07,
+      "loss": 0.9472,
+      "step": 61940
+    },
+    {
+      "epoch": 18.53,
+      "grad_norm": 2.7384274005889893,
+      "learning_rate": 6.652043682717907e-07,
+      "loss": 0.922,
+      "step": 61945
+    },
+    {
+      "epoch": 18.53,
+      "grad_norm": 2.7123560905456543,
+      "learning_rate": 6.638587532874219e-07,
+      "loss": 0.9753,
+      "step": 61950
+    },
+    {
+      "epoch": 18.54,
+      "grad_norm": 2.5400397777557373,
+      "learning_rate": 6.625144823628332e-07,
+      "loss": 0.7639,
+      "step": 61955
+    },
+    {
+      "epoch": 18.54,
+      "grad_norm": 4.62942361831665,
+      "learning_rate": 6.611715555722764e-07,
+      "loss": 1.0445,
+      "step": 61960
+    },
+    {
+      "epoch": 18.54,
+      "grad_norm": 2.880040168762207,
+      "learning_rate": 6.598299729899088e-07,
+      "loss": 0.9792,
+      "step": 61965
+    },
+    {
+      "epoch": 18.54,
+      "grad_norm": 2.0064916610717773,
+      "learning_rate": 6.584897346898405e-07,
+      "loss": 1.0655,
+      "step": 61970
+    },
+    {
+      "epoch": 18.54,
+      "grad_norm": 2.9998416900634766,
+      "learning_rate": 6.571508407460764e-07,
+      "loss": 0.9663,
+      "step": 61975
+    },
+    {
+      "epoch": 18.54,
+      "grad_norm": 2.280630350112915,
+      "learning_rate": 6.55813291232571e-07,
+      "loss": 1.0144,
+      "step": 61980
+    },
+    {
+      "epoch": 18.55,
+      "grad_norm": 2.611635208129883,
+      "learning_rate": 6.54477086223193e-07,
+      "loss": 1.2247,
+      "step": 61985
+    },
+    {
+      "epoch": 18.55,
+      "grad_norm": 5.889001369476318,
+      "learning_rate": 6.531422257917391e-07,
+      "loss": 1.0725,
+      "step": 61990
+    },
+    {
+      "epoch": 18.55,
+      "grad_norm": 3.5222485065460205,
+      "learning_rate": 6.518087100119335e-07,
+      "loss": 0.8588,
+      "step": 61995
+    },
+    {
+      "epoch": 18.55,
+      "grad_norm": 2.528952121734619,
+      "learning_rate": 6.50476538957423e-07,
+      "loss": 0.9481,
+      "step": 62000
+    },
+    {
+      "epoch": 18.55,
+      "grad_norm": 2.8010430335998535,
+      "learning_rate": 6.491457127017847e-07,
+      "loss": 0.8445,
+      "step": 62005
+    },
+    {
+      "epoch": 18.55,
+      "grad_norm": 1.7682600021362305,
+      "learning_rate": 6.478162313185182e-07,
+      "loss": 0.9884,
+      "step": 62010
+    },
+    {
+      "epoch": 18.55,
+      "grad_norm": 4.158107280731201,
+      "learning_rate": 6.464880948810453e-07,
+      "loss": 0.8942,
+      "step": 62015
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 2.5172414779663086,
+      "learning_rate": 6.45161303462724e-07,
+      "loss": 0.9485,
+      "step": 62020
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 1.1113386154174805,
+      "learning_rate": 6.438358571368263e-07,
+      "loss": 0.9776,
+      "step": 62025
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 2.090869426727295,
+      "learning_rate": 6.425117559765576e-07,
+      "loss": 1.0483,
+      "step": 62030
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 3.047197103500366,
+      "learning_rate": 6.41189000055048e-07,
+      "loss": 0.8976,
+      "step": 62035
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 2.634687662124634,
+      "learning_rate": 6.398675894453477e-07,
+      "loss": 0.9284,
+      "step": 62040
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 3.5582494735717773,
+      "learning_rate": 6.385475242204453e-07,
+      "loss": 0.8967,
+      "step": 62045
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 5.443139553070068,
+      "learning_rate": 6.372288044532326e-07,
+      "loss": 0.9113,
+      "step": 62050
+    },
+    {
+      "epoch": 18.57,
+      "grad_norm": 1.4090752601623535,
+      "learning_rate": 6.35911430216557e-07,
+      "loss": 1.0798,
+      "step": 62055
+    },
+    {
+      "epoch": 18.57,
+      "grad_norm": 1.6593273878097534,
+      "learning_rate": 6.345954015831629e-07,
+      "loss": 0.6439,
+      "step": 62060
+    },
+    {
+      "epoch": 18.57,
+      "grad_norm": 1.541206955909729,
+      "learning_rate": 6.332807186257423e-07,
+      "loss": 0.9511,
+      "step": 62065
+    },
+    {
+      "epoch": 18.57,
+      "grad_norm": 4.009206295013428,
+      "learning_rate": 6.319673814168953e-07,
+      "loss": 0.8826,
+      "step": 62070
+    },
+    {
+      "epoch": 18.57,
+      "grad_norm": 2.005704879760742,
+      "learning_rate": 6.30655390029164e-07,
+      "loss": 0.9057,
+      "step": 62075
+    },
+    {
+      "epoch": 18.57,
+      "grad_norm": 2.7836337089538574,
+      "learning_rate": 6.293447445350043e-07,
+      "loss": 0.9222,
+      "step": 62080
+    },
+    {
+      "epoch": 18.58,
+      "grad_norm": 4.371108055114746,
+      "learning_rate": 6.280354450068026e-07,
+      "loss": 0.8937,
+      "step": 62085
+    },
+    {
+      "epoch": 18.58,
+      "grad_norm": 2.3152432441711426,
+      "learning_rate": 6.26727491516868e-07,
+      "loss": 1.1751,
+      "step": 62090
+    },
+    {
+      "epoch": 18.58,
+      "grad_norm": 2.359133720397949,
+      "learning_rate": 6.254208841374426e-07,
+      "loss": 0.9842,
+      "step": 62095
+    },
+    {
+      "epoch": 18.58,
+      "grad_norm": 2.8773648738861084,
+      "learning_rate": 6.241156229406825e-07,
+      "loss": 0.9049,
+      "step": 62100
+    },
+    {
+      "epoch": 18.58,
+      "grad_norm": 2.3904407024383545,
+      "learning_rate": 6.228117079986829e-07,
+      "loss": 0.9434,
+      "step": 62105
+    },
+    {
+      "epoch": 18.58,
+      "grad_norm": 2.5811879634857178,
+      "learning_rate": 6.215091393834499e-07,
+      "loss": 1.0067,
+      "step": 62110
+    },
+    {
+      "epoch": 18.58,
+      "grad_norm": 5.010559558868408,
+      "learning_rate": 6.202079171669289e-07,
+      "loss": 0.9329,
+      "step": 62115
+    },
+    {
+      "epoch": 18.59,
+      "grad_norm": 6.2016167640686035,
+      "learning_rate": 6.189080414209847e-07,
+      "loss": 0.8496,
+      "step": 62120
+    },
+    {
+      "epoch": 18.59,
+      "grad_norm": 1.5568153858184814,
+      "learning_rate": 6.176095122174041e-07,
+      "loss": 0.9277,
+      "step": 62125
+    },
+    {
+      "epoch": 18.59,
+      "grad_norm": 1.9093276262283325,
+      "learning_rate": 6.163123296279077e-07,
+      "loss": 0.9473,
+      "step": 62130
+    },
+    {
+      "epoch": 18.59,
+      "grad_norm": 7.955539703369141,
+      "learning_rate": 6.150164937241354e-07,
+      "loss": 0.876,
+      "step": 62135
+    },
+    {
+      "epoch": 18.59,
+      "grad_norm": 2.9364593029022217,
+      "learning_rate": 6.137220045776604e-07,
+      "loss": 0.8387,
+      "step": 62140
+    },
+    {
+      "epoch": 18.59,
+      "grad_norm": 3.2200205326080322,
+      "learning_rate": 6.124288622599617e-07,
+      "loss": 0.9952,
+      "step": 62145
+    },
+    {
+      "epoch": 18.59,
+      "grad_norm": 2.081777572631836,
+      "learning_rate": 6.111370668424765e-07,
+      "loss": 0.8266,
+      "step": 62150
+    },
+    {
+      "epoch": 18.6,
+      "grad_norm": 2.223055601119995,
+      "learning_rate": 6.098466183965313e-07,
+      "loss": 0.9303,
+      "step": 62155
+    },
+    {
+      "epoch": 18.6,
+      "grad_norm": 1.2533679008483887,
+      "learning_rate": 6.085575169934132e-07,
+      "loss": 0.949,
+      "step": 62160
+    },
+    {
+      "epoch": 18.6,
+      "grad_norm": 2.873736619949341,
+      "learning_rate": 6.072697627043045e-07,
+      "loss": 1.0112,
+      "step": 62165
+    },
+    {
+      "epoch": 18.6,
+      "grad_norm": 1.682570219039917,
+      "learning_rate": 6.059833556003341e-07,
+      "loss": 1.0301,
+      "step": 62170
+    },
+    {
+      "epoch": 18.6,
+      "grad_norm": 4.487571716308594,
+      "learning_rate": 6.046982957525482e-07,
+      "loss": 1.1361,
+      "step": 62175
+    },
+    {
+      "epoch": 18.6,
+      "grad_norm": 4.013219833374023,
+      "learning_rate": 6.03414583231915e-07,
+      "loss": 0.8556,
+      "step": 62180
+    },
+    {
+      "epoch": 18.61,
+      "grad_norm": 3.0844905376434326,
+      "learning_rate": 6.021322181093386e-07,
+      "loss": 0.9482,
+      "step": 62185
+    },
+    {
+      "epoch": 18.61,
+      "grad_norm": 3.159336805343628,
+      "learning_rate": 6.008512004556377e-07,
+      "loss": 0.9925,
+      "step": 62190
+    },
+    {
+      "epoch": 18.61,
+      "grad_norm": 2.8269448280334473,
+      "learning_rate": 5.995715303415639e-07,
+      "loss": 1.0819,
+      "step": 62195
+    },
+    {
+      "epoch": 18.61,
+      "grad_norm": 3.574028730392456,
+      "learning_rate": 5.982932078377912e-07,
+      "loss": 0.9489,
+      "step": 62200
+    },
+    {
+      "epoch": 18.61,
+      "grad_norm": 2.7778759002685547,
+      "learning_rate": 5.970162330149187e-07,
+      "loss": 0.9698,
+      "step": 62205
+    },
+    {
+      "epoch": 18.61,
+      "grad_norm": 2.205036163330078,
+      "learning_rate": 5.957406059434761e-07,
+      "loss": 0.9544,
+      "step": 62210
+    },
+    {
+      "epoch": 18.61,
+      "grad_norm": 2.8578882217407227,
+      "learning_rate": 5.944663266939127e-07,
+      "loss": 0.7805,
+      "step": 62215
+    },
+    {
+      "epoch": 18.62,
+      "grad_norm": 2.122692346572876,
+      "learning_rate": 5.931933953366081e-07,
+      "loss": 0.9865,
+      "step": 62220
+    },
+    {
+      "epoch": 18.62,
+      "grad_norm": 1.806129813194275,
+      "learning_rate": 5.919218119418591e-07,
+      "loss": 0.909,
+      "step": 62225
+    },
+    {
+      "epoch": 18.62,
+      "grad_norm": 3.1925947666168213,
+      "learning_rate": 5.906515765799009e-07,
+      "loss": 0.97,
+      "step": 62230
+    },
+    {
+      "epoch": 18.62,
+      "grad_norm": 2.054266929626465,
+      "learning_rate": 5.89382689320886e-07,
+      "loss": 0.9196,
+      "step": 62235
+    },
+    {
+      "epoch": 18.62,
+      "grad_norm": 3.7723171710968018,
+      "learning_rate": 5.881151502348859e-07,
+      "loss": 1.0202,
+      "step": 62240
+    },
+    {
+      "epoch": 18.62,
+      "grad_norm": 1.945148229598999,
+      "learning_rate": 5.868489593919141e-07,
+      "loss": 1.1337,
+      "step": 62245
+    },
+    {
+      "epoch": 18.62,
+      "grad_norm": 2.7918272018432617,
+      "learning_rate": 5.855841168618953e-07,
+      "loss": 0.9992,
+      "step": 62250
+    },
+    {
+      "epoch": 18.63,
+      "grad_norm": 4.473152160644531,
+      "learning_rate": 5.843206227146958e-07,
+      "loss": 0.9585,
+      "step": 62255
+    },
+    {
+      "epoch": 18.63,
+      "grad_norm": 3.2821218967437744,
+      "learning_rate": 5.83058477020082e-07,
+      "loss": 0.8887,
+      "step": 62260
+    },
+    {
+      "epoch": 18.63,
+      "grad_norm": 3.8282477855682373,
+      "learning_rate": 5.817976798477731e-07,
+      "loss": 0.7917,
+      "step": 62265
+    },
+    {
+      "epoch": 18.63,
+      "grad_norm": 3.6935458183288574,
+      "learning_rate": 5.805382312673941e-07,
+      "loss": 0.82,
+      "step": 62270
+    },
+    {
+      "epoch": 18.63,
+      "grad_norm": 1.9937669038772583,
+      "learning_rate": 5.792801313485086e-07,
+      "loss": 1.113,
+      "step": 62275
+    },
+    {
+      "epoch": 18.63,
+      "grad_norm": 1.7843409776687622,
+      "learning_rate": 5.780233801605944e-07,
+      "loss": 0.9841,
+      "step": 62280
+    },
+    {
+      "epoch": 18.63,
+      "grad_norm": 2.7850582599639893,
+      "learning_rate": 5.767679777730656e-07,
+      "loss": 1.0303,
+      "step": 62285
+    },
+    {
+      "epoch": 18.64,
+      "grad_norm": 3.4673726558685303,
+      "learning_rate": 5.755139242552554e-07,
+      "loss": 0.9193,
+      "step": 62290
+    },
+    {
+      "epoch": 18.64,
+      "grad_norm": 1.380431890487671,
+      "learning_rate": 5.742612196764224e-07,
+      "loss": 0.7912,
+      "step": 62295
+    },
+    {
+      "epoch": 18.64,
+      "grad_norm": 2.402841567993164,
+      "learning_rate": 5.73009864105753e-07,
+      "loss": 0.8712,
+      "step": 62300
+    },
+    {
+      "epoch": 18.64,
+      "grad_norm": 3.139500379562378,
+      "learning_rate": 5.717598576123584e-07,
+      "loss": 0.9437,
+      "step": 62305
+    },
+    {
+      "epoch": 18.64,
+      "grad_norm": 2.209512233734131,
+      "learning_rate": 5.705112002652752e-07,
+      "loss": 0.9299,
+      "step": 62310
+    },
+    {
+      "epoch": 18.64,
+      "grad_norm": 2.3451426029205322,
+      "learning_rate": 5.692638921334647e-07,
+      "loss": 0.9714,
+      "step": 62315
+    },
+    {
+      "epoch": 18.65,
+      "grad_norm": 3.0532517433166504,
+      "learning_rate": 5.680179332858165e-07,
+      "loss": 1.1268,
+      "step": 62320
+    },
+    {
+      "epoch": 18.65,
+      "grad_norm": 6.812263488769531,
+      "learning_rate": 5.667733237911421e-07,
+      "loss": 0.786,
+      "step": 62325
+    },
+    {
+      "epoch": 18.65,
+      "grad_norm": 2.6777634620666504,
+      "learning_rate": 5.655300637181809e-07,
+      "loss": 0.8496,
+      "step": 62330
+    },
+    {
+      "epoch": 18.65,
+      "grad_norm": 1.7131210565567017,
+      "learning_rate": 5.64288153135592e-07,
+      "loss": 1.0718,
+      "step": 62335
+    },
+    {
+      "epoch": 18.65,
+      "grad_norm": 3.845153570175171,
+      "learning_rate": 5.63047592111976e-07,
+      "loss": 0.7881,
+      "step": 62340
+    },
+    {
+      "epoch": 18.65,
+      "grad_norm": 2.2182199954986572,
+      "learning_rate": 5.618083807158337e-07,
+      "loss": 0.9474,
+      "step": 62345
+    },
+    {
+      "epoch": 18.65,
+      "grad_norm": 7.1650390625,
+      "learning_rate": 5.605705190156158e-07,
+      "loss": 1.0506,
+      "step": 62350
+    },
+    {
+      "epoch": 18.66,
+      "grad_norm": 4.3074259757995605,
+      "learning_rate": 5.593340070796843e-07,
+      "loss": 1.0749,
+      "step": 62355
+    },
+    {
+      "epoch": 18.66,
+      "grad_norm": 5.269190311431885,
+      "learning_rate": 5.580988449763319e-07,
+      "loss": 0.986,
+      "step": 62360
+    },
+    {
+      "epoch": 18.66,
+      "grad_norm": 1.8356404304504395,
+      "learning_rate": 5.568650327737734e-07,
+      "loss": 0.9098,
+      "step": 62365
+    },
+    {
+      "epoch": 18.66,
+      "grad_norm": 3.584327459335327,
+      "learning_rate": 5.556325705401517e-07,
+      "loss": 1.004,
+      "step": 62370
+    },
+    {
+      "epoch": 18.66,
+      "grad_norm": 2.812471628189087,
+      "learning_rate": 5.544014583435342e-07,
+      "loss": 0.9846,
+      "step": 62375
+    },
+    {
+      "epoch": 18.66,
+      "grad_norm": 1.4029104709625244,
+      "learning_rate": 5.53171696251914e-07,
+      "loss": 0.8815,
+      "step": 62380
+    },
+    {
+      "epoch": 18.66,
+      "grad_norm": 2.412611722946167,
+      "learning_rate": 5.519432843332117e-07,
+      "loss": 1.0807,
+      "step": 62385
+    },
+    {
+      "epoch": 18.67,
+      "grad_norm": 2.5085036754608154,
+      "learning_rate": 5.507162226552676e-07,
+      "loss": 1.0554,
+      "step": 62390
+    },
+    {
+      "epoch": 18.67,
+      "grad_norm": 2.622960090637207,
+      "learning_rate": 5.494905112858551e-07,
+      "loss": 1.0882,
+      "step": 62395
+    },
+    {
+      "epoch": 18.67,
+      "grad_norm": 7.016479015350342,
+      "learning_rate": 5.482661502926645e-07,
+      "loss": 0.971,
+      "step": 62400
+    },
+    {
+      "epoch": 18.67,
+      "grad_norm": 2.4067859649658203,
+      "learning_rate": 5.470431397433196e-07,
+      "loss": 0.943,
+      "step": 62405
+    },
+    {
+      "epoch": 18.67,
+      "grad_norm": 1.8276368379592896,
+      "learning_rate": 5.458214797053634e-07,
+      "loss": 0.8402,
+      "step": 62410
+    },
+    {
+      "epoch": 18.67,
+      "grad_norm": 3.209845781326294,
+      "learning_rate": 5.4460117024627e-07,
+      "loss": 1.0835,
+      "step": 62415
+    },
+    {
+      "epoch": 18.68,
+      "grad_norm": 3.5742673873901367,
+      "learning_rate": 5.433822114334325e-07,
+      "loss": 0.9627,
+      "step": 62420
+    },
+    {
+      "epoch": 18.68,
+      "grad_norm": 3.4893949031829834,
+      "learning_rate": 5.421646033341748e-07,
+      "loss": 1.057,
+      "step": 62425
+    },
+    {
+      "epoch": 18.68,
+      "grad_norm": 3.8481268882751465,
+      "learning_rate": 5.409483460157433e-07,
+      "loss": 1.0174,
+      "step": 62430
+    },
+    {
+      "epoch": 18.68,
+      "grad_norm": 1.257150650024414,
+      "learning_rate": 5.397334395453147e-07,
+      "loss": 0.8821,
+      "step": 62435
+    },
+    {
+      "epoch": 18.68,
+      "grad_norm": 2.7487339973449707,
+      "learning_rate": 5.385198839899769e-07,
+      "loss": 0.9345,
+      "step": 62440
+    },
+    {
+      "epoch": 18.68,
+      "grad_norm": 1.8233064413070679,
+      "learning_rate": 5.373076794167653e-07,
+      "loss": 0.8765,
+      "step": 62445
+    },
+    {
+      "epoch": 18.68,
+      "grad_norm": 3.476785898208618,
+      "learning_rate": 5.360968258926152e-07,
+      "loss": 0.978,
+      "step": 62450
+    },
+    {
+      "epoch": 18.69,
+      "grad_norm": 1.5684432983398438,
+      "learning_rate": 5.348873234844176e-07,
+      "loss": 0.8142,
+      "step": 62455
+    },
+    {
+      "epoch": 18.69,
+      "grad_norm": 4.222447395324707,
+      "learning_rate": 5.336791722589579e-07,
+      "loss": 0.9887,
+      "step": 62460
+    },
+    {
+      "epoch": 18.69,
+      "grad_norm": 3.2067711353302,
+      "learning_rate": 5.324723722829661e-07,
+      "loss": 1.0059,
+      "step": 62465
+    },
+    {
+      "epoch": 18.69,
+      "grad_norm": 1.711020827293396,
+      "learning_rate": 5.312669236230944e-07,
+      "loss": 1.1326,
+      "step": 62470
+    },
+    {
+      "epoch": 18.69,
+      "grad_norm": 5.746004581451416,
+      "learning_rate": 5.300628263459173e-07,
+      "loss": 0.8101,
+      "step": 62475
+    },
+    {
+      "epoch": 18.69,
+      "grad_norm": 3.5213847160339355,
+      "learning_rate": 5.288600805179345e-07,
+      "loss": 0.9395,
+      "step": 62480
+    },
+    {
+      "epoch": 18.69,
+      "grad_norm": 3.3688313961029053,
+      "learning_rate": 5.276586862055704e-07,
+      "loss": 0.8759,
+      "step": 62485
+    },
+    {
+      "epoch": 18.7,
+      "grad_norm": 1.9452404975891113,
+      "learning_rate": 5.264586434751833e-07,
+      "loss": 0.9038,
+      "step": 62490
+    },
+    {
+      "epoch": 18.7,
+      "grad_norm": 2.7685489654541016,
+      "learning_rate": 5.252599523930452e-07,
+      "loss": 0.8489,
+      "step": 62495
+    },
+    {
+      "epoch": 18.7,
+      "grad_norm": 2.2779135704040527,
+      "learning_rate": 5.240626130253585e-07,
+      "loss": 1.0238,
+      "step": 62500
+    },
+    {
+      "epoch": 18.7,
+      "grad_norm": 2.2392897605895996,
+      "learning_rate": 5.228666254382536e-07,
+      "loss": 1.1407,
+      "step": 62505
+    },
+    {
+      "epoch": 18.7,
+      "grad_norm": 4.152332782745361,
+      "learning_rate": 5.216719896977806e-07,
+      "loss": 0.7601,
+      "step": 62510
+    },
+    {
+      "epoch": 18.7,
+      "grad_norm": 2.4010398387908936,
+      "learning_rate": 5.204787058699228e-07,
+      "loss": 0.6884,
+      "step": 62515
+    },
+    {
+      "epoch": 18.71,
+      "grad_norm": 3.122080087661743,
+      "learning_rate": 5.192867740205775e-07,
+      "loss": 1.0505,
+      "step": 62520
+    },
+    {
+      "epoch": 18.71,
+      "grad_norm": 4.57648229598999,
+      "learning_rate": 5.18096194215581e-07,
+      "loss": 0.9048,
+      "step": 62525
+    },
+    {
+      "epoch": 18.71,
+      "grad_norm": 5.7991414070129395,
+      "learning_rate": 5.169069665206833e-07,
+      "loss": 0.8917,
+      "step": 62530
+    },
+    {
+      "epoch": 18.71,
+      "grad_norm": 1.6808985471725464,
+      "learning_rate": 5.157190910015625e-07,
+      "loss": 0.9275,
+      "step": 62535
+    },
+    {
+      "epoch": 18.71,
+      "grad_norm": 2.2939209938049316,
+      "learning_rate": 5.145325677238327e-07,
+      "loss": 1.0068,
+      "step": 62540
+    },
+    {
+      "epoch": 18.71,
+      "grad_norm": 2.1694259643554688,
+      "learning_rate": 5.133473967530111e-07,
+      "loss": 1.0079,
+      "step": 62545
+    },
+    {
+      "epoch": 18.71,
+      "grad_norm": 3.0563623905181885,
+      "learning_rate": 5.121635781545647e-07,
+      "loss": 0.9808,
+      "step": 62550
+    },
+    {
+      "epoch": 18.72,
+      "grad_norm": 2.7795627117156982,
+      "learning_rate": 5.109811119938663e-07,
+      "loss": 0.9278,
+      "step": 62555
+    },
+    {
+      "epoch": 18.72,
+      "grad_norm": 2.1813108921051025,
+      "learning_rate": 5.0979999833623e-07,
+      "loss": 0.973,
+      "step": 62560
+    },
+    {
+      "epoch": 18.72,
+      "grad_norm": 1.735756754875183,
+      "learning_rate": 5.086202372468818e-07,
+      "loss": 1.0072,
+      "step": 62565
+    },
+    {
+      "epoch": 18.72,
+      "grad_norm": 1.2512030601501465,
+      "learning_rate": 5.074418287909804e-07,
+      "loss": 1.0268,
+      "step": 62570
+    },
+    {
+      "epoch": 18.72,
+      "grad_norm": 6.556924343109131,
+      "learning_rate": 5.062647730336073e-07,
+      "loss": 0.9008,
+      "step": 62575
+    },
+    {
+      "epoch": 18.72,
+      "grad_norm": 2.471876382827759,
+      "learning_rate": 5.050890700397715e-07,
+      "loss": 0.7419,
+      "step": 62580
+    },
+    {
+      "epoch": 18.72,
+      "grad_norm": 2.7054262161254883,
+      "learning_rate": 5.039147198744071e-07,
+      "loss": 1.0026,
+      "step": 62585
+    },
+    {
+      "epoch": 18.73,
+      "grad_norm": 3.4873416423797607,
+      "learning_rate": 5.027417226023678e-07,
+      "loss": 1.0207,
+      "step": 62590
+    },
+    {
+      "epoch": 18.73,
+      "grad_norm": 2.8062655925750732,
+      "learning_rate": 5.015700782884408e-07,
+      "loss": 0.9565,
+      "step": 62595
+    },
+    {
+      "epoch": 18.73,
+      "grad_norm": 1.282105565071106,
+      "learning_rate": 5.003997869973353e-07,
+      "loss": 1.0715,
+      "step": 62600
+    },
+    {
+      "epoch": 18.73,
+      "grad_norm": 2.5329504013061523,
+      "learning_rate": 4.992308487936804e-07,
+      "loss": 1.0151,
+      "step": 62605
+    },
+    {
+      "epoch": 18.73,
+      "grad_norm": 4.997149467468262,
+      "learning_rate": 4.980632637420408e-07,
+      "loss": 0.9544,
+      "step": 62610
+    },
+    {
+      "epoch": 18.73,
+      "grad_norm": 2.9110753536224365,
+      "learning_rate": 4.968970319068983e-07,
+      "loss": 0.8164,
+      "step": 62615
+    },
+    {
+      "epoch": 18.74,
+      "grad_norm": 2.0976929664611816,
+      "learning_rate": 4.957321533526626e-07,
+      "loss": 0.878,
+      "step": 62620
+    },
+    {
+      "epoch": 18.74,
+      "grad_norm": 3.1105329990386963,
+      "learning_rate": 4.94568628143674e-07,
+      "loss": 1.0102,
+      "step": 62625
+    },
+    {
+      "epoch": 18.74,
+      "grad_norm": 2.1142258644104004,
+      "learning_rate": 4.93406456344181e-07,
+      "loss": 0.9267,
+      "step": 62630
+    },
+    {
+      "epoch": 18.74,
+      "grad_norm": 2.38909912109375,
+      "learning_rate": 4.922456380183821e-07,
+      "loss": 0.8664,
+      "step": 62635
+    },
+    {
+      "epoch": 18.74,
+      "grad_norm": 3.103832721710205,
+      "learning_rate": 4.910861732303762e-07,
+      "loss": 0.9541,
+      "step": 62640
+    },
+    {
+      "epoch": 18.74,
+      "grad_norm": 3.287440299987793,
+      "learning_rate": 4.899280620442121e-07,
+      "loss": 0.986,
+      "step": 62645
+    },
+    {
+      "epoch": 18.74,
+      "grad_norm": 3.8672590255737305,
+      "learning_rate": 4.887713045238385e-07,
+      "loss": 0.9699,
+      "step": 62650
+    },
+    {
+      "epoch": 18.75,
+      "grad_norm": 2.768521785736084,
+      "learning_rate": 4.876159007331516e-07,
+      "loss": 0.9764,
+      "step": 62655
+    },
+    {
+      "epoch": 18.75,
+      "grad_norm": 1.9124183654785156,
+      "learning_rate": 4.864618507359558e-07,
+      "loss": 0.925,
+      "step": 62660
+    },
+    {
+      "epoch": 18.75,
+      "grad_norm": 6.206687927246094,
+      "learning_rate": 4.853091545959948e-07,
+      "loss": 1.0686,
+      "step": 62665
+    },
+    {
+      "epoch": 18.75,
+      "grad_norm": 1.3544474840164185,
+      "learning_rate": 4.841578123769258e-07,
+      "loss": 0.8745,
+      "step": 62670
+    },
+    {
+      "epoch": 18.75,
+      "grad_norm": 2.0739264488220215,
+      "learning_rate": 4.830078241423369e-07,
+      "loss": 0.8589,
+      "step": 62675
+    },
+    {
+      "epoch": 18.75,
+      "grad_norm": 2.00089430809021,
+      "learning_rate": 4.818591899557467e-07,
+      "loss": 0.908,
+      "step": 62680
+    },
+    {
+      "epoch": 18.75,
+      "grad_norm": 2.7553093433380127,
+      "learning_rate": 4.807119098805851e-07,
+      "loss": 1.0617,
+      "step": 62685
+    },
+    {
+      "epoch": 18.76,
+      "grad_norm": 1.7284417152404785,
+      "learning_rate": 4.795659839802181e-07,
+      "loss": 0.952,
+      "step": 62690
+    },
+    {
+      "epoch": 18.76,
+      "grad_norm": 2.9970293045043945,
+      "learning_rate": 4.784214123179337e-07,
+      "loss": 0.878,
+      "step": 62695
+    },
+    {
+      "epoch": 18.76,
+      "grad_norm": 2.638425350189209,
+      "learning_rate": 4.772781949569455e-07,
+      "loss": 0.9725,
+      "step": 62700
+    },
+    {
+      "epoch": 18.76,
+      "grad_norm": 2.378786087036133,
+      "learning_rate": 4.7613633196039455e-07,
+      "loss": 0.912,
+      "step": 62705
+    },
+    {
+      "epoch": 18.76,
+      "grad_norm": 2.5562052726745605,
+      "learning_rate": 4.749958233913415e-07,
+      "loss": 0.9821,
+      "step": 62710
+    },
+    {
+      "epoch": 18.76,
+      "grad_norm": 4.997159957885742,
+      "learning_rate": 4.7385666931277774e-07,
+      "loss": 0.998,
+      "step": 62715
+    },
+    {
+      "epoch": 18.77,
+      "grad_norm": 2.2791659832000732,
+      "learning_rate": 4.727188697876195e-07,
+      "loss": 1.0484,
+      "step": 62720
+    },
+    {
+      "epoch": 18.77,
+      "grad_norm": 3.7600674629211426,
+      "learning_rate": 4.715824248786971e-07,
+      "loss": 0.9918,
+      "step": 62725
+    },
+    {
+      "epoch": 18.77,
+      "grad_norm": 5.7446417808532715,
+      "learning_rate": 4.704473346487881e-07,
+      "loss": 0.8191,
+      "step": 62730
+    },
+    {
+      "epoch": 18.77,
+      "grad_norm": 2.3552982807159424,
+      "learning_rate": 4.693135991605702e-07,
+      "loss": 1.0591,
+      "step": 62735
+    },
+    {
+      "epoch": 18.77,
+      "grad_norm": 2.88034987449646,
+      "learning_rate": 4.681812184766682e-07,
+      "loss": 1.1121,
+      "step": 62740
+    },
+    {
+      "epoch": 18.77,
+      "grad_norm": 1.8043874502182007,
+      "learning_rate": 4.670501926596127e-07,
+      "loss": 1.0824,
+      "step": 62745
+    },
+    {
+      "epoch": 18.77,
+      "grad_norm": 2.6473546028137207,
+      "learning_rate": 4.659205217718815e-07,
+      "loss": 0.866,
+      "step": 62750
+    },
+    {
+      "epoch": 18.78,
+      "grad_norm": 2.442850351333618,
+      "learning_rate": 4.647922058758525e-07,
+      "loss": 0.9233,
+      "step": 62755
+    },
+    {
+      "epoch": 18.78,
+      "grad_norm": 2.945636510848999,
+      "learning_rate": 4.63665245033848e-07,
+      "loss": 0.8995,
+      "step": 62760
+    },
+    {
+      "epoch": 18.78,
+      "grad_norm": 2.816267490386963,
+      "learning_rate": 4.6253963930810995e-07,
+      "loss": 0.9573,
+      "step": 62765
+    },
+    {
+      "epoch": 18.78,
+      "grad_norm": 3.510456085205078,
+      "learning_rate": 4.614153887607997e-07,
+      "loss": 1.0292,
+      "step": 62770
+    },
+    {
+      "epoch": 18.78,
+      "grad_norm": 1.8083337545394897,
+      "learning_rate": 4.6029249345401484e-07,
+      "loss": 0.8119,
+      "step": 62775
+    },
+    {
+      "epoch": 18.78,
+      "grad_norm": 3.959989309310913,
+      "learning_rate": 4.5917095344976403e-07,
+      "loss": 0.9996,
+      "step": 62780
+    },
+    {
+      "epoch": 18.78,
+      "grad_norm": 6.885543346405029,
+      "learning_rate": 4.58050768809995e-07,
+      "loss": 0.9633,
+      "step": 62785
+    },
+    {
+      "epoch": 18.79,
+      "grad_norm": 2.3350930213928223,
+      "learning_rate": 4.569319395965721e-07,
+      "loss": 1.0368,
+      "step": 62790
+    },
+    {
+      "epoch": 18.79,
+      "grad_norm": 4.399080276489258,
+      "learning_rate": 4.5581446587128763e-07,
+      "loss": 1.014,
+      "step": 62795
+    },
+    {
+      "epoch": 18.79,
+      "grad_norm": 1.9063431024551392,
+      "learning_rate": 4.5469834769585604e-07,
+      "loss": 1.0557,
+      "step": 62800
+    },
+    {
+      "epoch": 18.79,
+      "grad_norm": 1.7454651594161987,
+      "learning_rate": 4.5358358513192257e-07,
+      "loss": 1.1915,
+      "step": 62805
+    },
+    {
+      "epoch": 18.79,
+      "grad_norm": 2.7510390281677246,
+      "learning_rate": 4.5247017824105174e-07,
+      "loss": 0.9034,
+      "step": 62810
+    },
+    {
+      "epoch": 18.79,
+      "grad_norm": 4.9659600257873535,
+      "learning_rate": 4.513581270847389e-07,
+      "loss": 0.7884,
+      "step": 62815
+    },
+    {
+      "epoch": 18.8,
+      "grad_norm": 4.02251672744751,
+      "learning_rate": 4.5024743172439877e-07,
+      "loss": 0.8928,
+      "step": 62820
+    },
+    {
+      "epoch": 18.8,
+      "grad_norm": 4.709519863128662,
+      "learning_rate": 4.491380922213767e-07,
+      "loss": 0.9039,
+      "step": 62825
+    },
+    {
+      "epoch": 18.8,
+      "grad_norm": 1.9701640605926514,
+      "learning_rate": 4.4803010863693483e-07,
+      "loss": 0.937,
+      "step": 62830
+    },
+    {
+      "epoch": 18.8,
+      "grad_norm": 2.070518970489502,
+      "learning_rate": 4.4692348103227143e-07,
+      "loss": 0.9715,
+      "step": 62835
+    },
+    {
+      "epoch": 18.8,
+      "grad_norm": 1.8517335653305054,
+      "learning_rate": 4.4581820946850427e-07,
+      "loss": 0.9508,
+      "step": 62840
+    },
+    {
+      "epoch": 18.8,
+      "grad_norm": 1.6975722312927246,
+      "learning_rate": 4.4471429400667343e-07,
+      "loss": 1.0122,
+      "step": 62845
+    },
+    {
+      "epoch": 18.8,
+      "grad_norm": 2.018352746963501,
+      "learning_rate": 4.436117347077495e-07,
+      "loss": 1.058,
+      "step": 62850
+    },
+    {
+      "epoch": 18.81,
+      "grad_norm": 4.113342761993408,
+      "learning_rate": 4.425105316326228e-07,
+      "loss": 0.9022,
+      "step": 62855
+    },
+    {
+      "epoch": 18.81,
+      "grad_norm": 3.3095271587371826,
+      "learning_rate": 4.414106848421168e-07,
+      "loss": 1.0346,
+      "step": 62860
+    },
+    {
+      "epoch": 18.81,
+      "grad_norm": 3.63959002494812,
+      "learning_rate": 4.4031219439696625e-07,
+      "loss": 0.9892,
+      "step": 62865
+    },
+    {
+      "epoch": 18.81,
+      "grad_norm": 4.62831974029541,
+      "learning_rate": 4.3921506035784766e-07,
+      "loss": 0.9858,
+      "step": 62870
+    },
+    {
+      "epoch": 18.81,
+      "grad_norm": 3.9106874465942383,
+      "learning_rate": 4.381192827853514e-07,
+      "loss": 0.943,
+      "step": 62875
+    },
+    {
+      "epoch": 18.81,
+      "grad_norm": 3.417293071746826,
+      "learning_rate": 4.3702486173999856e-07,
+      "loss": 0.8616,
+      "step": 62880
+    },
+    {
+      "epoch": 18.81,
+      "grad_norm": 2.981015682220459,
+      "learning_rate": 4.359317972822269e-07,
+      "loss": 1.0777,
+      "step": 62885
+    },
+    {
+      "epoch": 18.82,
+      "grad_norm": 9.559314727783203,
+      "learning_rate": 4.3484008947241304e-07,
+      "loss": 0.9715,
+      "step": 62890
+    },
+    {
+      "epoch": 18.82,
+      "grad_norm": 2.853968858718872,
+      "learning_rate": 4.33749738370845e-07,
+      "loss": 1.0899,
+      "step": 62895
+    },
+    {
+      "epoch": 18.82,
+      "grad_norm": 2.9519448280334473,
+      "learning_rate": 4.326607440377439e-07,
+      "loss": 0.9593,
+      "step": 62900
+    },
+    {
+      "epoch": 18.82,
+      "grad_norm": 3.299821376800537,
+      "learning_rate": 4.3157310653325344e-07,
+      "loss": 0.9734,
+      "step": 62905
+    },
+    {
+      "epoch": 18.82,
+      "grad_norm": 2.5960991382598877,
+      "learning_rate": 4.3048682591744495e-07,
+      "loss": 1.0205,
+      "step": 62910
+    },
+    {
+      "epoch": 18.82,
+      "grad_norm": 1.4669679403305054,
+      "learning_rate": 4.294019022503037e-07,
+      "loss": 0.9927,
+      "step": 62915
+    },
+    {
+      "epoch": 18.82,
+      "grad_norm": 3.0329182147979736,
+      "learning_rate": 4.283183355917625e-07,
+      "loss": 1.1614,
+      "step": 62920
+    },
+    {
+      "epoch": 18.83,
+      "grad_norm": 3.468059778213501,
+      "learning_rate": 4.272361260016566e-07,
+      "loss": 0.7337,
+      "step": 62925
+    },
+    {
+      "epoch": 18.83,
+      "grad_norm": 8.143178939819336,
+      "learning_rate": 4.2615527353975783e-07,
+      "loss": 0.833,
+      "step": 62930
+    },
+    {
+      "epoch": 18.83,
+      "grad_norm": 1.7140194177627563,
+      "learning_rate": 4.2507577826575995e-07,
+      "loss": 0.9572,
+      "step": 62935
+    },
+    {
+      "epoch": 18.83,
+      "grad_norm": 2.0888423919677734,
+      "learning_rate": 4.23997640239282e-07,
+      "loss": 1.0221,
+      "step": 62940
+    },
+    {
+      "epoch": 18.83,
+      "grad_norm": 2.820460319519043,
+      "learning_rate": 4.2292085951987084e-07,
+      "loss": 0.8878,
+      "step": 62945
+    },
+    {
+      "epoch": 18.83,
+      "grad_norm": 4.351651668548584,
+      "learning_rate": 4.218454361669899e-07,
+      "loss": 0.8856,
+      "step": 62950
+    },
+    {
+      "epoch": 18.84,
+      "grad_norm": 1.5819222927093506,
+      "learning_rate": 4.2077137024004455e-07,
+      "loss": 1.04,
+      "step": 62955
+    },
+    {
+      "epoch": 18.84,
+      "grad_norm": 3.579989194869995,
+      "learning_rate": 4.1969866179834284e-07,
+      "loss": 1.0235,
+      "step": 62960
+    },
+    {
+      "epoch": 18.84,
+      "grad_norm": 2.6302943229675293,
+      "learning_rate": 4.1862731090113736e-07,
+      "loss": 0.9574,
+      "step": 62965
+    },
+    {
+      "epoch": 18.84,
+      "grad_norm": 4.829039573669434,
+      "learning_rate": 4.175573176075892e-07,
+      "loss": 1.0367,
+      "step": 62970
+    },
+    {
+      "epoch": 18.84,
+      "grad_norm": 3.0657598972320557,
+      "learning_rate": 4.164886819768038e-07,
+      "loss": 0.855,
+      "step": 62975
+    },
+    {
+      "epoch": 18.84,
+      "grad_norm": 3.936453342437744,
+      "learning_rate": 4.1542140406779504e-07,
+      "loss": 1.1114,
+      "step": 62980
+    },
+    {
+      "epoch": 18.84,
+      "grad_norm": 3.4842371940612793,
+      "learning_rate": 4.1435548393950474e-07,
+      "loss": 0.9725,
+      "step": 62985
+    },
+    {
+      "epoch": 18.85,
+      "grad_norm": 2.537079334259033,
+      "learning_rate": 4.1329092165080794e-07,
+      "loss": 0.9701,
+      "step": 62990
+    },
+    {
+      "epoch": 18.85,
+      "grad_norm": 2.407586097717285,
+      "learning_rate": 4.122277172604966e-07,
+      "loss": 0.8825,
+      "step": 62995
+    },
+    {
+      "epoch": 18.85,
+      "grad_norm": 1.6151124238967896,
+      "learning_rate": 4.111658708272903e-07,
+      "loss": 1.1248,
+      "step": 63000
+    },
+    {
+      "epoch": 18.85,
+      "grad_norm": 3.798203706741333,
+      "learning_rate": 4.1010538240983667e-07,
+      "loss": 1.0407,
+      "step": 63005
+    },
+    {
+      "epoch": 18.85,
+      "grad_norm": 3.2758753299713135,
+      "learning_rate": 4.090462520666999e-07,
+      "loss": 1.033,
+      "step": 63010
+    },
+    {
+      "epoch": 18.85,
+      "grad_norm": 3.0684893131256104,
+      "learning_rate": 4.0798847985637766e-07,
+      "loss": 0.8769,
+      "step": 63015
+    },
+    {
+      "epoch": 18.85,
+      "grad_norm": 2.2949230670928955,
+      "learning_rate": 4.0693206583729273e-07,
+      "loss": 1.0484,
+      "step": 63020
+    },
+    {
+      "epoch": 18.86,
+      "grad_norm": 3.716773271560669,
+      "learning_rate": 4.0587701006778443e-07,
+      "loss": 1.0033,
+      "step": 63025
+    },
+    {
+      "epoch": 18.86,
+      "grad_norm": 1.4224724769592285,
+      "learning_rate": 4.048233126061257e-07,
+      "loss": 0.9085,
+      "step": 63030
+    },
+    {
+      "epoch": 18.86,
+      "grad_norm": 1.496864676475525,
+      "learning_rate": 4.037709735105089e-07,
+      "loss": 0.8798,
+      "step": 63035
+    },
+    {
+      "epoch": 18.86,
+      "grad_norm": 3.2750120162963867,
+      "learning_rate": 4.0271999283905683e-07,
+      "loss": 0.954,
+      "step": 63040
+    },
+    {
+      "epoch": 18.86,
+      "grad_norm": 3.3160552978515625,
+      "learning_rate": 4.0167037064980925e-07,
+      "loss": 0.9068,
+      "step": 63045
+    },
+    {
+      "epoch": 18.86,
+      "grad_norm": 4.469260215759277,
+      "learning_rate": 4.006221070007421e-07,
+      "loss": 0.8265,
+      "step": 63050
+    },
+    {
+      "epoch": 18.87,
+      "grad_norm": 2.6699278354644775,
+      "learning_rate": 3.9957520194974505e-07,
+      "loss": 1.0278,
+      "step": 63055
+    },
+    {
+      "epoch": 18.87,
+      "grad_norm": 2.068089485168457,
+      "learning_rate": 3.9852965555463863e-07,
+      "loss": 0.9476,
+      "step": 63060
+    },
+    {
+      "epoch": 18.87,
+      "grad_norm": 1.7394936084747314,
+      "learning_rate": 3.974854678731654e-07,
+      "loss": 0.9202,
+      "step": 63065
+    },
+    {
+      "epoch": 18.87,
+      "grad_norm": 5.004978656768799,
+      "learning_rate": 3.964426389630016e-07,
+      "loss": 1.0277,
+      "step": 63070
+    },
+    {
+      "epoch": 18.87,
+      "grad_norm": 2.368013620376587,
+      "learning_rate": 3.954011688817344e-07,
+      "loss": 0.9778,
+      "step": 63075
+    },
+    {
+      "epoch": 18.87,
+      "grad_norm": 1.3884702920913696,
+      "learning_rate": 3.943610576868845e-07,
+      "loss": 0.8678,
+      "step": 63080
+    },
+    {
+      "epoch": 18.87,
+      "grad_norm": 1.4735065698623657,
+      "learning_rate": 3.9332230543589753e-07,
+      "loss": 1.0603,
+      "step": 63085
+    },
+    {
+      "epoch": 18.88,
+      "grad_norm": 2.9235053062438965,
+      "learning_rate": 3.922849121861416e-07,
+      "loss": 0.9802,
+      "step": 63090
+    },
+    {
+      "epoch": 18.88,
+      "grad_norm": 1.8659436702728271,
+      "learning_rate": 3.912488779949153e-07,
+      "loss": 1.0448,
+      "step": 63095
+    },
+    {
+      "epoch": 18.88,
+      "grad_norm": 3.8090038299560547,
+      "learning_rate": 3.902142029194311e-07,
+      "loss": 0.9592,
+      "step": 63100
+    },
+    {
+      "epoch": 18.88,
+      "grad_norm": 2.0518362522125244,
+      "learning_rate": 3.891808870168351e-07,
+      "loss": 0.8467,
+      "step": 63105
+    },
+    {
+      "epoch": 18.88,
+      "grad_norm": 2.6214382648468018,
+      "learning_rate": 3.8814893034420097e-07,
+      "loss": 0.9282,
+      "step": 63110
+    },
+    {
+      "epoch": 18.88,
+      "grad_norm": 4.084232807159424,
+      "learning_rate": 3.8711833295851654e-07,
+      "loss": 0.9546,
+      "step": 63115
+    },
+    {
+      "epoch": 18.88,
+      "grad_norm": 1.4593278169631958,
+      "learning_rate": 3.8608909491670287e-07,
+      "loss": 0.803,
+      "step": 63120
+    },
+    {
+      "epoch": 18.89,
+      "grad_norm": 3.1608002185821533,
+      "learning_rate": 3.8506121627560345e-07,
+      "loss": 0.9568,
+      "step": 63125
+    },
+    {
+      "epoch": 18.89,
+      "grad_norm": 2.8863463401794434,
+      "learning_rate": 3.840346970919867e-07,
+      "loss": 1.0497,
+      "step": 63130
+    },
+    {
+      "epoch": 18.89,
+      "grad_norm": 1.954045057296753,
+      "learning_rate": 3.8300953742254895e-07,
+      "loss": 0.9635,
+      "step": 63135
+    },
+    {
+      "epoch": 18.89,
+      "grad_norm": 2.1377789974212646,
+      "learning_rate": 3.819857373239033e-07,
+      "loss": 0.7992,
+      "step": 63140
+    },
+    {
+      "epoch": 18.89,
+      "grad_norm": 2.493187665939331,
+      "learning_rate": 3.809632968526017e-07,
+      "loss": 1.0234,
+      "step": 63145
+    },
+    {
+      "epoch": 18.89,
+      "grad_norm": 2.3102526664733887,
+      "learning_rate": 3.799422160651017e-07,
+      "loss": 0.8951,
+      "step": 63150
+    },
+    {
+      "epoch": 18.9,
+      "grad_norm": 2.773439645767212,
+      "learning_rate": 3.7892249501780554e-07,
+      "loss": 0.8641,
+      "step": 63155
+    },
+    {
+      "epoch": 18.9,
+      "grad_norm": 2.822577953338623,
+      "learning_rate": 3.779041337670236e-07,
+      "loss": 0.9386,
+      "step": 63160
+    },
+    {
+      "epoch": 18.9,
+      "grad_norm": 3.764552116394043,
+      "learning_rate": 3.7688713236900816e-07,
+      "loss": 0.9324,
+      "step": 63165
+    },
+    {
+      "epoch": 18.9,
+      "grad_norm": 2.5591769218444824,
+      "learning_rate": 3.758714908799199e-07,
+      "loss": 0.925,
+      "step": 63170
+    },
+    {
+      "epoch": 18.9,
+      "grad_norm": 1.3700811862945557,
+      "learning_rate": 3.7485720935585553e-07,
+      "loss": 0.8529,
+      "step": 63175
+    },
+    {
+      "epoch": 18.9,
+      "grad_norm": 2.099714517593384,
+      "learning_rate": 3.7384428785282867e-07,
+      "loss": 0.9949,
+      "step": 63180
+    },
+    {
+      "epoch": 18.9,
+      "grad_norm": 3.3494229316711426,
+      "learning_rate": 3.7283272642678335e-07,
+      "loss": 0.7835,
+      "step": 63185
+    },
+    {
+      "epoch": 18.91,
+      "grad_norm": 2.3654308319091797,
+      "learning_rate": 3.718225251335916e-07,
+      "loss": 0.9557,
+      "step": 63190
+    },
+    {
+      "epoch": 18.91,
+      "grad_norm": 3.252896547317505,
+      "learning_rate": 3.7081368402903935e-07,
+      "loss": 1.0649,
+      "step": 63195
+    },
+    {
+      "epoch": 18.91,
+      "grad_norm": 2.0601794719696045,
+      "learning_rate": 3.6980620316884876e-07,
+      "loss": 0.9137,
+      "step": 63200
+    },
+    {
+      "epoch": 18.91,
+      "grad_norm": 3.0151379108428955,
+      "learning_rate": 3.688000826086585e-07,
+      "loss": 1.0903,
+      "step": 63205
+    },
+    {
+      "epoch": 18.91,
+      "grad_norm": 2.6133904457092285,
+      "learning_rate": 3.6779532240403537e-07,
+      "loss": 1.0124,
+      "step": 63210
+    },
+    {
+      "epoch": 18.91,
+      "grad_norm": 5.491546154022217,
+      "learning_rate": 3.6679192261047655e-07,
+      "loss": 0.9035,
+      "step": 63215
+    },
+    {
+      "epoch": 18.91,
+      "grad_norm": 2.6490890979766846,
+      "learning_rate": 3.6578988328339334e-07,
+      "loss": 1.0075,
+      "step": 63220
+    },
+    {
+      "epoch": 18.92,
+      "grad_norm": 2.415147542953491,
+      "learning_rate": 3.647892044781276e-07,
+      "loss": 1.0301,
+      "step": 63225
+    },
+    {
+      "epoch": 18.92,
+      "grad_norm": 6.828542709350586,
+      "learning_rate": 3.637898862499489e-07,
+      "loss": 1.1634,
+      "step": 63230
+    },
+    {
+      "epoch": 18.92,
+      "grad_norm": 1.5809046030044556,
+      "learning_rate": 3.6279192865404654e-07,
+      "loss": 0.9714,
+      "step": 63235
+    },
+    {
+      "epoch": 18.92,
+      "grad_norm": 2.416311264038086,
+      "learning_rate": 3.6179533174553746e-07,
+      "loss": 1.1117,
+      "step": 63240
+    },
+    {
+      "epoch": 18.92,
+      "grad_norm": 3.4020678997039795,
+      "learning_rate": 3.6080009557946104e-07,
+      "loss": 0.7682,
+      "step": 63245
+    },
+    {
+      "epoch": 18.92,
+      "grad_norm": 1.7761116027832031,
+      "learning_rate": 3.5980622021078716e-07,
+      "loss": 0.959,
+      "step": 63250
+    },
+    {
+      "epoch": 18.93,
+      "grad_norm": 1.5695528984069824,
+      "learning_rate": 3.5881370569439964e-07,
+      "loss": 0.9388,
+      "step": 63255
+    },
+    {
+      "epoch": 18.93,
+      "grad_norm": 1.8310515880584717,
+      "learning_rate": 3.5782255208512136e-07,
+      "loss": 0.9259,
+      "step": 63260
+    },
+    {
+      "epoch": 18.93,
+      "grad_norm": 2.843670129776001,
+      "learning_rate": 3.5683275943768914e-07,
+      "loss": 1.0009,
+      "step": 63265
+    },
+    {
+      "epoch": 18.93,
+      "grad_norm": 3.0472795963287354,
+      "learning_rate": 3.5584432780676745e-07,
+      "loss": 0.8153,
+      "step": 63270
+    },
+    {
+      "epoch": 18.93,
+      "grad_norm": 2.187629461288452,
+      "learning_rate": 3.548572572469461e-07,
+      "loss": 0.9146,
+      "step": 63275
+    },
+    {
+      "epoch": 18.93,
+      "grad_norm": 2.447866201400757,
+      "learning_rate": 3.5387154781274244e-07,
+      "loss": 0.929,
+      "step": 63280
+    },
+    {
+      "epoch": 18.93,
+      "grad_norm": 2.634216785430908,
+      "learning_rate": 3.528871995585964e-07,
+      "loss": 0.9642,
+      "step": 63285
+    },
+    {
+      "epoch": 18.94,
+      "grad_norm": 1.427452802658081,
+      "learning_rate": 3.5190421253886717e-07,
+      "loss": 1.1127,
+      "step": 63290
+    },
+    {
+      "epoch": 18.94,
+      "grad_norm": 2.087641716003418,
+      "learning_rate": 3.5092258680785305e-07,
+      "loss": 1.0386,
+      "step": 63295
+    },
+    {
+      "epoch": 18.94,
+      "grad_norm": 1.5409181118011475,
+      "learning_rate": 3.4994232241976065e-07,
+      "loss": 0.9281,
+      "step": 63300
+    },
+    {
+      "epoch": 18.94,
+      "grad_norm": 2.0599822998046875,
+      "learning_rate": 3.4896341942873e-07,
+      "loss": 0.9306,
+      "step": 63305
+    },
+    {
+      "epoch": 18.94,
+      "grad_norm": 3.1682193279266357,
+      "learning_rate": 3.4798587788882895e-07,
+      "loss": 1.0272,
+      "step": 63310
+    },
+    {
+      "epoch": 18.94,
+      "grad_norm": 2.4055306911468506,
+      "learning_rate": 3.4700969785404214e-07,
+      "loss": 0.9128,
+      "step": 63315
+    },
+    {
+      "epoch": 18.94,
+      "grad_norm": 2.880706548690796,
+      "learning_rate": 3.460348793782847e-07,
+      "loss": 0.8306,
+      "step": 63320
+    },
+    {
+      "epoch": 18.95,
+      "grad_norm": 1.1103742122650146,
+      "learning_rate": 3.450614225153942e-07,
+      "loss": 1.0068,
+      "step": 63325
+    },
+    {
+      "epoch": 18.95,
+      "grad_norm": 3.7383155822753906,
+      "learning_rate": 3.440893273191331e-07,
+      "loss": 0.9928,
+      "step": 63330
+    },
+    {
+      "epoch": 18.95,
+      "grad_norm": 3.4909322261810303,
+      "learning_rate": 3.431185938431919e-07,
+      "loss": 1.0227,
+      "step": 63335
+    },
+    {
+      "epoch": 18.95,
+      "grad_norm": 2.2891860008239746,
+      "learning_rate": 3.421492221411804e-07,
+      "loss": 0.9881,
+      "step": 63340
+    },
+    {
+      "epoch": 18.95,
+      "grad_norm": 1.8985334634780884,
+      "learning_rate": 3.4118121226663913e-07,
+      "loss": 0.9726,
+      "step": 63345
+    },
+    {
+      "epoch": 18.95,
+      "grad_norm": 2.307614803314209,
+      "learning_rate": 3.402145642730226e-07,
+      "loss": 1.0593,
+      "step": 63350
+    },
+    {
+      "epoch": 18.96,
+      "grad_norm": 1.9816555976867676,
+      "learning_rate": 3.3924927821372965e-07,
+      "loss": 0.8479,
+      "step": 63355
+    },
+    {
+      "epoch": 18.96,
+      "grad_norm": 3.931323289871216,
+      "learning_rate": 3.3828535414206217e-07,
+      "loss": 0.8215,
+      "step": 63360
+    },
+    {
+      "epoch": 18.96,
+      "grad_norm": 2.6754205226898193,
+      "learning_rate": 3.373227921112609e-07,
+      "loss": 1.1578,
+      "step": 63365
+    },
+    {
+      "epoch": 18.96,
+      "grad_norm": 3.3269999027252197,
+      "learning_rate": 3.3636159217448617e-07,
+      "loss": 0.9295,
+      "step": 63370
+    },
+    {
+      "epoch": 18.96,
+      "grad_norm": 3.251188278198242,
+      "learning_rate": 3.354017543848259e-07,
+      "loss": 0.9049,
+      "step": 63375
+    },
+    {
+      "epoch": 18.96,
+      "grad_norm": 1.9303770065307617,
+      "learning_rate": 3.344432787952878e-07,
+      "loss": 0.9681,
+      "step": 63380
+    },
+    {
+      "epoch": 18.96,
+      "grad_norm": 3.30509614944458,
+      "learning_rate": 3.3348616545880727e-07,
+      "loss": 0.9311,
+      "step": 63385
+    },
+    {
+      "epoch": 18.97,
+      "grad_norm": 3.3238556385040283,
+      "learning_rate": 3.325304144282476e-07,
+      "loss": 0.9782,
+      "step": 63390
+    },
+    {
+      "epoch": 18.97,
+      "grad_norm": 1.8988010883331299,
+      "learning_rate": 3.315760257563943e-07,
+      "loss": 0.8005,
+      "step": 63395
+    },
+    {
+      "epoch": 18.97,
+      "grad_norm": 2.3984742164611816,
+      "learning_rate": 3.306229994959553e-07,
+      "loss": 0.8219,
+      "step": 63400
+    },
+    {
+      "epoch": 18.97,
+      "grad_norm": 1.133448839187622,
+      "learning_rate": 3.2967133569956344e-07,
+      "loss": 1.0835,
+      "step": 63405
+    },
+    {
+      "epoch": 18.97,
+      "grad_norm": 2.583820343017578,
+      "learning_rate": 3.287210344197822e-07,
+      "loss": 0.8689,
+      "step": 63410
+    },
+    {
+      "epoch": 18.97,
+      "grad_norm": 2.5700812339782715,
+      "learning_rate": 3.2777209570909464e-07,
+      "loss": 0.8236,
+      "step": 63415
+    },
+    {
+      "epoch": 18.97,
+      "grad_norm": 3.197221279144287,
+      "learning_rate": 3.2682451961990603e-07,
+      "loss": 1.0251,
+      "step": 63420
+    },
+    {
+      "epoch": 18.98,
+      "grad_norm": 2.2375690937042236,
+      "learning_rate": 3.2587830620455507e-07,
+      "loss": 0.8646,
+      "step": 63425
+    },
+    {
+      "epoch": 18.98,
+      "grad_norm": 2.5171115398406982,
+      "learning_rate": 3.249334555153e-07,
+      "loss": 1.1793,
+      "step": 63430
+    },
+    {
+      "epoch": 18.98,
+      "grad_norm": 2.9206202030181885,
+      "learning_rate": 3.239899676043184e-07,
+      "loss": 1.1318,
+      "step": 63435
+    },
+    {
+      "epoch": 18.98,
+      "grad_norm": 2.8934311866760254,
+      "learning_rate": 3.230478425237243e-07,
+      "loss": 1.0891,
+      "step": 63440
+    },
+    {
+      "epoch": 18.98,
+      "grad_norm": 2.022392988204956,
+      "learning_rate": 3.221070803255427e-07,
+      "loss": 1.0109,
+      "step": 63445
+    },
+    {
+      "epoch": 18.98,
+      "grad_norm": 2.2447142601013184,
+      "learning_rate": 3.2116768106174035e-07,
+      "loss": 1.0619,
+      "step": 63450
+    },
+    {
+      "epoch": 18.99,
+      "grad_norm": 3.3541789054870605,
+      "learning_rate": 3.2022964478419235e-07,
+      "loss": 1.1107,
+      "step": 63455
+    },
+    {
+      "epoch": 18.99,
+      "grad_norm": 2.8497142791748047,
+      "learning_rate": 3.192929715447102e-07,
+      "loss": 0.9282,
+      "step": 63460
+    },
+    {
+      "epoch": 18.99,
+      "grad_norm": 2.122657537460327,
+      "learning_rate": 3.1835766139502174e-07,
+      "loss": 0.87,
+      "step": 63465
+    },
+    {
+      "epoch": 18.99,
+      "grad_norm": 2.5856778621673584,
+      "learning_rate": 3.1742371438678586e-07,
+      "loss": 1.0344,
+      "step": 63470
+    },
+    {
+      "epoch": 18.99,
+      "grad_norm": 3.2936534881591797,
+      "learning_rate": 3.1649113057158066e-07,
+      "loss": 0.9697,
+      "step": 63475
+    },
+    {
+      "epoch": 18.99,
+      "grad_norm": 1.6257001161575317,
+      "learning_rate": 3.1555991000091214e-07,
+      "loss": 0.9946,
+      "step": 63480
+    },
+    {
+      "epoch": 18.99,
+      "grad_norm": 3.2520930767059326,
+      "learning_rate": 3.1463005272621416e-07,
+      "loss": 0.8589,
+      "step": 63485
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 3.0452685356140137,
+      "learning_rate": 3.1370155879883735e-07,
+      "loss": 0.9231,
+      "step": 63490
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 1.2502514123916626,
+      "learning_rate": 3.1277442827006564e-07,
+      "loss": 0.9966,
+      "step": 63495
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 3.0072884559631348,
+      "learning_rate": 3.118486611910998e-07,
+      "loss": 0.9674,
+      "step": 63500
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 1.7312711477279663,
+      "learning_rate": 3.109242576130711e-07,
+      "loss": 0.857,
+      "step": 63505
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 1.629128336906433,
+      "learning_rate": 3.100012175870304e-07,
+      "loss": 0.9265,
+      "step": 63510
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 1.9464771747589111,
+      "learning_rate": 3.0907954116396185e-07,
+      "loss": 0.9412,
+      "step": 63515
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 1.9227700233459473,
+      "learning_rate": 3.0815922839476376e-07,
+      "loss": 0.6606,
+      "step": 63520
+    },
+    {
+      "epoch": 19.01,
+      "grad_norm": 3.476877450942993,
+      "learning_rate": 3.0724027933026757e-07,
+      "loss": 1.0086,
+      "step": 63525
+    },
+    {
+      "epoch": 19.01,
+      "grad_norm": 3.3601372241973877,
+      "learning_rate": 3.063226940212216e-07,
+      "loss": 0.9579,
+      "step": 63530
+    },
+    {
+      "epoch": 19.01,
+      "grad_norm": 2.9032952785491943,
+      "learning_rate": 3.0540647251830755e-07,
+      "loss": 1.0735,
+      "step": 63535
+    },
+    {
+      "epoch": 19.01,
+      "grad_norm": 1.6308649778366089,
+      "learning_rate": 3.0449161487212384e-07,
+      "loss": 1.1542,
+      "step": 63540
+    },
+    {
+      "epoch": 19.01,
+      "grad_norm": 2.7309157848358154,
+      "learning_rate": 3.035781211331995e-07,
+      "loss": 0.8411,
+      "step": 63545
+    },
+    {
+      "epoch": 19.01,
+      "grad_norm": 9.992263793945312,
+      "learning_rate": 3.026659913519858e-07,
+      "loss": 0.9904,
+      "step": 63550
+    },
+    {
+      "epoch": 19.01,
+      "grad_norm": 1.4755282402038574,
+      "learning_rate": 3.017552255788564e-07,
+      "loss": 1.1822,
+      "step": 63555
+    },
+    {
+      "epoch": 19.02,
+      "grad_norm": 3.669390916824341,
+      "learning_rate": 3.008458238641154e-07,
+      "loss": 1.1826,
+      "step": 63560
+    },
+    {
+      "epoch": 19.02,
+      "grad_norm": 2.0326449871063232,
+      "learning_rate": 2.9993778625798383e-07,
+      "loss": 0.9136,
+      "step": 63565
+    },
+    {
+      "epoch": 19.02,
+      "grad_norm": 2.4299309253692627,
+      "learning_rate": 2.9903111281061604e-07,
+      "loss": 0.891,
+      "step": 63570
+    },
+    {
+      "epoch": 19.02,
+      "grad_norm": 1.997877836227417,
+      "learning_rate": 2.98125803572083e-07,
+      "loss": 0.8897,
+      "step": 63575
+    },
+    {
+      "epoch": 19.02,
+      "grad_norm": 2.001340866088867,
+      "learning_rate": 2.972218585923864e-07,
+      "loss": 0.9927,
+      "step": 63580
+    },
+    {
+      "epoch": 19.02,
+      "grad_norm": 1.1060036420822144,
+      "learning_rate": 2.963192779214502e-07,
+      "loss": 1.1492,
+      "step": 63585
+    },
+    {
+      "epoch": 19.03,
+      "grad_norm": 2.4950671195983887,
+      "learning_rate": 2.9541806160912346e-07,
+      "loss": 0.8886,
+      "step": 63590
+    },
+    {
+      "epoch": 19.03,
+      "grad_norm": 1.5963852405548096,
+      "learning_rate": 2.9451820970517466e-07,
+      "loss": 0.9515,
+      "step": 63595
+    },
+    {
+      "epoch": 19.03,
+      "grad_norm": 2.513756036758423,
+      "learning_rate": 2.9361972225930845e-07,
+      "loss": 0.9299,
+      "step": 63600
+    },
+    {
+      "epoch": 19.03,
+      "grad_norm": 2.3485164642333984,
+      "learning_rate": 2.927225993211408e-07,
+      "loss": 0.7861,
+      "step": 63605
+    },
+    {
+      "epoch": 19.03,
+      "grad_norm": 4.739649772644043,
+      "learning_rate": 2.9182684094022363e-07,
+      "loss": 1.0509,
+      "step": 63610
+    },
+    {
+      "epoch": 19.03,
+      "grad_norm": 2.6401422023773193,
+      "learning_rate": 2.909324471660285e-07,
+      "loss": 0.9599,
+      "step": 63615
+    },
+    {
+      "epoch": 19.03,
+      "grad_norm": 3.8304197788238525,
+      "learning_rate": 2.900394180479521e-07,
+      "loss": 0.9171,
+      "step": 63620
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 1.489724040031433,
+      "learning_rate": 2.891477536353104e-07,
+      "loss": 1.1215,
+      "step": 63625
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 3.6274325847625732,
+      "learning_rate": 2.8825745397735584e-07,
+      "loss": 0.8733,
+      "step": 63630
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 2.480381965637207,
+      "learning_rate": 2.873685191232517e-07,
+      "loss": 0.9707,
+      "step": 63635
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 3.279183864593506,
+      "learning_rate": 2.8648094912210044e-07,
+      "loss": 0.9859,
+      "step": 63640
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 6.223374843597412,
+      "learning_rate": 2.8559474402291564e-07,
+      "loss": 0.9295,
+      "step": 63645
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 2.4016520977020264,
+      "learning_rate": 2.8470990387464424e-07,
+      "loss": 0.8226,
+      "step": 63650
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 4.2338666915893555,
+      "learning_rate": 2.8382642872615826e-07,
+      "loss": 0.955,
+      "step": 63655
+    },
+    {
+      "epoch": 19.05,
+      "grad_norm": 3.539830446243286,
+      "learning_rate": 2.829443186262437e-07,
+      "loss": 1.1181,
+      "step": 63660
+    },
+    {
+      "epoch": 19.05,
+      "grad_norm": 3.2982068061828613,
+      "learning_rate": 2.8206357362362543e-07,
+      "loss": 1.0266,
+      "step": 63665
+    },
+    {
+      "epoch": 19.05,
+      "grad_norm": 1.85117769241333,
+      "learning_rate": 2.811841937669396e-07,
+      "loss": 1.1086,
+      "step": 63670
+    },
+    {
+      "epoch": 19.05,
+      "grad_norm": 4.956904888153076,
+      "learning_rate": 2.8030617910475845e-07,
+      "loss": 0.9842,
+      "step": 63675
+    },
+    {
+      "epoch": 19.05,
+      "grad_norm": 2.6763927936553955,
+      "learning_rate": 2.794295296855709e-07,
+      "loss": 1.1752,
+      "step": 63680
+    },
+    {
+      "epoch": 19.05,
+      "grad_norm": 2.8290627002716064,
+      "learning_rate": 2.785542455577994e-07,
+      "loss": 0.9809,
+      "step": 63685
+    },
+    {
+      "epoch": 19.06,
+      "grad_norm": 2.090080499649048,
+      "learning_rate": 2.776803267697775e-07,
+      "loss": 1.065,
+      "step": 63690
+    },
+    {
+      "epoch": 19.06,
+      "grad_norm": 1.995445728302002,
+      "learning_rate": 2.768077733697749e-07,
+      "loss": 0.9295,
+      "step": 63695
+    },
+    {
+      "epoch": 19.06,
+      "grad_norm": 3.277020215988159,
+      "learning_rate": 2.759365854059781e-07,
+      "loss": 0.9197,
+      "step": 63700
+    },
+    {
+      "epoch": 19.06,
+      "grad_norm": 2.4329395294189453,
+      "learning_rate": 2.750667629265069e-07,
+      "loss": 1.0294,
+      "step": 63705
+    },
+    {
+      "epoch": 19.06,
+      "grad_norm": 3.564478874206543,
+      "learning_rate": 2.741983059793979e-07,
+      "loss": 1.0323,
+      "step": 63710
+    },
+    {
+      "epoch": 19.06,
+      "grad_norm": 1.9448779821395874,
+      "learning_rate": 2.7333121461261545e-07,
+      "loss": 0.8142,
+      "step": 63715
+    },
+    {
+      "epoch": 19.06,
+      "grad_norm": 2.6165874004364014,
+      "learning_rate": 2.724654888740463e-07,
+      "loss": 0.8353,
+      "step": 63720
+    },
+    {
+      "epoch": 19.07,
+      "grad_norm": 3.1223433017730713,
+      "learning_rate": 2.716011288115078e-07,
+      "loss": 0.9984,
+      "step": 63725
+    },
+    {
+      "epoch": 19.07,
+      "grad_norm": 3.356553554534912,
+      "learning_rate": 2.7073813447273386e-07,
+      "loss": 0.9196,
+      "step": 63730
+    },
+    {
+      "epoch": 19.07,
+      "grad_norm": 1.8804484605789185,
+      "learning_rate": 2.698765059053865e-07,
+      "loss": 0.9712,
+      "step": 63735
+    },
+    {
+      "epoch": 19.07,
+      "grad_norm": 2.733215570449829,
+      "learning_rate": 2.690162431570553e-07,
+      "loss": 0.9342,
+      "step": 63740
+    },
+    {
+      "epoch": 19.07,
+      "grad_norm": 2.573293685913086,
+      "learning_rate": 2.6815734627525235e-07,
+      "loss": 0.8668,
+      "step": 63745
+    },
+    {
+      "epoch": 19.07,
+      "grad_norm": 2.0941619873046875,
+      "learning_rate": 2.672998153074119e-07,
+      "loss": 0.8658,
+      "step": 63750
+    },
+    {
+      "epoch": 19.07,
+      "grad_norm": 4.143740653991699,
+      "learning_rate": 2.6644365030089046e-07,
+      "loss": 0.9723,
+      "step": 63755
+    },
+    {
+      "epoch": 19.08,
+      "grad_norm": 2.708742141723633,
+      "learning_rate": 2.6558885130298363e-07,
+      "loss": 0.8001,
+      "step": 63760
+    },
+    {
+      "epoch": 19.08,
+      "grad_norm": 1.1027653217315674,
+      "learning_rate": 2.647354183608869e-07,
+      "loss": 0.8632,
+      "step": 63765
+    },
+    {
+      "epoch": 19.08,
+      "grad_norm": 5.099153995513916,
+      "learning_rate": 2.638833515217487e-07,
+      "loss": 0.9992,
+      "step": 63770
+    },
+    {
+      "epoch": 19.08,
+      "grad_norm": 2.2473297119140625,
+      "learning_rate": 2.6303265083261753e-07,
+      "loss": 1.179,
+      "step": 63775
+    },
+    {
+      "epoch": 19.08,
+      "grad_norm": 2.4921000003814697,
+      "learning_rate": 2.621833163404835e-07,
+      "loss": 0.8985,
+      "step": 63780
+    },
+    {
+      "epoch": 19.08,
+      "grad_norm": 2.435410261154175,
+      "learning_rate": 2.6133534809224813e-07,
+      "loss": 1.155,
+      "step": 63785
+    },
+    {
+      "epoch": 19.09,
+      "grad_norm": 1.8930751085281372,
+      "learning_rate": 2.6048874613474884e-07,
+      "loss": 0.9032,
+      "step": 63790
+    },
+    {
+      "epoch": 19.09,
+      "grad_norm": 2.620929718017578,
+      "learning_rate": 2.596435105147399e-07,
+      "loss": 1.0091,
+      "step": 63795
+    },
+    {
+      "epoch": 19.09,
+      "grad_norm": 3.0546348094940186,
+      "learning_rate": 2.587996412789034e-07,
+      "loss": 0.9915,
+      "step": 63800
+    },
+    {
+      "epoch": 19.09,
+      "grad_norm": 2.1018009185791016,
+      "learning_rate": 2.579571384738466e-07,
+      "loss": 0.9352,
+      "step": 63805
+    },
+    {
+      "epoch": 19.09,
+      "grad_norm": 3.1756997108459473,
+      "learning_rate": 2.5711600214609885e-07,
+      "loss": 1.0105,
+      "step": 63810
+    },
+    {
+      "epoch": 19.09,
+      "grad_norm": 4.981276512145996,
+      "learning_rate": 2.562762323421147e-07,
+      "loss": 1.0216,
+      "step": 63815
+    },
+    {
+      "epoch": 19.09,
+      "grad_norm": 3.7036938667297363,
+      "learning_rate": 2.554378291082765e-07,
+      "loss": 1.0003,
+      "step": 63820
+    },
+    {
+      "epoch": 19.1,
+      "grad_norm": 2.4953665733337402,
+      "learning_rate": 2.5460079249088606e-07,
+      "loss": 0.7268,
+      "step": 63825
+    },
+    {
+      "epoch": 19.1,
+      "grad_norm": 2.5604419708251953,
+      "learning_rate": 2.5376512253617034e-07,
+      "loss": 1.0928,
+      "step": 63830
+    },
+    {
+      "epoch": 19.1,
+      "grad_norm": 3.0103671550750732,
+      "learning_rate": 2.5293081929028686e-07,
+      "loss": 1.0442,
+      "step": 63835
+    },
+    {
+      "epoch": 19.1,
+      "grad_norm": 4.777401447296143,
+      "learning_rate": 2.5209788279930977e-07,
+      "loss": 0.9435,
+      "step": 63840
+    },
+    {
+      "epoch": 19.1,
+      "grad_norm": 2.0667479038238525,
+      "learning_rate": 2.512663131092441e-07,
+      "loss": 1.15,
+      "step": 63845
+    },
+    {
+      "epoch": 19.1,
+      "grad_norm": 1.9874635934829712,
+      "learning_rate": 2.504361102660141e-07,
+      "loss": 0.983,
+      "step": 63850
+    },
+    {
+      "epoch": 19.1,
+      "grad_norm": 4.316436290740967,
+      "learning_rate": 2.4960727431547206e-07,
+      "loss": 0.9662,
+      "step": 63855
+    },
+    {
+      "epoch": 19.11,
+      "grad_norm": 2.294708013534546,
+      "learning_rate": 2.487798053033924e-07,
+      "loss": 0.7309,
+      "step": 63860
+    },
+    {
+      "epoch": 19.11,
+      "grad_norm": 3.665914535522461,
+      "learning_rate": 2.479537032754803e-07,
+      "loss": 0.9579,
+      "step": 63865
+    },
+    {
+      "epoch": 19.11,
+      "grad_norm": 2.2551567554473877,
+      "learning_rate": 2.4712896827735197e-07,
+      "loss": 1.0002,
+      "step": 63870
+    },
+    {
+      "epoch": 19.11,
+      "grad_norm": 3.8116965293884277,
+      "learning_rate": 2.463056003545655e-07,
+      "loss": 0.8443,
+      "step": 63875
+    },
+    {
+      "epoch": 19.11,
+      "grad_norm": 2.817349433898926,
+      "learning_rate": 2.4548359955259003e-07,
+      "loss": 1.0352,
+      "step": 63880
+    },
+    {
+      "epoch": 19.11,
+      "grad_norm": 1.7628093957901,
+      "learning_rate": 2.4466296591682256e-07,
+      "loss": 0.9593,
+      "step": 63885
+    },
+    {
+      "epoch": 19.12,
+      "grad_norm": 4.2066755294799805,
+      "learning_rate": 2.43843699492588e-07,
+      "loss": 1.024,
+      "step": 63890
+    },
+    {
+      "epoch": 19.12,
+      "grad_norm": 2.6177399158477783,
+      "learning_rate": 2.430258003251334e-07,
+      "loss": 1.1829,
+      "step": 63895
+    },
+    {
+      "epoch": 19.12,
+      "grad_norm": 4.167474746704102,
+      "learning_rate": 2.422092684596311e-07,
+      "loss": 0.765,
+      "step": 63900
+    },
+    {
+      "epoch": 19.12,
+      "grad_norm": 2.174982786178589,
+      "learning_rate": 2.4139410394117825e-07,
+      "loss": 0.7503,
+      "step": 63905
+    },
+    {
+      "epoch": 19.12,
+      "grad_norm": 4.269941329956055,
+      "learning_rate": 2.405803068147916e-07,
+      "loss": 0.8784,
+      "step": 63910
+    },
+    {
+      "epoch": 19.12,
+      "grad_norm": 1.9396120309829712,
+      "learning_rate": 2.397678771254186e-07,
+      "loss": 1.0015,
+      "step": 63915
+    },
+    {
+      "epoch": 19.12,
+      "grad_norm": 5.77046537399292,
+      "learning_rate": 2.3895681491792886e-07,
+      "loss": 1.1133,
+      "step": 63920
+    },
+    {
+      "epoch": 19.13,
+      "grad_norm": 2.1925339698791504,
+      "learning_rate": 2.3814712023711995e-07,
+      "loss": 1.1291,
+      "step": 63925
+    },
+    {
+      "epoch": 19.13,
+      "grad_norm": 1.8961554765701294,
+      "learning_rate": 2.3733879312770324e-07,
+      "loss": 0.8816,
+      "step": 63930
+    },
+    {
+      "epoch": 19.13,
+      "grad_norm": 1.7458471059799194,
+      "learning_rate": 2.365318336343264e-07,
+      "loss": 1.0096,
+      "step": 63935
+    },
+    {
+      "epoch": 19.13,
+      "grad_norm": 1.646514892578125,
+      "learning_rate": 2.3572624180155655e-07,
+      "loss": 1.1617,
+      "step": 63940
+    },
+    {
+      "epoch": 19.13,
+      "grad_norm": 2.5368010997772217,
+      "learning_rate": 2.3492201767388589e-07,
+      "loss": 0.8202,
+      "step": 63945
+    },
+    {
+      "epoch": 19.13,
+      "grad_norm": 1.8663653135299683,
+      "learning_rate": 2.3411916129573164e-07,
+      "loss": 0.961,
+      "step": 63950
+    },
+    {
+      "epoch": 19.13,
+      "grad_norm": 3.437864065170288,
+      "learning_rate": 2.3331767271143057e-07,
+      "loss": 1.0659,
+      "step": 63955
+    },
+    {
+      "epoch": 19.14,
+      "grad_norm": 2.4234211444854736,
+      "learning_rate": 2.325175519652528e-07,
+      "loss": 0.881,
+      "step": 63960
+    },
+    {
+      "epoch": 19.14,
+      "grad_norm": 2.6147170066833496,
+      "learning_rate": 2.317187991013825e-07,
+      "loss": 0.8791,
+      "step": 63965
+    },
+    {
+      "epoch": 19.14,
+      "grad_norm": 3.2492587566375732,
+      "learning_rate": 2.3092141416394264e-07,
+      "loss": 0.9265,
+      "step": 63970
+    },
+    {
+      "epoch": 19.14,
+      "grad_norm": 2.781677722930908,
+      "learning_rate": 2.301253971969647e-07,
+      "loss": 0.8303,
+      "step": 63975
+    },
+    {
+      "epoch": 19.14,
+      "grad_norm": 2.981367826461792,
+      "learning_rate": 2.2933074824441624e-07,
+      "loss": 0.8374,
+      "step": 63980
+    },
+    {
+      "epoch": 19.14,
+      "grad_norm": 2.767667055130005,
+      "learning_rate": 2.2853746735017888e-07,
+      "loss": 1.0309,
+      "step": 63985
+    },
+    {
+      "epoch": 19.15,
+      "grad_norm": 2.9164721965789795,
+      "learning_rate": 2.2774555455807311e-07,
+      "loss": 0.9418,
+      "step": 63990
+    },
+    {
+      "epoch": 19.15,
+      "grad_norm": 2.0985426902770996,
+      "learning_rate": 2.2695500991182783e-07,
+      "loss": 1.0022,
+      "step": 63995
+    },
+    {
+      "epoch": 19.15,
+      "grad_norm": 2.751217842102051,
+      "learning_rate": 2.2616583345510812e-07,
+      "loss": 0.7166,
+      "step": 64000
+    },
+    {
+      "epoch": 19.15,
+      "grad_norm": 2.930095911026001,
+      "learning_rate": 2.2537802523149853e-07,
+      "loss": 0.921,
+      "step": 64005
+    },
+    {
+      "epoch": 19.15,
+      "grad_norm": 3.8547110557556152,
+      "learning_rate": 2.2459158528450874e-07,
+      "loss": 0.9999,
+      "step": 64010
+    },
+    {
+      "epoch": 19.15,
+      "grad_norm": 1.5911672115325928,
+      "learning_rate": 2.2380651365757343e-07,
+      "loss": 1.0612,
+      "step": 64015
+    },
+    {
+      "epoch": 19.15,
+      "grad_norm": 4.613253593444824,
+      "learning_rate": 2.2302281039405238e-07,
+      "loss": 1.0091,
+      "step": 64020
+    },
+    {
+      "epoch": 19.16,
+      "grad_norm": 2.566091299057007,
+      "learning_rate": 2.2224047553722484e-07,
+      "loss": 1.0461,
+      "step": 64025
+    },
+    {
+      "epoch": 19.16,
+      "grad_norm": 2.502011299133301,
+      "learning_rate": 2.2145950913030066e-07,
+      "loss": 1.0877,
+      "step": 64030
+    },
+    {
+      "epoch": 19.16,
+      "grad_norm": 3.6345396041870117,
+      "learning_rate": 2.2067991121641484e-07,
+      "loss": 0.9155,
+      "step": 64035
+    },
+    {
+      "epoch": 19.16,
+      "grad_norm": 1.3162267208099365,
+      "learning_rate": 2.1990168183861904e-07,
+      "loss": 0.8879,
+      "step": 64040
+    },
+    {
+      "epoch": 19.16,
+      "grad_norm": 2.207672595977783,
+      "learning_rate": 2.191248210398955e-07,
+      "loss": 0.9934,
+      "step": 64045
+    },
+    {
+      "epoch": 19.16,
+      "grad_norm": 3.507061719894409,
+      "learning_rate": 2.1834932886314885e-07,
+      "loss": 0.9936,
+      "step": 64050
+    },
+    {
+      "epoch": 19.16,
+      "grad_norm": 3.270716428756714,
+      "learning_rate": 2.1757520535121423e-07,
+      "loss": 0.9819,
+      "step": 64055
+    },
+    {
+      "epoch": 19.17,
+      "grad_norm": 2.573598623275757,
+      "learning_rate": 2.1680245054683524e-07,
+      "loss": 1.0214,
+      "step": 64060
+    },
+    {
+      "epoch": 19.17,
+      "grad_norm": 3.2568507194519043,
+      "learning_rate": 2.1603106449269993e-07,
+      "loss": 0.9246,
+      "step": 64065
+    },
+    {
+      "epoch": 19.17,
+      "grad_norm": 1.5595399141311646,
+      "learning_rate": 2.1526104723140484e-07,
+      "loss": 0.9696,
+      "step": 64070
+    },
+    {
+      "epoch": 19.17,
+      "grad_norm": 3.1589651107788086,
+      "learning_rate": 2.1449239880548254e-07,
+      "loss": 0.9985,
+      "step": 64075
+    },
+    {
+      "epoch": 19.17,
+      "grad_norm": 2.8172848224639893,
+      "learning_rate": 2.137251192573797e-07,
+      "loss": 0.7963,
+      "step": 64080
+    },
+    {
+      "epoch": 19.17,
+      "grad_norm": 2.691014528274536,
+      "learning_rate": 2.1295920862947628e-07,
+      "loss": 0.9184,
+      "step": 64085
+    },
+    {
+      "epoch": 19.17,
+      "grad_norm": 3.006046772003174,
+      "learning_rate": 2.1219466696407176e-07,
+      "loss": 0.8609,
+      "step": 64090
+    },
+    {
+      "epoch": 19.18,
+      "grad_norm": 2.1390583515167236,
+      "learning_rate": 2.1143149430338793e-07,
+      "loss": 1.1105,
+      "step": 64095
+    },
+    {
+      "epoch": 19.18,
+      "grad_norm": 1.9489213228225708,
+      "learning_rate": 2.106696906895772e-07,
+      "loss": 0.9752,
+      "step": 64100
+    },
+    {
+      "epoch": 19.18,
+      "grad_norm": 1.616949439048767,
+      "learning_rate": 2.0990925616471424e-07,
+      "loss": 0.9205,
+      "step": 64105
+    },
+    {
+      "epoch": 19.18,
+      "grad_norm": 1.416013240814209,
+      "learning_rate": 2.0915019077079322e-07,
+      "loss": 0.9529,
+      "step": 64110
+    },
+    {
+      "epoch": 19.18,
+      "grad_norm": 1.7403104305267334,
+      "learning_rate": 2.0839249454973896e-07,
+      "loss": 0.9774,
+      "step": 64115
+    },
+    {
+      "epoch": 19.18,
+      "grad_norm": 3.7459676265716553,
+      "learning_rate": 2.0763616754339577e-07,
+      "loss": 0.9201,
+      "step": 64120
+    },
+    {
+      "epoch": 19.19,
+      "grad_norm": 3.3942503929138184,
+      "learning_rate": 2.0688120979353853e-07,
+      "loss": 1.0433,
+      "step": 64125
+    },
+    {
+      "epoch": 19.19,
+      "grad_norm": 2.123073101043701,
+      "learning_rate": 2.061276213418617e-07,
+      "loss": 0.8314,
+      "step": 64130
+    },
+    {
+      "epoch": 19.19,
+      "grad_norm": 1.3835378885269165,
+      "learning_rate": 2.0537540222998474e-07,
+      "loss": 0.7707,
+      "step": 64135
+    },
+    {
+      "epoch": 19.19,
+      "grad_norm": 3.362372398376465,
+      "learning_rate": 2.0462455249945222e-07,
+      "loss": 0.9923,
+      "step": 64140
+    },
+    {
+      "epoch": 19.19,
+      "grad_norm": 3.691894769668579,
+      "learning_rate": 2.0387507219173098e-07,
+      "loss": 0.9268,
+      "step": 64145
+    },
+    {
+      "epoch": 19.19,
+      "grad_norm": 3.2239859104156494,
+      "learning_rate": 2.0312696134821562e-07,
+      "loss": 1.0184,
+      "step": 64150
+    },
+    {
+      "epoch": 19.19,
+      "grad_norm": 3.0610954761505127,
+      "learning_rate": 2.0238022001022315e-07,
+      "loss": 0.8632,
+      "step": 64155
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 1.395918846130371,
+      "learning_rate": 2.0163484821899557e-07,
+      "loss": 0.8581,
+      "step": 64160
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 4.282837390899658,
+      "learning_rate": 2.0089084601569718e-07,
+      "loss": 0.931,
+      "step": 64165
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 1.898026704788208,
+      "learning_rate": 2.0014821344142286e-07,
+      "loss": 0.866,
+      "step": 64170
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 2.9928948879241943,
+      "learning_rate": 1.9940695053718428e-07,
+      "loss": 1.0112,
+      "step": 64175
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 2.1379966735839844,
+      "learning_rate": 1.9866705734392088e-07,
+      "loss": 0.868,
+      "step": 64180
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 3.2108120918273926,
+      "learning_rate": 1.9792853390249444e-07,
+      "loss": 0.8312,
+      "step": 64185
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 6.297685146331787,
+      "learning_rate": 1.971913802536973e-07,
+      "loss": 0.9006,
+      "step": 64190
+    },
+    {
+      "epoch": 19.21,
+      "grad_norm": 3.2357826232910156,
+      "learning_rate": 1.9645559643823863e-07,
+      "loss": 0.9766,
+      "step": 64195
+    },
+    {
+      "epoch": 19.21,
+      "grad_norm": 2.2781829833984375,
+      "learning_rate": 1.9572118249675532e-07,
+      "loss": 1.1408,
+      "step": 64200
+    },
+    {
+      "epoch": 19.21,
+      "grad_norm": 3.3626670837402344,
+      "learning_rate": 1.9498813846980658e-07,
+      "loss": 0.9252,
+      "step": 64205
+    },
+    {
+      "epoch": 19.21,
+      "grad_norm": 6.3693108558654785,
+      "learning_rate": 1.9425646439788224e-07,
+      "loss": 1.0737,
+      "step": 64210
+    },
+    {
+      "epoch": 19.21,
+      "grad_norm": 3.2453861236572266,
+      "learning_rate": 1.9352616032138614e-07,
+      "loss": 0.8096,
+      "step": 64215
+    },
+    {
+      "epoch": 19.21,
+      "grad_norm": 2.069312572479248,
+      "learning_rate": 1.9279722628065823e-07,
+      "loss": 0.8934,
+      "step": 64220
+    },
+    {
+      "epoch": 19.22,
+      "grad_norm": 3.1376097202301025,
+      "learning_rate": 1.9206966231595236e-07,
+      "loss": 0.9529,
+      "step": 64225
+    },
+    {
+      "epoch": 19.22,
+      "grad_norm": 1.8707741498947144,
+      "learning_rate": 1.913434684674531e-07,
+      "loss": 1.1012,
+      "step": 64230
+    },
+    {
+      "epoch": 19.22,
+      "grad_norm": 4.028426170349121,
+      "learning_rate": 1.9061864477527004e-07,
+      "loss": 0.9865,
+      "step": 64235
+    },
+    {
+      "epoch": 19.22,
+      "grad_norm": 2.0778110027313232,
+      "learning_rate": 1.8989519127942667e-07,
+      "loss": 0.8687,
+      "step": 64240
+    },
+    {
+      "epoch": 19.22,
+      "grad_norm": 2.294110059738159,
+      "learning_rate": 1.8917310801988552e-07,
+      "loss": 1.0524,
+      "step": 64245
+    },
+    {
+      "epoch": 19.22,
+      "grad_norm": 2.6258926391601562,
+      "learning_rate": 1.88452395036523e-07,
+      "loss": 0.7745,
+      "step": 64250
+    },
+    {
+      "epoch": 19.22,
+      "grad_norm": 2.8713653087615967,
+      "learning_rate": 1.8773305236914618e-07,
+      "loss": 1.0878,
+      "step": 64255
+    },
+    {
+      "epoch": 19.23,
+      "grad_norm": 2.246710777282715,
+      "learning_rate": 1.8701508005747882e-07,
+      "loss": 1.0726,
+      "step": 64260
+    },
+    {
+      "epoch": 19.23,
+      "grad_norm": 3.4345037937164307,
+      "learning_rate": 1.8629847814118084e-07,
+      "loss": 0.826,
+      "step": 64265
+    },
+    {
+      "epoch": 19.23,
+      "grad_norm": 3.921339273452759,
+      "learning_rate": 1.8558324665982341e-07,
+      "loss": 0.9873,
+      "step": 64270
+    },
+    {
+      "epoch": 19.23,
+      "grad_norm": 3.5189566612243652,
+      "learning_rate": 1.8486938565290822e-07,
+      "loss": 1.0737,
+      "step": 64275
+    },
+    {
+      "epoch": 19.23,
+      "grad_norm": 3.8996822834014893,
+      "learning_rate": 1.8415689515986488e-07,
+      "loss": 0.9291,
+      "step": 64280
+    },
+    {
+      "epoch": 19.23,
+      "grad_norm": 1.6882737874984741,
+      "learning_rate": 1.8344577522004248e-07,
+      "loss": 0.8509,
+      "step": 64285
+    },
+    {
+      "epoch": 19.23,
+      "grad_norm": 1.2842025756835938,
+      "learning_rate": 1.827360258727151e-07,
+      "loss": 0.9767,
+      "step": 64290
+    },
+    {
+      "epoch": 19.24,
+      "grad_norm": 3.280601739883423,
+      "learning_rate": 1.820276471570792e-07,
+      "loss": 0.843,
+      "step": 64295
+    },
+    {
+      "epoch": 19.24,
+      "grad_norm": 1.3017467260360718,
+      "learning_rate": 1.8132063911225905e-07,
+      "loss": 1.0419,
+      "step": 64300
+    },
+    {
+      "epoch": 19.24,
+      "grad_norm": 2.4863226413726807,
+      "learning_rate": 1.8061500177730396e-07,
+      "loss": 0.8428,
+      "step": 64305
+    },
+    {
+      "epoch": 19.24,
+      "grad_norm": 3.148401975631714,
+      "learning_rate": 1.7991073519118274e-07,
+      "loss": 0.9154,
+      "step": 64310
+    },
+    {
+      "epoch": 19.24,
+      "grad_norm": 3.4491004943847656,
+      "learning_rate": 1.7920783939279207e-07,
+      "loss": 1.0461,
+      "step": 64315
+    },
+    {
+      "epoch": 19.24,
+      "grad_norm": 1.5797641277313232,
+      "learning_rate": 1.7850631442095367e-07,
+      "loss": 1.1385,
+      "step": 64320
+    },
+    {
+      "epoch": 19.25,
+      "grad_norm": 2.7437055110931396,
+      "learning_rate": 1.7780616031441156e-07,
+      "loss": 1.0647,
+      "step": 64325
+    },
+    {
+      "epoch": 19.25,
+      "grad_norm": 3.9407639503479004,
+      "learning_rate": 1.7710737711183478e-07,
+      "loss": 0.6554,
+      "step": 64330
+    },
+    {
+      "epoch": 19.25,
+      "grad_norm": 1.4198654890060425,
+      "learning_rate": 1.7640996485181472e-07,
+      "loss": 1.0993,
+      "step": 64335
+    },
+    {
+      "epoch": 19.25,
+      "grad_norm": 1.583427906036377,
+      "learning_rate": 1.7571392357287053e-07,
+      "loss": 1.1437,
+      "step": 64340
+    },
+    {
+      "epoch": 19.25,
+      "grad_norm": 6.862940311431885,
+      "learning_rate": 1.7501925331343817e-07,
+      "loss": 0.953,
+      "step": 64345
+    },
+    {
+      "epoch": 19.25,
+      "grad_norm": 4.871342182159424,
+      "learning_rate": 1.7432595411189524e-07,
+      "loss": 0.8867,
+      "step": 64350
+    },
+    {
+      "epoch": 19.25,
+      "grad_norm": 2.5354645252227783,
+      "learning_rate": 1.7363402600651945e-07,
+      "loss": 1.0432,
+      "step": 64355
+    },
+    {
+      "epoch": 19.26,
+      "grad_norm": 2.395785331726074,
+      "learning_rate": 1.7294346903553305e-07,
+      "loss": 0.8861,
+      "step": 64360
+    },
+    {
+      "epoch": 19.26,
+      "grad_norm": 2.0268208980560303,
+      "learning_rate": 1.7225428323707494e-07,
+      "loss": 0.8826,
+      "step": 64365
+    },
+    {
+      "epoch": 19.26,
+      "grad_norm": 1.8643012046813965,
+      "learning_rate": 1.7156646864920077e-07,
+      "loss": 1.0994,
+      "step": 64370
+    },
+    {
+      "epoch": 19.26,
+      "grad_norm": 1.6344417333602905,
+      "learning_rate": 1.708800253099052e-07,
+      "loss": 0.9345,
+      "step": 64375
+    },
+    {
+      "epoch": 19.26,
+      "grad_norm": 2.7925565242767334,
+      "learning_rate": 1.7019495325709677e-07,
+      "loss": 1.0531,
+      "step": 64380
+    },
+    {
+      "epoch": 19.26,
+      "grad_norm": 3.3944101333618164,
+      "learning_rate": 1.6951125252861466e-07,
+      "loss": 0.9223,
+      "step": 64385
+    },
+    {
+      "epoch": 19.26,
+      "grad_norm": 3.003763437271118,
+      "learning_rate": 1.68828923162212e-07,
+      "loss": 0.8319,
+      "step": 64390
+    },
+    {
+      "epoch": 19.27,
+      "grad_norm": 2.5167884826660156,
+      "learning_rate": 1.6814796519558084e-07,
+      "loss": 1.1937,
+      "step": 64395
+    },
+    {
+      "epoch": 19.27,
+      "grad_norm": 1.8801358938217163,
+      "learning_rate": 1.6746837866632725e-07,
+      "loss": 1.0366,
+      "step": 64400
+    },
+    {
+      "epoch": 19.27,
+      "grad_norm": 2.1538734436035156,
+      "learning_rate": 1.6679016361197951e-07,
+      "loss": 1.0713,
+      "step": 64405
+    },
+    {
+      "epoch": 19.27,
+      "grad_norm": 3.144021987915039,
+      "learning_rate": 1.6611332006999935e-07,
+      "loss": 0.9265,
+      "step": 64410
+    },
+    {
+      "epoch": 19.27,
+      "grad_norm": 2.700958490371704,
+      "learning_rate": 1.6543784807776795e-07,
+      "loss": 0.8548,
+      "step": 64415
+    },
+    {
+      "epoch": 19.27,
+      "grad_norm": 2.5718274116516113,
+      "learning_rate": 1.6476374767258883e-07,
+      "loss": 0.9646,
+      "step": 64420
+    },
+    {
+      "epoch": 19.28,
+      "grad_norm": 3.631052255630493,
+      "learning_rate": 1.640910188916961e-07,
+      "loss": 1.0448,
+      "step": 64425
+    },
+    {
+      "epoch": 19.28,
+      "grad_norm": 2.219627618789673,
+      "learning_rate": 1.6341966177223777e-07,
+      "loss": 1.1337,
+      "step": 64430
+    },
+    {
+      "epoch": 19.28,
+      "grad_norm": 1.4927635192871094,
+      "learning_rate": 1.6274967635129811e-07,
+      "loss": 0.9727,
+      "step": 64435
+    },
+    {
+      "epoch": 19.28,
+      "grad_norm": 3.119546890258789,
+      "learning_rate": 1.6208106266587253e-07,
+      "loss": 0.9235,
+      "step": 64440
+    },
+    {
+      "epoch": 19.28,
+      "grad_norm": 1.84986412525177,
+      "learning_rate": 1.6141382075289813e-07,
+      "loss": 0.8477,
+      "step": 64445
+    },
+    {
+      "epoch": 19.28,
+      "grad_norm": 2.341019868850708,
+      "learning_rate": 1.607479506492149e-07,
+      "loss": 1.1153,
+      "step": 64450
+    },
+    {
+      "epoch": 19.28,
+      "grad_norm": 2.070584535598755,
+      "learning_rate": 1.600834523916045e-07,
+      "loss": 0.8931,
+      "step": 64455
+    },
+    {
+      "epoch": 19.29,
+      "grad_norm": 4.45391845703125,
+      "learning_rate": 1.5942032601676815e-07,
+      "loss": 0.9916,
+      "step": 64460
+    },
+    {
+      "epoch": 19.29,
+      "grad_norm": 3.5736074447631836,
+      "learning_rate": 1.587585715613238e-07,
+      "loss": 1.1195,
+      "step": 64465
+    },
+    {
+      "epoch": 19.29,
+      "grad_norm": 2.6194748878479004,
+      "learning_rate": 1.580981890618255e-07,
+      "loss": 1.0239,
+      "step": 64470
+    },
+    {
+      "epoch": 19.29,
+      "grad_norm": 2.58370304107666,
+      "learning_rate": 1.5743917855473855e-07,
+      "loss": 1.0816,
+      "step": 64475
+    },
+    {
+      "epoch": 19.29,
+      "grad_norm": 2.7932465076446533,
+      "learning_rate": 1.5678154007646718e-07,
+      "loss": 1.0808,
+      "step": 64480
+    },
+    {
+      "epoch": 19.29,
+      "grad_norm": 5.6625471115112305,
+      "learning_rate": 1.5612527366332674e-07,
+      "loss": 0.9235,
+      "step": 64485
+    },
+    {
+      "epoch": 19.29,
+      "grad_norm": 3.3300294876098633,
+      "learning_rate": 1.5547037935156327e-07,
+      "loss": 0.8833,
+      "step": 64490
+    },
+    {
+      "epoch": 19.3,
+      "grad_norm": 2.539818525314331,
+      "learning_rate": 1.5481685717734783e-07,
+      "loss": 0.9428,
+      "step": 64495
+    },
+    {
+      "epoch": 19.3,
+      "grad_norm": 3.0235023498535156,
+      "learning_rate": 1.54164707176771e-07,
+      "loss": 1.1307,
+      "step": 64500
+    },
+    {
+      "epoch": 19.3,
+      "grad_norm": 2.7196106910705566,
+      "learning_rate": 1.5351392938585118e-07,
+      "loss": 0.946,
+      "step": 64505
+    },
+    {
+      "epoch": 19.3,
+      "grad_norm": 2.6924359798431396,
+      "learning_rate": 1.5286452384053184e-07,
+      "loss": 0.7011,
+      "step": 64510
+    },
+    {
+      "epoch": 19.3,
+      "grad_norm": 1.7030616998672485,
+      "learning_rate": 1.52216490576676e-07,
+      "loss": 1.0243,
+      "step": 64515
+    },
+    {
+      "epoch": 19.3,
+      "grad_norm": 2.473074197769165,
+      "learning_rate": 1.5156982963007715e-07,
+      "loss": 1.0579,
+      "step": 64520
+    },
+    {
+      "epoch": 19.31,
+      "grad_norm": 1.5015933513641357,
+      "learning_rate": 1.5092454103644572e-07,
+      "loss": 0.9168,
+      "step": 64525
+    },
+    {
+      "epoch": 19.31,
+      "grad_norm": 2.3585140705108643,
+      "learning_rate": 1.5028062483142537e-07,
+      "loss": 0.8898,
+      "step": 64530
+    },
+    {
+      "epoch": 19.31,
+      "grad_norm": 2.2519643306732178,
+      "learning_rate": 1.4963808105057653e-07,
+      "loss": 0.9104,
+      "step": 64535
+    },
+    {
+      "epoch": 19.31,
+      "grad_norm": 3.287801504135132,
+      "learning_rate": 1.4899690972938474e-07,
+      "loss": 1.0621,
+      "step": 64540
+    },
+    {
+      "epoch": 19.31,
+      "grad_norm": 2.565260648727417,
+      "learning_rate": 1.4835711090326054e-07,
+      "loss": 0.8879,
+      "step": 64545
+    },
+    {
+      "epoch": 19.31,
+      "grad_norm": 1.9417716264724731,
+      "learning_rate": 1.4771868460754234e-07,
+      "loss": 0.9862,
+      "step": 64550
+    },
+    {
+      "epoch": 19.31,
+      "grad_norm": 1.4697206020355225,
+      "learning_rate": 1.4708163087749082e-07,
+      "loss": 0.9105,
+      "step": 64555
+    },
+    {
+      "epoch": 19.32,
+      "grad_norm": 6.190129280090332,
+      "learning_rate": 1.4644594974828618e-07,
+      "loss": 0.9255,
+      "step": 64560
+    },
+    {
+      "epoch": 19.32,
+      "grad_norm": 2.575528383255005,
+      "learning_rate": 1.4581164125503922e-07,
+      "loss": 1.0289,
+      "step": 64565
+    },
+    {
+      "epoch": 19.32,
+      "grad_norm": 2.064091205596924,
+      "learning_rate": 1.4517870543277745e-07,
+      "loss": 1.0171,
+      "step": 64570
+    },
+    {
+      "epoch": 19.32,
+      "grad_norm": 1.4593415260314941,
+      "learning_rate": 1.445471423164646e-07,
+      "loss": 0.8089,
+      "step": 64575
+    },
+    {
+      "epoch": 19.32,
+      "grad_norm": 4.081628799438477,
+      "learning_rate": 1.4391695194097553e-07,
+      "loss": 0.9634,
+      "step": 64580
+    },
+    {
+      "epoch": 19.32,
+      "grad_norm": 1.9155333042144775,
+      "learning_rate": 1.4328813434111577e-07,
+      "loss": 0.8419,
+      "step": 64585
+    },
+    {
+      "epoch": 19.32,
+      "grad_norm": 1.9294395446777344,
+      "learning_rate": 1.426606895516186e-07,
+      "loss": 0.8925,
+      "step": 64590
+    },
+    {
+      "epoch": 19.33,
+      "grad_norm": 4.205466270446777,
+      "learning_rate": 1.4203461760713133e-07,
+      "loss": 0.9417,
+      "step": 64595
+    },
+    {
+      "epoch": 19.33,
+      "grad_norm": 2.388813018798828,
+      "learning_rate": 1.4140991854223183e-07,
+      "loss": 1.0479,
+      "step": 64600
+    },
+    {
+      "epoch": 19.33,
+      "grad_norm": 2.137241840362549,
+      "learning_rate": 1.407865923914259e-07,
+      "loss": 1.0903,
+      "step": 64605
+    },
+    {
+      "epoch": 19.33,
+      "grad_norm": 3.590604543685913,
+      "learning_rate": 1.4016463918913592e-07,
+      "loss": 0.8402,
+      "step": 64610
+    },
+    {
+      "epoch": 19.33,
+      "grad_norm": 3.1465625762939453,
+      "learning_rate": 1.3954405896971224e-07,
+      "loss": 1.1066,
+      "step": 64615
+    },
+    {
+      "epoch": 19.33,
+      "grad_norm": 4.3162922859191895,
+      "learning_rate": 1.389248517674302e-07,
+      "loss": 1.0114,
+      "step": 64620
+    },
+    {
+      "epoch": 19.34,
+      "grad_norm": 2.18872332572937,
+      "learning_rate": 1.3830701761648744e-07,
+      "loss": 0.954,
+      "step": 64625
+    },
+    {
+      "epoch": 19.34,
+      "grad_norm": 3.4795877933502197,
+      "learning_rate": 1.3769055655100395e-07,
+      "loss": 0.7991,
+      "step": 64630
+    },
+    {
+      "epoch": 19.34,
+      "grad_norm": 3.6978468894958496,
+      "learning_rate": 1.3707546860503019e-07,
+      "loss": 0.9708,
+      "step": 64635
+    },
+    {
+      "epoch": 19.34,
+      "grad_norm": 3.991750478744507,
+      "learning_rate": 1.3646175381253345e-07,
+      "loss": 0.9624,
+      "step": 64640
+    },
+    {
+      "epoch": 19.34,
+      "grad_norm": 3.2379486560821533,
+      "learning_rate": 1.3584941220740888e-07,
+      "loss": 1.0254,
+      "step": 64645
+    },
+    {
+      "epoch": 19.34,
+      "grad_norm": 3.404052257537842,
+      "learning_rate": 1.352384438234794e-07,
+      "loss": 0.8464,
+      "step": 64650
+    },
+    {
+      "epoch": 19.34,
+      "grad_norm": 2.1188278198242188,
+      "learning_rate": 1.346288486944819e-07,
+      "loss": 0.8731,
+      "step": 64655
+    },
+    {
+      "epoch": 19.35,
+      "grad_norm": 5.067503452301025,
+      "learning_rate": 1.3402062685408945e-07,
+      "loss": 0.8417,
+      "step": 64660
+    },
+    {
+      "epoch": 19.35,
+      "grad_norm": 3.299839973449707,
+      "learning_rate": 1.3341377833588908e-07,
+      "loss": 1.0166,
+      "step": 64665
+    },
+    {
+      "epoch": 19.35,
+      "grad_norm": 3.843200922012329,
+      "learning_rate": 1.3280830317340122e-07,
+      "loss": 0.9164,
+      "step": 64670
+    },
+    {
+      "epoch": 19.35,
+      "grad_norm": 3.3282971382141113,
+      "learning_rate": 1.3220420140006017e-07,
+      "loss": 1.0903,
+      "step": 64675
+    },
+    {
+      "epoch": 19.35,
+      "grad_norm": 3.1337714195251465,
+      "learning_rate": 1.3160147304923375e-07,
+      "loss": 0.953,
+      "step": 64680
+    },
+    {
+      "epoch": 19.35,
+      "grad_norm": 4.363522529602051,
+      "learning_rate": 1.3100011815420642e-07,
+      "loss": 0.7601,
+      "step": 64685
+    },
+    {
+      "epoch": 19.35,
+      "grad_norm": 3.619309902191162,
+      "learning_rate": 1.3040013674819605e-07,
+      "loss": 0.8317,
+      "step": 64690
+    },
+    {
+      "epoch": 19.36,
+      "grad_norm": 4.793463230133057,
+      "learning_rate": 1.298015288643345e-07,
+      "loss": 1.1246,
+      "step": 64695
+    },
+    {
+      "epoch": 19.36,
+      "grad_norm": 4.220412731170654,
+      "learning_rate": 1.292042945356814e-07,
+      "loss": 1.2013,
+      "step": 64700
+    },
+    {
+      "epoch": 19.36,
+      "grad_norm": 1.8279774188995361,
+      "learning_rate": 1.2860843379522425e-07,
+      "loss": 0.9528,
+      "step": 64705
+    },
+    {
+      "epoch": 19.36,
+      "grad_norm": 3.403494119644165,
+      "learning_rate": 1.2801394667587285e-07,
+      "loss": 0.9246,
+      "step": 64710
+    },
+    {
+      "epoch": 19.36,
+      "grad_norm": 3.680722713470459,
+      "learning_rate": 1.274208332104565e-07,
+      "loss": 0.9582,
+      "step": 64715
+    },
+    {
+      "epoch": 19.36,
+      "grad_norm": 1.425324559211731,
+      "learning_rate": 1.268290934317351e-07,
+      "loss": 1.0058,
+      "step": 64720
+    },
+    {
+      "epoch": 19.36,
+      "grad_norm": 2.55713152885437,
+      "learning_rate": 1.262387273723853e-07,
+      "loss": 1.0049,
+      "step": 64725
+    },
+    {
+      "epoch": 19.37,
+      "grad_norm": 3.1014976501464844,
+      "learning_rate": 1.2564973506501987e-07,
+      "loss": 0.9631,
+      "step": 64730
+    },
+    {
+      "epoch": 19.37,
+      "grad_norm": 2.44319224357605,
+      "learning_rate": 1.2506211654216282e-07,
+      "loss": 0.9343,
+      "step": 64735
+    },
+    {
+      "epoch": 19.37,
+      "grad_norm": 1.407696008682251,
+      "learning_rate": 1.2447587183626874e-07,
+      "loss": 0.9746,
+      "step": 64740
+    },
+    {
+      "epoch": 19.37,
+      "grad_norm": 3.8315956592559814,
+      "learning_rate": 1.238910009797145e-07,
+      "loss": 0.9999,
+      "step": 64745
+    },
+    {
+      "epoch": 19.37,
+      "grad_norm": 2.9506800174713135,
+      "learning_rate": 1.2330750400480483e-07,
+      "loss": 0.9509,
+      "step": 64750
+    },
+    {
+      "epoch": 19.37,
+      "grad_norm": 1.5093969106674194,
+      "learning_rate": 1.2272538094376396e-07,
+      "loss": 1.0732,
+      "step": 64755
+    },
+    {
+      "epoch": 19.38,
+      "grad_norm": 3.4581491947174072,
+      "learning_rate": 1.2214463182873836e-07,
+      "loss": 0.9692,
+      "step": 64760
+    },
+    {
+      "epoch": 19.38,
+      "grad_norm": 6.015048027038574,
+      "learning_rate": 1.2156525669181074e-07,
+      "loss": 0.8817,
+      "step": 64765
+    },
+    {
+      "epoch": 19.38,
+      "grad_norm": 1.234391450881958,
+      "learning_rate": 1.2098725556496938e-07,
+      "loss": 0.9878,
+      "step": 64770
+    },
+    {
+      "epoch": 19.38,
+      "grad_norm": 2.1239829063415527,
+      "learning_rate": 1.2041062848014428e-07,
+      "loss": 1.0328,
+      "step": 64775
+    },
+    {
+      "epoch": 19.38,
+      "grad_norm": 2.3981587886810303,
+      "learning_rate": 1.1983537546917944e-07,
+      "loss": 0.9892,
+      "step": 64780
+    },
+    {
+      "epoch": 19.38,
+      "grad_norm": 4.929365158081055,
+      "learning_rate": 1.1926149656384667e-07,
+      "loss": 1.0319,
+      "step": 64785
+    },
+    {
+      "epoch": 19.38,
+      "grad_norm": 2.424349308013916,
+      "learning_rate": 1.1868899179583726e-07,
+      "loss": 1.0278,
+      "step": 64790
+    },
+    {
+      "epoch": 19.39,
+      "grad_norm": 3.70564866065979,
+      "learning_rate": 1.1811786119677037e-07,
+      "loss": 0.8346,
+      "step": 64795
+    },
+    {
+      "epoch": 19.39,
+      "grad_norm": 1.7846856117248535,
+      "learning_rate": 1.1754810479819301e-07,
+      "loss": 0.8833,
+      "step": 64800
+    },
+    {
+      "epoch": 19.39,
+      "grad_norm": 2.44340181350708,
+      "learning_rate": 1.1697972263157164e-07,
+      "loss": 0.951,
+      "step": 64805
+    },
+    {
+      "epoch": 19.39,
+      "grad_norm": 8.20602798461914,
+      "learning_rate": 1.1641271472829229e-07,
+      "loss": 0.9087,
+      "step": 64810
+    },
+    {
+      "epoch": 19.39,
+      "grad_norm": 5.193169593811035,
+      "learning_rate": 1.1584708111967435e-07,
+      "loss": 1.0248,
+      "step": 64815
+    },
+    {
+      "epoch": 19.39,
+      "grad_norm": 5.007645130157471,
+      "learning_rate": 1.1528282183695949e-07,
+      "loss": 0.9085,
+      "step": 64820
+    },
+    {
+      "epoch": 19.39,
+      "grad_norm": 2.5000011920928955,
+      "learning_rate": 1.1471993691130612e-07,
+      "loss": 1.1319,
+      "step": 64825
+    },
+    {
+      "epoch": 19.4,
+      "grad_norm": 1.7281321287155151,
+      "learning_rate": 1.1415842637380326e-07,
+      "loss": 0.9048,
+      "step": 64830
+    },
+    {
+      "epoch": 19.4,
+      "grad_norm": 3.866112470626831,
+      "learning_rate": 1.135982902554622e-07,
+      "loss": 0.9548,
+      "step": 64835
+    },
+    {
+      "epoch": 19.4,
+      "grad_norm": 3.7432994842529297,
+      "learning_rate": 1.1303952858722211e-07,
+      "loss": 0.977,
+      "step": 64840
+    },
+    {
+      "epoch": 19.4,
+      "grad_norm": 2.450726270675659,
+      "learning_rate": 1.1248214139993884e-07,
+      "loss": 1.0289,
+      "step": 64845
+    },
+    {
+      "epoch": 19.4,
+      "grad_norm": 4.173949718475342,
+      "learning_rate": 1.1192612872439889e-07,
+      "loss": 1.0554,
+      "step": 64850
+    },
+    {
+      "epoch": 19.4,
+      "grad_norm": 2.5136771202087402,
+      "learning_rate": 1.1137149059130825e-07,
+      "loss": 0.8978,
+      "step": 64855
+    },
+    {
+      "epoch": 19.41,
+      "grad_norm": 3.791132688522339,
+      "learning_rate": 1.1081822703130074e-07,
+      "loss": 1.076,
+      "step": 64860
+    },
+    {
+      "epoch": 19.41,
+      "grad_norm": 2.5353622436523438,
+      "learning_rate": 1.102663380749297e-07,
+      "loss": 1.0079,
+      "step": 64865
+    },
+    {
+      "epoch": 19.41,
+      "grad_norm": 1.9366538524627686,
+      "learning_rate": 1.0971582375267908e-07,
+      "loss": 1.0107,
+      "step": 64870
+    },
+    {
+      "epoch": 19.41,
+      "grad_norm": 2.1806907653808594,
+      "learning_rate": 1.0916668409494957e-07,
+      "loss": 0.959,
+      "step": 64875
+    },
+    {
+      "epoch": 19.41,
+      "grad_norm": 1.3208969831466675,
+      "learning_rate": 1.0861891913207523e-07,
+      "loss": 0.7762,
+      "step": 64880
+    },
+    {
+      "epoch": 19.41,
+      "grad_norm": 1.4078150987625122,
+      "learning_rate": 1.080725288943013e-07,
+      "loss": 0.9522,
+      "step": 64885
+    },
+    {
+      "epoch": 19.41,
+      "grad_norm": 3.181238889694214,
+      "learning_rate": 1.0752751341180922e-07,
+      "loss": 0.9171,
+      "step": 64890
+    },
+    {
+      "epoch": 19.42,
+      "grad_norm": 2.8408191204071045,
+      "learning_rate": 1.0698387271469712e-07,
+      "loss": 0.9459,
+      "step": 64895
+    },
+    {
+      "epoch": 19.42,
+      "grad_norm": 2.00119686126709,
+      "learning_rate": 1.0644160683299375e-07,
+      "loss": 1.0576,
+      "step": 64900
+    },
+    {
+      "epoch": 19.42,
+      "grad_norm": 1.823187232017517,
+      "learning_rate": 1.0590071579664185e-07,
+      "loss": 1.192,
+      "step": 64905
+    },
+    {
+      "epoch": 19.42,
+      "grad_norm": 2.4590156078338623,
+      "learning_rate": 1.0536119963552026e-07,
+      "loss": 0.9914,
+      "step": 64910
+    },
+    {
+      "epoch": 19.42,
+      "grad_norm": 3.0153346061706543,
+      "learning_rate": 1.0482305837942185e-07,
+      "loss": 0.9732,
+      "step": 64915
+    },
+    {
+      "epoch": 19.42,
+      "grad_norm": 1.55227530002594,
+      "learning_rate": 1.0428629205806728e-07,
+      "loss": 0.9852,
+      "step": 64920
+    },
+    {
+      "epoch": 19.42,
+      "grad_norm": 1.4424388408660889,
+      "learning_rate": 1.0375090070110505e-07,
+      "loss": 1.0208,
+      "step": 64925
+    },
+    {
+      "epoch": 19.43,
+      "grad_norm": 1.9990373849868774,
+      "learning_rate": 1.032168843381004e-07,
+      "loss": 0.9683,
+      "step": 64930
+    },
+    {
+      "epoch": 19.43,
+      "grad_norm": 4.354613304138184,
+      "learning_rate": 1.0268424299855195e-07,
+      "loss": 0.9116,
+      "step": 64935
+    },
+    {
+      "epoch": 19.43,
+      "grad_norm": 4.074637413024902,
+      "learning_rate": 1.0215297671186952e-07,
+      "loss": 0.846,
+      "step": 64940
+    },
+    {
+      "epoch": 19.43,
+      "grad_norm": 2.377812385559082,
+      "learning_rate": 1.0162308550740185e-07,
+      "loss": 1.0472,
+      "step": 64945
+    },
+    {
+      "epoch": 19.43,
+      "grad_norm": 1.2441447973251343,
+      "learning_rate": 1.0109456941440886e-07,
+      "loss": 1.0943,
+      "step": 64950
+    },
+    {
+      "epoch": 19.43,
+      "grad_norm": 35.32939910888672,
+      "learning_rate": 1.0056742846208389e-07,
+      "loss": 0.8993,
+      "step": 64955
+    },
+    {
+      "epoch": 19.44,
+      "grad_norm": 4.968936443328857,
+      "learning_rate": 1.0004166267953419e-07,
+      "loss": 0.976,
+      "step": 64960
+    },
+    {
+      "epoch": 19.44,
+      "grad_norm": 6.838953971862793,
+      "learning_rate": 9.951727209580598e-08,
+      "loss": 0.9147,
+      "step": 64965
+    },
+    {
+      "epoch": 19.44,
+      "grad_norm": 1.3050497770309448,
+      "learning_rate": 9.899425673985107e-08,
+      "loss": 1.0613,
+      "step": 64970
+    },
+    {
+      "epoch": 19.44,
+      "grad_norm": 3.5567336082458496,
+      "learning_rate": 9.847261664056584e-08,
+      "loss": 0.9914,
+      "step": 64975
+    },
+    {
+      "epoch": 19.44,
+      "grad_norm": 3.1688625812530518,
+      "learning_rate": 9.795235182674944e-08,
+      "loss": 1.0523,
+      "step": 64980
+    },
+    {
+      "epoch": 19.44,
+      "grad_norm": 1.9151177406311035,
+      "learning_rate": 9.743346232714279e-08,
+      "loss": 1.0164,
+      "step": 64985
+    },
+    {
+      "epoch": 19.44,
+      "grad_norm": 3.5026047229766846,
+      "learning_rate": 9.691594817040073e-08,
+      "loss": 0.9442,
+      "step": 64990
+    },
+    {
+      "epoch": 19.45,
+      "grad_norm": 2.427326202392578,
+      "learning_rate": 9.639980938510595e-08,
+      "loss": 0.9183,
+      "step": 64995
+    },
+    {
+      "epoch": 19.45,
+      "grad_norm": 1.6219308376312256,
+      "learning_rate": 9.588504599976345e-08,
+      "loss": 1.0548,
+      "step": 65000
+    },
+    {
+      "epoch": 19.45,
+      "grad_norm": 2.265070676803589,
+      "learning_rate": 9.537165804280324e-08,
+      "loss": 0.9385,
+      "step": 65005
+    },
+    {
+      "epoch": 19.45,
+      "grad_norm": 4.394448280334473,
+      "learning_rate": 9.485964554258043e-08,
+      "loss": 0.9525,
+      "step": 65010
+    },
+    {
+      "epoch": 19.45,
+      "grad_norm": 3.270338296890259,
+      "learning_rate": 9.434900852737238e-08,
+      "loss": 0.9407,
+      "step": 65015
+    },
+    {
+      "epoch": 19.45,
+      "grad_norm": 2.3003993034362793,
+      "learning_rate": 9.383974702537878e-08,
+      "loss": 0.8104,
+      "step": 65020
+    },
+    {
+      "epoch": 19.45,
+      "grad_norm": 2.146404981613159,
+      "learning_rate": 9.33318610647299e-08,
+      "loss": 0.9752,
+      "step": 65025
+    },
+    {
+      "epoch": 19.46,
+      "grad_norm": 1.9496128559112549,
+      "learning_rate": 9.282535067346998e-08,
+      "loss": 0.8462,
+      "step": 65030
+    },
+    {
+      "epoch": 19.46,
+      "grad_norm": 1.7979656457901,
+      "learning_rate": 9.23202158795794e-08,
+      "loss": 0.8072,
+      "step": 65035
+    },
+    {
+      "epoch": 19.46,
+      "grad_norm": 2.433094024658203,
+      "learning_rate": 9.181645671095252e-08,
+      "loss": 0.8688,
+      "step": 65040
+    },
+    {
+      "epoch": 19.46,
+      "grad_norm": 3.4137656688690186,
+      "learning_rate": 9.131407319541153e-08,
+      "loss": 1.0364,
+      "step": 65045
+    },
+    {
+      "epoch": 19.46,
+      "grad_norm": 1.888330340385437,
+      "learning_rate": 9.081306536070366e-08,
+      "loss": 0.8318,
+      "step": 65050
+    },
+    {
+      "epoch": 19.46,
+      "grad_norm": 3.29191255569458,
+      "learning_rate": 9.03134332344957e-08,
+      "loss": 0.7606,
+      "step": 65055
+    },
+    {
+      "epoch": 19.47,
+      "grad_norm": 1.3745579719543457,
+      "learning_rate": 8.981517684438779e-08,
+      "loss": 0.9285,
+      "step": 65060
+    },
+    {
+      "epoch": 19.47,
+      "grad_norm": 2.466064929962158,
+      "learning_rate": 8.9318296217894e-08,
+      "loss": 0.9789,
+      "step": 65065
+    },
+    {
+      "epoch": 19.47,
+      "grad_norm": 2.370903730392456,
+      "learning_rate": 8.882279138245908e-08,
+      "loss": 0.8135,
+      "step": 65070
+    },
+    {
+      "epoch": 19.47,
+      "grad_norm": 3.7410812377929688,
+      "learning_rate": 8.832866236544446e-08,
+      "loss": 0.9207,
+      "step": 65075
+    },
+    {
+      "epoch": 19.47,
+      "grad_norm": 2.961038589477539,
+      "learning_rate": 8.783590919414497e-08,
+      "loss": 1.0046,
+      "step": 65080
+    },
+    {
+      "epoch": 19.47,
+      "grad_norm": 2.3268706798553467,
+      "learning_rate": 8.734453189577497e-08,
+      "loss": 0.9985,
+      "step": 65085
+    },
+    {
+      "epoch": 19.47,
+      "grad_norm": 2.1825544834136963,
+      "learning_rate": 8.685453049747105e-08,
+      "loss": 0.9957,
+      "step": 65090
+    },
+    {
+      "epoch": 19.48,
+      "grad_norm": 3.99306321144104,
+      "learning_rate": 8.636590502629494e-08,
+      "loss": 0.9198,
+      "step": 65095
+    },
+    {
+      "epoch": 19.48,
+      "grad_norm": 4.728055000305176,
+      "learning_rate": 8.587865550923057e-08,
+      "loss": 1.1148,
+      "step": 65100
+    },
+    {
+      "epoch": 19.48,
+      "grad_norm": 1.5131349563598633,
+      "learning_rate": 8.539278197319533e-08,
+      "loss": 0.8645,
+      "step": 65105
+    },
+    {
+      "epoch": 19.48,
+      "grad_norm": 4.779293537139893,
+      "learning_rate": 8.490828444501775e-08,
+      "loss": 0.9724,
+      "step": 65110
+    },
+    {
+      "epoch": 19.48,
+      "grad_norm": 2.5396671295166016,
+      "learning_rate": 8.442516295145974e-08,
+      "loss": 0.9741,
+      "step": 65115
+    },
+    {
+      "epoch": 19.48,
+      "grad_norm": 2.4737155437469482,
+      "learning_rate": 8.394341751919998e-08,
+      "loss": 1.1152,
+      "step": 65120
+    },
+    {
+      "epoch": 19.48,
+      "grad_norm": 3.827990770339966,
+      "learning_rate": 8.346304817484496e-08,
+      "loss": 0.9559,
+      "step": 65125
+    },
+    {
+      "epoch": 19.49,
+      "grad_norm": 2.1786043643951416,
+      "learning_rate": 8.307974350048597e-08,
+      "loss": 0.8171,
+      "step": 65130
+    },
+    {
+      "epoch": 19.49,
+      "grad_norm": 2.547623634338379,
+      "learning_rate": 8.260185118116881e-08,
+      "loss": 0.9985,
+      "step": 65135
+    },
+    {
+      "epoch": 19.49,
+      "grad_norm": 2.0980887413024902,
+      "learning_rate": 8.212533502385267e-08,
+      "loss": 0.9029,
+      "step": 65140
+    },
+    {
+      "epoch": 19.49,
+      "grad_norm": 3.6388769149780273,
+      "learning_rate": 8.165019505485261e-08,
+      "loss": 1.0007,
+      "step": 65145
+    },
+    {
+      "epoch": 19.49,
+      "grad_norm": 3.2435462474823,
+      "learning_rate": 8.117643130041152e-08,
+      "loss": 1.0546,
+      "step": 65150
+    },
+    {
+      "epoch": 19.49,
+      "grad_norm": 2.3990752696990967,
+      "learning_rate": 8.070404378669461e-08,
+      "loss": 0.9811,
+      "step": 65155
+    },
+    {
+      "epoch": 19.5,
+      "grad_norm": 2.219163179397583,
+      "learning_rate": 8.02330325397893e-08,
+      "loss": 1.179,
+      "step": 65160
+    },
+    {
+      "epoch": 19.5,
+      "grad_norm": 1.6309661865234375,
+      "learning_rate": 7.976339758571094e-08,
+      "loss": 0.8849,
+      "step": 65165
+    },
+    {
+      "epoch": 19.5,
+      "grad_norm": 1.6608452796936035,
+      "learning_rate": 7.929513895039987e-08,
+      "loss": 0.8809,
+      "step": 65170
+    },
+    {
+      "epoch": 19.5,
+      "grad_norm": 3.3001632690429688,
+      "learning_rate": 7.882825665971316e-08,
+      "loss": 0.9709,
+      "step": 65175
+    },
+    {
+      "epoch": 19.5,
+      "grad_norm": 3.3454103469848633,
+      "learning_rate": 7.836275073943577e-08,
+      "loss": 0.9296,
+      "step": 65180
+    },
+    {
+      "epoch": 19.5,
+      "grad_norm": 3.6163206100463867,
+      "learning_rate": 7.789862121528325e-08,
+      "loss": 1.0389,
+      "step": 65185
+    },
+    {
+      "epoch": 19.5,
+      "grad_norm": 2.1792218685150146,
+      "learning_rate": 7.743586811288228e-08,
+      "loss": 0.8764,
+      "step": 65190
+    },
+    {
+      "epoch": 19.51,
+      "grad_norm": 36.940330505371094,
+      "learning_rate": 7.6974491457793e-08,
+      "loss": 1.1318,
+      "step": 65195
+    },
+    {
+      "epoch": 19.51,
+      "grad_norm": 1.098024845123291,
+      "learning_rate": 7.651449127549503e-08,
+      "loss": 0.9647,
+      "step": 65200
+    },
+    {
+      "epoch": 19.51,
+      "grad_norm": 2.7333242893218994,
+      "learning_rate": 7.605586759139582e-08,
+      "loss": 1.0234,
+      "step": 65205
+    },
+    {
+      "epoch": 19.51,
+      "grad_norm": 3.17170786857605,
+      "learning_rate": 7.559862043082511e-08,
+      "loss": 1.0791,
+      "step": 65210
+    },
+    {
+      "epoch": 19.51,
+      "grad_norm": 3.0869386196136475,
+      "learning_rate": 7.514274981903491e-08,
+      "loss": 0.8448,
+      "step": 65215
+    },
+    {
+      "epoch": 19.51,
+      "grad_norm": 1.4254027605056763,
+      "learning_rate": 7.468825578120508e-08,
+      "loss": 1.0125,
+      "step": 65220
+    },
+    {
+      "epoch": 19.51,
+      "grad_norm": 4.486727237701416,
+      "learning_rate": 7.423513834242945e-08,
+      "loss": 0.7673,
+      "step": 65225
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 2.695744514465332,
+      "learning_rate": 7.378339752774077e-08,
+      "loss": 0.9713,
+      "step": 65230
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 1.8701138496398926,
+      "learning_rate": 7.333303336208574e-08,
+      "loss": 1.0323,
+      "step": 65235
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 2.85798978805542,
+      "learning_rate": 7.288404587033892e-08,
+      "loss": 0.9666,
+      "step": 65240
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 2.4904799461364746,
+      "learning_rate": 7.243643507729437e-08,
+      "loss": 1.0193,
+      "step": 65245
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 1.5817186832427979,
+      "learning_rate": 7.199020100767395e-08,
+      "loss": 0.9865,
+      "step": 65250
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 4.257795333862305,
+      "learning_rate": 7.154534368612465e-08,
+      "loss": 0.7875,
+      "step": 65255
+    },
+    {
+      "epoch": 19.53,
+      "grad_norm": 4.011660099029541,
+      "learning_rate": 7.11018631372129e-08,
+      "loss": 0.87,
+      "step": 65260
+    },
+    {
+      "epoch": 19.53,
+      "grad_norm": 4.863183975219727,
+      "learning_rate": 7.065975938543579e-08,
+      "loss": 0.9798,
+      "step": 65265
+    },
+    {
+      "epoch": 19.53,
+      "grad_norm": 2.2304656505584717,
+      "learning_rate": 7.021903245520433e-08,
+      "loss": 0.9956,
+      "step": 65270
+    },
+    {
+      "epoch": 19.53,
+      "grad_norm": 2.4296457767486572,
+      "learning_rate": 6.977968237086574e-08,
+      "loss": 0.9869,
+      "step": 65275
+    },
+    {
+      "epoch": 19.53,
+      "grad_norm": 1.2431954145431519,
+      "learning_rate": 6.934170915668114e-08,
+      "loss": 1.1323,
+      "step": 65280
+    },
+    {
+      "epoch": 19.53,
+      "grad_norm": 2.272568464279175,
+      "learning_rate": 6.890511283683953e-08,
+      "loss": 1.0307,
+      "step": 65285
+    },
+    {
+      "epoch": 19.53,
+      "grad_norm": 3.0335166454315186,
+      "learning_rate": 6.846989343545219e-08,
+      "loss": 1.0689,
+      "step": 65290
+    },
+    {
+      "epoch": 19.54,
+      "grad_norm": 2.1746132373809814,
+      "learning_rate": 6.803605097656096e-08,
+      "loss": 1.0758,
+      "step": 65295
+    },
+    {
+      "epoch": 19.54,
+      "grad_norm": 1.310981035232544,
+      "learning_rate": 6.760358548411893e-08,
+      "loss": 1.0002,
+      "step": 65300
+    },
+    {
+      "epoch": 19.54,
+      "grad_norm": 3.3575997352600098,
+      "learning_rate": 6.717249698202088e-08,
+      "loss": 0.8939,
+      "step": 65305
+    },
+    {
+      "epoch": 19.54,
+      "grad_norm": 1.3703774213790894,
+      "learning_rate": 6.674278549406444e-08,
+      "loss": 0.9257,
+      "step": 65310
+    },
+    {
+      "epoch": 19.54,
+      "grad_norm": 1.6967878341674805,
+      "learning_rate": 6.63144510439917e-08,
+      "loss": 1.0022,
+      "step": 65315
+    },
+    {
+      "epoch": 19.54,
+      "grad_norm": 3.0534799098968506,
+      "learning_rate": 6.588749365545044e-08,
+      "loss": 1.0344,
+      "step": 65320
+    },
+    {
+      "epoch": 19.54,
+      "grad_norm": 2.0059475898742676,
+      "learning_rate": 6.546191335202734e-08,
+      "loss": 0.8697,
+      "step": 65325
+    },
+    {
+      "epoch": 19.55,
+      "grad_norm": 2.44685697555542,
+      "learning_rate": 6.503771015722582e-08,
+      "loss": 0.9656,
+      "step": 65330
+    },
+    {
+      "epoch": 19.55,
+      "grad_norm": 2.028308868408203,
+      "learning_rate": 6.461488409447159e-08,
+      "loss": 0.7818,
+      "step": 65335
+    },
+    {
+      "epoch": 19.55,
+      "grad_norm": 1.6491647958755493,
+      "learning_rate": 6.419343518711818e-08,
+      "loss": 0.822,
+      "step": 65340
+    },
+    {
+      "epoch": 19.55,
+      "grad_norm": 2.1462082862854004,
+      "learning_rate": 6.37733634584442e-08,
+      "loss": 1.0347,
+      "step": 65345
+    },
+    {
+      "epoch": 19.55,
+      "grad_norm": 2.555720329284668,
+      "learning_rate": 6.335466893164499e-08,
+      "loss": 0.9114,
+      "step": 65350
+    },
+    {
+      "epoch": 19.55,
+      "grad_norm": 1.5168747901916504,
+      "learning_rate": 6.293735162984926e-08,
+      "loss": 0.8913,
+      "step": 65355
+    },
+    {
+      "epoch": 19.55,
+      "grad_norm": 3.1510086059570312,
+      "learning_rate": 6.252141157610247e-08,
+      "loss": 1.0805,
+      "step": 65360
+    },
+    {
+      "epoch": 19.56,
+      "grad_norm": 1.3030726909637451,
+      "learning_rate": 6.210684879337514e-08,
+      "loss": 0.9486,
+      "step": 65365
+    },
+    {
+      "epoch": 19.56,
+      "grad_norm": 1.6003532409667969,
+      "learning_rate": 6.16936633045656e-08,
+      "loss": 1.0441,
+      "step": 65370
+    },
+    {
+      "epoch": 19.56,
+      "grad_norm": 4.755357265472412,
+      "learning_rate": 6.128185513249452e-08,
+      "loss": 0.9853,
+      "step": 65375
+    },
+    {
+      "epoch": 19.56,
+      "grad_norm": 2.193000316619873,
+      "learning_rate": 6.087142429990478e-08,
+      "loss": 1.1333,
+      "step": 65380
+    },
+    {
+      "epoch": 19.56,
+      "grad_norm": 2.2681541442871094,
+      "learning_rate": 6.04623708294616e-08,
+      "loss": 1.1406,
+      "step": 65385
+    },
+    {
+      "epoch": 19.56,
+      "grad_norm": 3.417823314666748,
+      "learning_rate": 6.00546947437608e-08,
+      "loss": 1.0069,
+      "step": 65390
+    },
+    {
+      "epoch": 19.57,
+      "grad_norm": 2.2752387523651123,
+      "learning_rate": 5.964839606531214e-08,
+      "loss": 1.0824,
+      "step": 65395
+    },
+    {
+      "epoch": 19.57,
+      "grad_norm": 8.419718742370605,
+      "learning_rate": 5.924347481656156e-08,
+      "loss": 0.9543,
+      "step": 65400
+    },
+    {
+      "epoch": 19.57,
+      "grad_norm": 6.15936279296875,
+      "learning_rate": 5.8839931019868955e-08,
+      "loss": 0.9819,
+      "step": 65405
+    },
+    {
+      "epoch": 19.57,
+      "grad_norm": 3.1836822032928467,
+      "learning_rate": 5.8437764697522044e-08,
+      "loss": 1.0345,
+      "step": 65410
+    },
+    {
+      "epoch": 19.57,
+      "grad_norm": 2.2897796630859375,
+      "learning_rate": 5.803697587173085e-08,
+      "loss": 0.9194,
+      "step": 65415
+    },
+    {
+      "epoch": 19.57,
+      "grad_norm": 2.0201714038848877,
+      "learning_rate": 5.7637564564633205e-08,
+      "loss": 1.1888,
+      "step": 65420
+    },
+    {
+      "epoch": 19.57,
+      "grad_norm": 3.7068679332733154,
+      "learning_rate": 5.7239530798283704e-08,
+      "loss": 0.9841,
+      "step": 65425
+    },
+    {
+      "epoch": 19.58,
+      "grad_norm": 1.6118470430374146,
+      "learning_rate": 5.684287459467308e-08,
+      "loss": 0.9835,
+      "step": 65430
+    },
+    {
+      "epoch": 19.58,
+      "grad_norm": 2.5250370502471924,
+      "learning_rate": 5.6447595975700485e-08,
+      "loss": 0.8138,
+      "step": 65435
+    },
+    {
+      "epoch": 19.58,
+      "grad_norm": 2.585517406463623,
+      "learning_rate": 5.6053694963198456e-08,
+      "loss": 0.9517,
+      "step": 65440
+    },
+    {
+      "epoch": 19.58,
+      "grad_norm": 1.3797327280044556,
+      "learning_rate": 5.566117157892459e-08,
+      "loss": 1.0316,
+      "step": 65445
+    },
+    {
+      "epoch": 19.58,
+      "grad_norm": 2.215745449066162,
+      "learning_rate": 5.527002584455876e-08,
+      "loss": 0.7985,
+      "step": 65450
+    },
+    {
+      "epoch": 19.58,
+      "grad_norm": 4.586230754852295,
+      "learning_rate": 5.488025778169759e-08,
+      "loss": 0.8436,
+      "step": 65455
+    },
+    {
+      "epoch": 19.58,
+      "grad_norm": 3.00388503074646,
+      "learning_rate": 5.4491867411871064e-08,
+      "loss": 1.0267,
+      "step": 65460
+    },
+    {
+      "epoch": 19.59,
+      "grad_norm": 2.888880729675293,
+      "learning_rate": 5.410485475652871e-08,
+      "loss": 0.9659,
+      "step": 65465
+    },
+    {
+      "epoch": 19.59,
+      "grad_norm": 1.5345345735549927,
+      "learning_rate": 5.3719219837047865e-08,
+      "loss": 1.161,
+      "step": 65470
+    },
+    {
+      "epoch": 19.59,
+      "grad_norm": 3.530027151107788,
+      "learning_rate": 5.33349626747226e-08,
+      "loss": 0.9582,
+      "step": 65475
+    },
+    {
+      "epoch": 19.59,
+      "grad_norm": 1.8898435831069946,
+      "learning_rate": 5.295208329077761e-08,
+      "loss": 1.1305,
+      "step": 65480
+    },
+    {
+      "epoch": 19.59,
+      "grad_norm": 1.663489580154419,
+      "learning_rate": 5.257058170635709e-08,
+      "loss": 0.8846,
+      "step": 65485
+    },
+    {
+      "epoch": 19.59,
+      "grad_norm": 2.8383398056030273,
+      "learning_rate": 5.2190457942533076e-08,
+      "loss": 0.8722,
+      "step": 65490
+    },
+    {
+      "epoch": 19.6,
+      "grad_norm": 2.9115426540374756,
+      "learning_rate": 5.181171202029711e-08,
+      "loss": 0.824,
+      "step": 65495
+    },
+    {
+      "epoch": 19.6,
+      "grad_norm": 7.482906341552734,
+      "learning_rate": 5.1434343960568565e-08,
+      "loss": 1.0373,
+      "step": 65500
+    },
+    {
+      "epoch": 19.6,
+      "grad_norm": 3.7145473957061768,
+      "learning_rate": 5.105835378418911e-08,
+      "loss": 0.8548,
+      "step": 65505
+    },
+    {
+      "epoch": 19.6,
+      "grad_norm": 2.6154332160949707,
+      "learning_rate": 5.068374151192268e-08,
+      "loss": 1.0704,
+      "step": 65510
+    },
+    {
+      "epoch": 19.6,
+      "grad_norm": 8.783279418945312,
+      "learning_rate": 5.031050716446106e-08,
+      "loss": 0.8265,
+      "step": 65515
+    },
+    {
+      "epoch": 19.6,
+      "grad_norm": 2.184469699859619,
+      "learning_rate": 4.993865076241555e-08,
+      "loss": 1.0476,
+      "step": 65520
+    },
+    {
+      "epoch": 19.6,
+      "grad_norm": 2.483076810836792,
+      "learning_rate": 4.9568172326325266e-08,
+      "loss": 0.8417,
+      "step": 65525
+    },
+    {
+      "epoch": 19.61,
+      "grad_norm": 2.2957425117492676,
+      "learning_rate": 4.919907187664885e-08,
+      "loss": 0.9391,
+      "step": 65530
+    },
+    {
+      "epoch": 19.61,
+      "grad_norm": 2.0007917881011963,
+      "learning_rate": 4.883134943377276e-08,
+      "loss": 0.8605,
+      "step": 65535
+    },
+    {
+      "epoch": 19.61,
+      "grad_norm": 2.37748384475708,
+      "learning_rate": 4.846500501800577e-08,
+      "loss": 1.2328,
+      "step": 65540
+    },
+    {
+      "epoch": 19.61,
+      "grad_norm": 8.686102867126465,
+      "learning_rate": 4.810003864958168e-08,
+      "loss": 0.8948,
+      "step": 65545
+    },
+    {
+      "epoch": 19.61,
+      "grad_norm": 2.3307371139526367,
+      "learning_rate": 4.773645034865659e-08,
+      "loss": 1.0579,
+      "step": 65550
+    },
+    {
+      "epoch": 19.61,
+      "grad_norm": 4.153103351593018,
+      "learning_rate": 4.737424013531166e-08,
+      "loss": 0.7621,
+      "step": 65555
+    },
+    {
+      "epoch": 19.61,
+      "grad_norm": 4.211757659912109,
+      "learning_rate": 4.701340802955034e-08,
+      "loss": 1.0984,
+      "step": 65560
+    },
+    {
+      "epoch": 19.62,
+      "grad_norm": 4.062090873718262,
+      "learning_rate": 4.6653954051298354e-08,
+      "loss": 0.9231,
+      "step": 65565
+    },
+    {
+      "epoch": 19.62,
+      "grad_norm": 1.6518398523330688,
+      "learning_rate": 4.6295878220414815e-08,
+      "loss": 0.8956,
+      "step": 65570
+    },
+    {
+      "epoch": 19.62,
+      "grad_norm": 2.662663698196411,
+      "learning_rate": 4.5939180556670016e-08,
+      "loss": 0.9944,
+      "step": 65575
+    },
+    {
+      "epoch": 19.62,
+      "grad_norm": 3.212043523788452,
+      "learning_rate": 4.5583861079767645e-08,
+      "loss": 0.8612,
+      "step": 65580
+    },
+    {
+      "epoch": 19.62,
+      "grad_norm": 2.235893964767456,
+      "learning_rate": 4.5229919809328115e-08,
+      "loss": 0.9491,
+      "step": 65585
+    },
+    {
+      "epoch": 19.62,
+      "grad_norm": 2.6770949363708496,
+      "learning_rate": 4.4877356764902455e-08,
+      "loss": 0.8584,
+      "step": 65590
+    },
+    {
+      "epoch": 19.63,
+      "grad_norm": 6.8069939613342285,
+      "learning_rate": 4.452617196595843e-08,
+      "loss": 0.895,
+      "step": 65595
+    },
+    {
+      "epoch": 19.63,
+      "grad_norm": 2.035522937774658,
+      "learning_rate": 4.417636543189718e-08,
+      "loss": 0.8317,
+      "step": 65600
+    },
+    {
+      "epoch": 19.63,
+      "grad_norm": 2.070621967315674,
+      "learning_rate": 4.3827937182033816e-08,
+      "loss": 0.9145,
+      "step": 65605
+    },
+    {
+      "epoch": 19.63,
+      "grad_norm": 1.5345319509506226,
+      "learning_rate": 4.348088723561128e-08,
+      "loss": 0.9873,
+      "step": 65610
+    },
+    {
+      "epoch": 19.63,
+      "grad_norm": 1.464187741279602,
+      "learning_rate": 4.313521561180034e-08,
+      "loss": 0.9745,
+      "step": 65615
+    },
+    {
+      "epoch": 19.63,
+      "grad_norm": 5.471027851104736,
+      "learning_rate": 4.2790922329691287e-08,
+      "loss": 0.9186,
+      "step": 65620
+    },
+    {
+      "epoch": 19.63,
+      "grad_norm": 1.9127711057662964,
+      "learning_rate": 4.2448007408296684e-08,
+      "loss": 0.9168,
+      "step": 65625
+    },
+    {
+      "epoch": 19.64,
+      "grad_norm": 2.524143934249878,
+      "learning_rate": 4.210647086655695e-08,
+      "loss": 1.0501,
+      "step": 65630
+    },
+    {
+      "epoch": 19.64,
+      "grad_norm": 3.043079137802124,
+      "learning_rate": 4.1766312723334755e-08,
+      "loss": 0.7822,
+      "step": 65635
+    },
+    {
+      "epoch": 19.64,
+      "grad_norm": 2.911064624786377,
+      "learning_rate": 4.1427532997415086e-08,
+      "loss": 1.012,
+      "step": 65640
+    },
+    {
+      "epoch": 19.64,
+      "grad_norm": 5.481101036071777,
+      "learning_rate": 4.109013170751075e-08,
+      "loss": 1.0,
+      "step": 65645
+    },
+    {
+      "epoch": 19.64,
+      "grad_norm": 2.473296880722046,
+      "learning_rate": 4.075410887225684e-08,
+      "loss": 0.8792,
+      "step": 65650
+    },
+    {
+      "epoch": 19.64,
+      "grad_norm": 2.742133855819702,
+      "learning_rate": 4.041946451020795e-08,
+      "loss": 0.9363,
+      "step": 65655
+    },
+    {
+      "epoch": 19.64,
+      "grad_norm": 1.7099344730377197,
+      "learning_rate": 4.008619863984653e-08,
+      "loss": 1.0857,
+      "step": 65660
+    },
+    {
+      "epoch": 19.65,
+      "grad_norm": 4.01974630355835,
+      "learning_rate": 3.975431127958285e-08,
+      "loss": 1.0116,
+      "step": 65665
+    },
+    {
+      "epoch": 19.65,
+      "grad_norm": 6.58184814453125,
+      "learning_rate": 3.9423802447743905e-08,
+      "loss": 0.7649,
+      "step": 65670
+    },
+    {
+      "epoch": 19.65,
+      "grad_norm": 3.7717185020446777,
+      "learning_rate": 3.909467216258178e-08,
+      "loss": 0.8852,
+      "step": 65675
+    },
+    {
+      "epoch": 19.65,
+      "grad_norm": 1.591399908065796,
+      "learning_rate": 3.876692044227359e-08,
+      "loss": 0.932,
+      "step": 65680
+    },
+    {
+      "epoch": 19.65,
+      "grad_norm": 1.8852194547653198,
+      "learning_rate": 3.8440547304927074e-08,
+      "loss": 0.9933,
+      "step": 65685
+    },
+    {
+      "epoch": 19.65,
+      "grad_norm": 5.670814037322998,
+      "learning_rate": 3.811555276855838e-08,
+      "loss": 1.0301,
+      "step": 65690
+    },
+    {
+      "epoch": 19.66,
+      "grad_norm": 1.9637709856033325,
+      "learning_rate": 3.7791936851125365e-08,
+      "loss": 0.9422,
+      "step": 65695
+    },
+    {
+      "epoch": 19.66,
+      "grad_norm": 2.995957612991333,
+      "learning_rate": 3.746969957049429e-08,
+      "loss": 0.8587,
+      "step": 65700
+    },
+    {
+      "epoch": 19.66,
+      "grad_norm": 2.3271963596343994,
+      "learning_rate": 3.7148840944464804e-08,
+      "loss": 0.8903,
+      "step": 65705
+    },
+    {
+      "epoch": 19.66,
+      "grad_norm": 2.014490842819214,
+      "learning_rate": 3.682936099075884e-08,
+      "loss": 1.0424,
+      "step": 65710
+    },
+    {
+      "epoch": 19.66,
+      "grad_norm": 3.5324313640594482,
+      "learning_rate": 3.6511259727017856e-08,
+      "loss": 0.7962,
+      "step": 65715
+    },
+    {
+      "epoch": 19.66,
+      "grad_norm": 2.035303831100464,
+      "learning_rate": 3.619453717081389e-08,
+      "loss": 0.9449,
+      "step": 65720
+    },
+    {
+      "epoch": 19.66,
+      "grad_norm": 12.83745288848877,
+      "learning_rate": 3.587919333963574e-08,
+      "loss": 1.0862,
+      "step": 65725
+    },
+    {
+      "epoch": 19.67,
+      "grad_norm": 1.7152767181396484,
+      "learning_rate": 3.556522825090281e-08,
+      "loss": 0.8594,
+      "step": 65730
+    },
+    {
+      "epoch": 19.67,
+      "grad_norm": 2.850773334503174,
+      "learning_rate": 3.525264192195121e-08,
+      "loss": 1.0139,
+      "step": 65735
+    },
+    {
+      "epoch": 19.67,
+      "grad_norm": 4.40149450302124,
+      "learning_rate": 3.494143437004771e-08,
+      "loss": 0.8235,
+      "step": 65740
+    },
+    {
+      "epoch": 19.67,
+      "grad_norm": 2.7569260597229004,
+      "learning_rate": 3.463160561237855e-08,
+      "loss": 0.8739,
+      "step": 65745
+    },
+    {
+      "epoch": 19.67,
+      "grad_norm": 2.746654510498047,
+      "learning_rate": 3.432315566605504e-08,
+      "loss": 0.9751,
+      "step": 65750
+    },
+    {
+      "epoch": 19.67,
+      "grad_norm": 3.1638407707214355,
+      "learning_rate": 3.4016084548116336e-08,
+      "loss": 0.9332,
+      "step": 65755
+    },
+    {
+      "epoch": 19.67,
+      "grad_norm": 2.6680147647857666,
+      "learning_rate": 3.371039227551553e-08,
+      "loss": 0.8459,
+      "step": 65760
+    },
+    {
+      "epoch": 19.68,
+      "grad_norm": 3.2998719215393066,
+      "learning_rate": 3.3406078865139137e-08,
+      "loss": 0.9761,
+      "step": 65765
+    },
+    {
+      "epoch": 19.68,
+      "grad_norm": 1.581969141960144,
+      "learning_rate": 3.3103144333793134e-08,
+      "loss": 1.1189,
+      "step": 65770
+    },
+    {
+      "epoch": 19.68,
+      "grad_norm": 5.630129337310791,
+      "learning_rate": 3.280158869821137e-08,
+      "loss": 0.8332,
+      "step": 65775
+    },
+    {
+      "epoch": 19.68,
+      "grad_norm": 2.6767990589141846,
+      "learning_rate": 3.250141197504442e-08,
+      "loss": 1.0059,
+      "step": 65780
+    },
+    {
+      "epoch": 19.68,
+      "grad_norm": 3.5217807292938232,
+      "learning_rate": 3.2202614180870674e-08,
+      "loss": 0.9764,
+      "step": 65785
+    },
+    {
+      "epoch": 19.68,
+      "grad_norm": 2.578888177871704,
+      "learning_rate": 3.190519533219638e-08,
+      "loss": 0.7636,
+      "step": 65790
+    },
+    {
+      "epoch": 19.69,
+      "grad_norm": 3.0376648902893066,
+      "learning_rate": 3.160915544544452e-08,
+      "loss": 0.9932,
+      "step": 65795
+    },
+    {
+      "epoch": 19.69,
+      "grad_norm": 4.299635887145996,
+      "learning_rate": 3.1314494536965886e-08,
+      "loss": 0.6652,
+      "step": 65800
+    },
+    {
+      "epoch": 19.69,
+      "grad_norm": 2.366783857345581,
+      "learning_rate": 3.1021212623033594e-08,
+      "loss": 1.089,
+      "step": 65805
+    },
+    {
+      "epoch": 19.69,
+      "grad_norm": 5.171496391296387,
+      "learning_rate": 3.0729309719845775e-08,
+      "loss": 0.9039,
+      "step": 65810
+    },
+    {
+      "epoch": 19.69,
+      "grad_norm": 3.2826364040374756,
+      "learning_rate": 3.043878584352566e-08,
+      "loss": 0.9896,
+      "step": 65815
+    },
+    {
+      "epoch": 19.69,
+      "grad_norm": 2.9492897987365723,
+      "learning_rate": 3.014964101011597e-08,
+      "loss": 1.0328,
+      "step": 65820
+    },
+    {
+      "epoch": 19.69,
+      "grad_norm": 3.130568265914917,
+      "learning_rate": 2.9861875235587255e-08,
+      "loss": 1.2715,
+      "step": 65825
+    },
+    {
+      "epoch": 19.7,
+      "grad_norm": 2.2713990211486816,
+      "learning_rate": 2.9575488535829587e-08,
+      "loss": 1.0102,
+      "step": 65830
+    },
+    {
+      "epoch": 19.7,
+      "grad_norm": 2.3380908966064453,
+      "learning_rate": 2.929048092666642e-08,
+      "loss": 0.9455,
+      "step": 65835
+    },
+    {
+      "epoch": 19.7,
+      "grad_norm": 4.085785865783691,
+      "learning_rate": 2.9006852423832386e-08,
+      "loss": 0.8658,
+      "step": 65840
+    },
+    {
+      "epoch": 19.7,
+      "grad_norm": 3.3306400775909424,
+      "learning_rate": 2.8724603042992735e-08,
+      "loss": 0.8673,
+      "step": 65845
+    },
+    {
+      "epoch": 19.7,
+      "grad_norm": 3.320685863494873,
+      "learning_rate": 2.8443732799740554e-08,
+      "loss": 1.0188,
+      "step": 65850
+    },
+    {
+      "epoch": 19.7,
+      "grad_norm": 3.164665460586548,
+      "learning_rate": 2.8164241709582873e-08,
+      "loss": 1.046,
+      "step": 65855
+    },
+    {
+      "epoch": 19.7,
+      "grad_norm": 3.0726613998413086,
+      "learning_rate": 2.7886129787954573e-08,
+      "loss": 0.8932,
+      "step": 65860
+    },
+    {
+      "epoch": 19.71,
+      "grad_norm": 2.211113452911377,
+      "learning_rate": 2.7609397050221143e-08,
+      "loss": 0.8676,
+      "step": 65865
+    },
+    {
+      "epoch": 19.71,
+      "grad_norm": 3.0560343265533447,
+      "learning_rate": 2.7334043511662023e-08,
+      "loss": 0.9694,
+      "step": 65870
+    },
+    {
+      "epoch": 19.71,
+      "grad_norm": 1.7159016132354736,
+      "learning_rate": 2.706006918748727e-08,
+      "loss": 0.8506,
+      "step": 65875
+    },
+    {
+      "epoch": 19.71,
+      "grad_norm": 2.9933714866638184,
+      "learning_rate": 2.678747409282645e-08,
+      "loss": 0.9774,
+      "step": 65880
+    },
+    {
+      "epoch": 19.71,
+      "grad_norm": 2.8719935417175293,
+      "learning_rate": 2.6516258242736958e-08,
+      "loss": 0.8631,
+      "step": 65885
+    },
+    {
+      "epoch": 19.71,
+      "grad_norm": 1.9159245491027832,
+      "learning_rate": 2.624642165219293e-08,
+      "loss": 0.7401,
+      "step": 65890
+    },
+    {
+      "epoch": 19.72,
+      "grad_norm": 5.0512590408325195,
+      "learning_rate": 2.597796433610189e-08,
+      "loss": 0.8452,
+      "step": 65895
+    },
+    {
+      "epoch": 19.72,
+      "grad_norm": 2.150611639022827,
+      "learning_rate": 2.571088630929086e-08,
+      "loss": 0.9875,
+      "step": 65900
+    },
+    {
+      "epoch": 19.72,
+      "grad_norm": 5.610554218292236,
+      "learning_rate": 2.5445187586503606e-08,
+      "loss": 0.9885,
+      "step": 65905
+    },
+    {
+      "epoch": 19.72,
+      "grad_norm": 2.4917097091674805,
+      "learning_rate": 2.5180868182422822e-08,
+      "loss": 1.0434,
+      "step": 65910
+    },
+    {
+      "epoch": 19.72,
+      "grad_norm": 2.9542877674102783,
+      "learning_rate": 2.4917928111639623e-08,
+      "loss": 1.0089,
+      "step": 65915
+    },
+    {
+      "epoch": 19.72,
+      "grad_norm": 2.203429937362671,
+      "learning_rate": 2.4656367388681268e-08,
+      "loss": 0.981,
+      "step": 65920
+    },
+    {
+      "epoch": 19.72,
+      "grad_norm": 3.659770965576172,
+      "learning_rate": 2.4396186027991764e-08,
+      "loss": 1.1026,
+      "step": 65925
+    },
+    {
+      "epoch": 19.73,
+      "grad_norm": 2.265361785888672,
+      "learning_rate": 2.413738404394017e-08,
+      "loss": 0.9166,
+      "step": 65930
+    },
+    {
+      "epoch": 19.73,
+      "grad_norm": 2.5812058448791504,
+      "learning_rate": 2.3879961450817835e-08,
+      "loss": 1.1483,
+      "step": 65935
+    },
+    {
+      "epoch": 19.73,
+      "grad_norm": 2.302971363067627,
+      "learning_rate": 2.3623918262846712e-08,
+      "loss": 0.8114,
+      "step": 65940
+    },
+    {
+      "epoch": 19.73,
+      "grad_norm": 2.793581008911133,
+      "learning_rate": 2.3369254494162718e-08,
+      "loss": 1.0621,
+      "step": 65945
+    },
+    {
+      "epoch": 19.73,
+      "grad_norm": 5.097170829772949,
+      "learning_rate": 2.3115970158835155e-08,
+      "loss": 0.9018,
+      "step": 65950
+    },
+    {
+      "epoch": 19.73,
+      "grad_norm": 3.7390031814575195,
+      "learning_rate": 2.2864065270850054e-08,
+      "loss": 1.0263,
+      "step": 65955
+    },
+    {
+      "epoch": 19.73,
+      "grad_norm": 2.8096506595611572,
+      "learning_rate": 2.2613539844118514e-08,
+      "loss": 1.0402,
+      "step": 65960
+    },
+    {
+      "epoch": 19.74,
+      "grad_norm": 1.8856858015060425,
+      "learning_rate": 2.236439389247946e-08,
+      "loss": 0.9717,
+      "step": 65965
+    },
+    {
+      "epoch": 19.74,
+      "grad_norm": 3.062253713607788,
+      "learning_rate": 2.2116627429694116e-08,
+      "loss": 0.8636,
+      "step": 65970
+    },
+    {
+      "epoch": 19.74,
+      "grad_norm": 2.116095781326294,
+      "learning_rate": 2.1870240469440417e-08,
+      "loss": 0.8674,
+      "step": 65975
+    },
+    {
+      "epoch": 19.74,
+      "grad_norm": 4.265397071838379,
+      "learning_rate": 2.162523302533248e-08,
+      "loss": 0.9501,
+      "step": 65980
+    },
+    {
+      "epoch": 19.74,
+      "grad_norm": 3.8397390842437744,
+      "learning_rate": 2.138160511090115e-08,
+      "loss": 0.9542,
+      "step": 65985
+    },
+    {
+      "epoch": 19.74,
+      "grad_norm": 5.092801570892334,
+      "learning_rate": 2.1139356739596773e-08,
+      "loss": 0.9933,
+      "step": 65990
+    },
+    {
+      "epoch": 19.74,
+      "grad_norm": 3.441525459289551,
+      "learning_rate": 2.0898487924803088e-08,
+      "loss": 0.8262,
+      "step": 65995
+    },
+    {
+      "epoch": 19.75,
+      "grad_norm": 2.899848461151123,
+      "learning_rate": 2.0658998679820573e-08,
+      "loss": 1.0899,
+      "step": 66000
+    },
+    {
+      "epoch": 19.75,
+      "grad_norm": 2.0572032928466797,
+      "learning_rate": 2.0420889017880305e-08,
+      "loss": 0.7595,
+      "step": 66005
+    },
+    {
+      "epoch": 19.75,
+      "grad_norm": 4.190646171569824,
+      "learning_rate": 2.0184158952124555e-08,
+      "loss": 0.8722,
+      "step": 66010
+    },
+    {
+      "epoch": 19.75,
+      "grad_norm": 2.0695345401763916,
+      "learning_rate": 1.9948808495637293e-08,
+      "loss": 1.0758,
+      "step": 66015
+    },
+    {
+      "epoch": 19.75,
+      "grad_norm": 3.038285732269287,
+      "learning_rate": 1.9714837661408138e-08,
+      "loss": 0.9091,
+      "step": 66020
+    },
+    {
+      "epoch": 19.75,
+      "grad_norm": 1.420379400253296,
+      "learning_rate": 1.9482246462365628e-08,
+      "loss": 0.9363,
+      "step": 66025
+    },
+    {
+      "epoch": 19.76,
+      "grad_norm": 2.4541139602661133,
+      "learning_rate": 1.9251034911352274e-08,
+      "loss": 0.9865,
+      "step": 66030
+    },
+    {
+      "epoch": 19.76,
+      "grad_norm": 1.3549059629440308,
+      "learning_rate": 1.9021203021135637e-08,
+      "loss": 1.0879,
+      "step": 66035
+    },
+    {
+      "epoch": 19.76,
+      "grad_norm": 3.1724984645843506,
+      "learning_rate": 1.8792750804413894e-08,
+      "loss": 0.9949,
+      "step": 66040
+    },
+    {
+      "epoch": 19.76,
+      "grad_norm": 2.643465995788574,
+      "learning_rate": 1.8565678273801956e-08,
+      "loss": 1.0872,
+      "step": 66045
+    },
+    {
+      "epoch": 19.76,
+      "grad_norm": 2.5308868885040283,
+      "learning_rate": 1.8339985441842567e-08,
+      "loss": 0.9338,
+      "step": 66050
+    },
+    {
+      "epoch": 19.76,
+      "grad_norm": 3.499067783355713,
+      "learning_rate": 1.8115672320995202e-08,
+      "loss": 1.0337,
+      "step": 66055
+    },
+    {
+      "epoch": 19.76,
+      "grad_norm": 3.2034714221954346,
+      "learning_rate": 1.7892738923655504e-08,
+      "loss": 0.859,
+      "step": 66060
+    },
+    {
+      "epoch": 19.77,
+      "grad_norm": 4.103240489959717,
+      "learning_rate": 1.7671185262130296e-08,
+      "loss": 0.8122,
+      "step": 66065
+    },
+    {
+      "epoch": 19.77,
+      "grad_norm": 2.9544129371643066,
+      "learning_rate": 1.745101134865701e-08,
+      "loss": 0.9386,
+      "step": 66070
+    },
+    {
+      "epoch": 19.77,
+      "grad_norm": 1.3058322668075562,
+      "learning_rate": 1.7232217195398138e-08,
+      "loss": 0.9956,
+      "step": 66075
+    },
+    {
+      "epoch": 19.77,
+      "grad_norm": 4.783046245574951,
+      "learning_rate": 1.7014802814435682e-08,
+      "loss": 0.8943,
+      "step": 66080
+    },
+    {
+      "epoch": 19.77,
+      "grad_norm": 1.3274574279785156,
+      "learning_rate": 1.6798768217776706e-08,
+      "loss": 1.0416,
+      "step": 66085
+    },
+    {
+      "epoch": 19.77,
+      "grad_norm": 1.9618406295776367,
+      "learning_rate": 1.658411341735333e-08,
+      "loss": 0.8388,
+      "step": 66090
+    },
+    {
+      "epoch": 19.77,
+      "grad_norm": 1.6042386293411255,
+      "learning_rate": 1.637083842502274e-08,
+      "loss": 1.0276,
+      "step": 66095
+    },
+    {
+      "epoch": 19.78,
+      "grad_norm": 1.6663395166397095,
+      "learning_rate": 1.6158943252558845e-08,
+      "loss": 0.885,
+      "step": 66100
+    },
+    {
+      "epoch": 19.78,
+      "grad_norm": 10.641562461853027,
+      "learning_rate": 1.594842791166895e-08,
+      "loss": 0.7403,
+      "step": 66105
+    },
+    {
+      "epoch": 19.78,
+      "grad_norm": 2.9796385765075684,
+      "learning_rate": 1.5739292413977093e-08,
+      "loss": 1.0609,
+      "step": 66110
+    },
+    {
+      "epoch": 19.78,
+      "grad_norm": 13.236249923706055,
+      "learning_rate": 1.553153677103236e-08,
+      "loss": 0.9377,
+      "step": 66115
+    },
+    {
+      "epoch": 19.78,
+      "grad_norm": 8.828291893005371,
+      "learning_rate": 1.5325160994314468e-08,
+      "loss": 1.1329,
+      "step": 66120
+    },
+    {
+      "epoch": 19.78,
+      "grad_norm": 1.7481231689453125,
+      "learning_rate": 1.5120165095217077e-08,
+      "loss": 0.9678,
+      "step": 66125
+    },
+    {
+      "epoch": 19.79,
+      "grad_norm": 5.550840854644775,
+      "learning_rate": 1.491654908506168e-08,
+      "loss": 1.0408,
+      "step": 66130
+    },
+    {
+      "epoch": 19.79,
+      "grad_norm": 1.4550460577011108,
+      "learning_rate": 1.4714312975094847e-08,
+      "loss": 0.9088,
+      "step": 66135
+    },
+    {
+      "epoch": 19.79,
+      "grad_norm": 3.302130937576294,
+      "learning_rate": 1.4513456776485413e-08,
+      "loss": 0.9667,
+      "step": 66140
+    },
+    {
+      "epoch": 19.79,
+      "grad_norm": 4.979493141174316,
+      "learning_rate": 1.4313980500327284e-08,
+      "loss": 1.0144,
+      "step": 66145
+    },
+    {
+      "epoch": 19.79,
+      "grad_norm": 3.0126538276672363,
+      "learning_rate": 1.411588415763665e-08,
+      "loss": 0.9664,
+      "step": 66150
+    },
+    {
+      "epoch": 19.79,
+      "grad_norm": 3.2353785037994385,
+      "learning_rate": 1.3919167759354756e-08,
+      "loss": 1.0019,
+      "step": 66155
+    },
+    {
+      "epoch": 19.79,
+      "grad_norm": 2.159238815307617,
+      "learning_rate": 1.3723831316345137e-08,
+      "loss": 0.8879,
+      "step": 66160
+    },
+    {
+      "epoch": 19.8,
+      "grad_norm": 2.4272637367248535,
+      "learning_rate": 1.3529874839396383e-08,
+      "loss": 0.984,
+      "step": 66165
+    },
+    {
+      "epoch": 19.8,
+      "grad_norm": 2.7232370376586914,
+      "learning_rate": 1.3337298339219372e-08,
+      "loss": 0.9735,
+      "step": 66170
+    },
+    {
+      "epoch": 19.8,
+      "grad_norm": 3.3963069915771484,
+      "learning_rate": 1.3146101826452818e-08,
+      "loss": 0.8962,
+      "step": 66175
+    },
+    {
+      "epoch": 19.8,
+      "grad_norm": 4.68864631652832,
+      "learning_rate": 1.2956285311654937e-08,
+      "loss": 0.9397,
+      "step": 66180
+    },
+    {
+      "epoch": 19.8,
+      "grad_norm": 1.9277212619781494,
+      "learning_rate": 1.2767848805309013e-08,
+      "loss": 0.9896,
+      "step": 66185
+    },
+    {
+      "epoch": 19.8,
+      "grad_norm": 1.3516850471496582,
+      "learning_rate": 1.258079231782061e-08,
+      "loss": 1.0026,
+      "step": 66190
+    },
+    {
+      "epoch": 19.8,
+      "grad_norm": 1.6840862035751343,
+      "learning_rate": 1.2395115859525907e-08,
+      "loss": 0.9797,
+      "step": 66195
+    },
+    {
+      "epoch": 19.81,
+      "grad_norm": 2.9334299564361572,
+      "learning_rate": 1.221081944067226e-08,
+      "loss": 1.1275,
+      "step": 66200
+    },
+    {
+      "epoch": 19.81,
+      "grad_norm": 16.37929916381836,
+      "learning_rate": 1.2027903071440416e-08,
+      "loss": 0.8693,
+      "step": 66205
+    },
+    {
+      "epoch": 19.81,
+      "grad_norm": 2.172717571258545,
+      "learning_rate": 1.1846366761936179e-08,
+      "loss": 0.8475,
+      "step": 66210
+    },
+    {
+      "epoch": 19.81,
+      "grad_norm": 2.417130947113037,
+      "learning_rate": 1.1666210522184861e-08,
+      "loss": 1.1559,
+      "step": 66215
+    },
+    {
+      "epoch": 19.81,
+      "grad_norm": 1.5092058181762695,
+      "learning_rate": 1.1487434362131288e-08,
+      "loss": 0.8431,
+      "step": 66220
+    },
+    {
+      "epoch": 19.81,
+      "grad_norm": 2.016624927520752,
+      "learning_rate": 1.1310038291656443e-08,
+      "loss": 1.1364,
+      "step": 66225
+    },
+    {
+      "epoch": 19.82,
+      "grad_norm": 4.214022159576416,
+      "learning_rate": 1.113402232054972e-08,
+      "loss": 0.9275,
+      "step": 66230
+    },
+    {
+      "epoch": 19.82,
+      "grad_norm": 1.9043920040130615,
+      "learning_rate": 1.0959386458539444e-08,
+      "loss": 0.9576,
+      "step": 66235
+    },
+    {
+      "epoch": 19.82,
+      "grad_norm": 2.4653420448303223,
+      "learning_rate": 1.0786130715267906e-08,
+      "loss": 0.9825,
+      "step": 66240
+    },
+    {
+      "epoch": 19.82,
+      "grad_norm": 1.7439837455749512,
+      "learning_rate": 1.0614255100299675e-08,
+      "loss": 0.9889,
+      "step": 66245
+    },
+    {
+      "epoch": 19.82,
+      "grad_norm": 2.955815076828003,
+      "learning_rate": 1.0443759623135484e-08,
+      "loss": 0.8389,
+      "step": 66250
+    },
+    {
+      "epoch": 19.82,
+      "grad_norm": 2.9133384227752686,
+      "learning_rate": 1.0274644293184477e-08,
+      "loss": 0.797,
+      "step": 66255
+    },
+    {
+      "epoch": 19.82,
+      "grad_norm": 2.106208562850952,
+      "learning_rate": 1.0106909119791952e-08,
+      "loss": 0.9157,
+      "step": 66260
+    },
+    {
+      "epoch": 19.83,
+      "grad_norm": 3.142747640609741,
+      "learning_rate": 9.940554112217171e-09,
+      "loss": 1.0129,
+      "step": 66265
+    },
+    {
+      "epoch": 19.83,
+      "grad_norm": 4.4805121421813965,
+      "learning_rate": 9.775579279650005e-09,
+      "loss": 0.9918,
+      "step": 66270
+    },
+    {
+      "epoch": 19.83,
+      "grad_norm": 1.981816053390503,
+      "learning_rate": 9.611984631202608e-09,
+      "loss": 0.9091,
+      "step": 66275
+    },
+    {
+      "epoch": 19.83,
+      "grad_norm": 2.474738359451294,
+      "learning_rate": 9.449770175909423e-09,
+      "loss": 0.9336,
+      "step": 66280
+    },
+    {
+      "epoch": 19.83,
+      "grad_norm": 3.66359281539917,
+      "learning_rate": 9.288935922727172e-09,
+      "loss": 0.9899,
+      "step": 66285
+    },
+    {
+      "epoch": 19.83,
+      "grad_norm": 3.463992118835449,
+      "learning_rate": 9.12948188054319e-09,
+      "loss": 0.6621,
+      "step": 66290
+    },
+    {
+      "epoch": 19.83,
+      "grad_norm": 2.546389102935791,
+      "learning_rate": 8.97140805815877e-09,
+      "loss": 0.9394,
+      "step": 66295
+    },
+    {
+      "epoch": 19.84,
+      "grad_norm": 2.605010509490967,
+      "learning_rate": 8.814714464308594e-09,
+      "loss": 0.9425,
+      "step": 66300
+    },
+    {
+      "epoch": 19.84,
+      "grad_norm": 1.4219180345535278,
+      "learning_rate": 8.659401107644072e-09,
+      "loss": 0.9586,
+      "step": 66305
+    },
+    {
+      "epoch": 19.84,
+      "grad_norm": 3.061012029647827,
+      "learning_rate": 8.505467996744455e-09,
+      "loss": 0.8155,
+      "step": 66310
+    },
+    {
+      "epoch": 19.84,
+      "grad_norm": 4.381716728210449,
+      "learning_rate": 8.352915140111273e-09,
+      "loss": 0.8649,
+      "step": 66315
+    },
+    {
+      "epoch": 19.84,
+      "grad_norm": 1.0354138612747192,
+      "learning_rate": 8.201742546168345e-09,
+      "loss": 0.9967,
+      "step": 66320
+    },
+    {
+      "epoch": 19.84,
+      "grad_norm": 3.057412624359131,
+      "learning_rate": 8.051950223267324e-09,
+      "loss": 1.1387,
+      "step": 66325
+    },
+    {
+      "epoch": 19.85,
+      "grad_norm": 2.2272489070892334,
+      "learning_rate": 7.903538179676595e-09,
+      "loss": 0.9963,
+      "step": 66330
+    },
+    {
+      "epoch": 19.85,
+      "grad_norm": 2.390744209289551,
+      "learning_rate": 7.756506423597932e-09,
+      "loss": 1.0075,
+      "step": 66335
+    },
+    {
+      "epoch": 19.85,
+      "grad_norm": 2.477290153503418,
+      "learning_rate": 7.610854963147063e-09,
+      "loss": 0.9847,
+      "step": 66340
+    },
+    {
+      "epoch": 19.85,
+      "grad_norm": 2.855509042739868,
+      "learning_rate": 7.466583806370331e-09,
+      "loss": 0.908,
+      "step": 66345
+    },
+    {
+      "epoch": 19.85,
+      "grad_norm": 2.9786510467529297,
+      "learning_rate": 7.323692961239137e-09,
+      "loss": 1.0602,
+      "step": 66350
+    },
+    {
+      "epoch": 19.85,
+      "grad_norm": 2.911388397216797,
+      "learning_rate": 7.18218243563884e-09,
+      "loss": 0.8163,
+      "step": 66355
+    },
+    {
+      "epoch": 19.85,
+      "grad_norm": 3.1563682556152344,
+      "learning_rate": 7.042052237390961e-09,
+      "loss": 0.9476,
+      "step": 66360
+    },
+    {
+      "epoch": 19.86,
+      "grad_norm": 3.7669315338134766,
+      "learning_rate": 6.903302374228204e-09,
+      "loss": 0.8065,
+      "step": 66365
+    },
+    {
+      "epoch": 19.86,
+      "grad_norm": 2.0955512523651123,
+      "learning_rate": 6.765932853819435e-09,
+      "loss": 1.1436,
+      "step": 66370
+    },
+    {
+      "epoch": 19.86,
+      "grad_norm": 2.7320590019226074,
+      "learning_rate": 6.62994368375025e-09,
+      "loss": 0.9004,
+      "step": 66375
+    },
+    {
+      "epoch": 19.86,
+      "grad_norm": 4.438745021820068,
+      "learning_rate": 6.495334871528536e-09,
+      "loss": 1.0057,
+      "step": 66380
+    },
+    {
+      "epoch": 19.86,
+      "grad_norm": 1.9494245052337646,
+      "learning_rate": 6.3621064245900085e-09,
+      "loss": 0.8972,
+      "step": 66385
+    },
+    {
+      "epoch": 19.86,
+      "grad_norm": 1.6904481649398804,
+      "learning_rate": 6.230258350292673e-09,
+      "loss": 0.9847,
+      "step": 66390
+    },
+    {
+      "epoch": 19.86,
+      "grad_norm": 2.8563525676727295,
+      "learning_rate": 6.09979065591959e-09,
+      "loss": 0.8644,
+      "step": 66395
+    },
+    {
+      "epoch": 19.87,
+      "grad_norm": 3.8716516494750977,
+      "learning_rate": 5.9707033486733345e-09,
+      "loss": 0.8536,
+      "step": 66400
+    },
+    {
+      "epoch": 19.87,
+      "grad_norm": 4.005788803100586,
+      "learning_rate": 5.842996435687087e-09,
+      "loss": 1.0165,
+      "step": 66405
+    },
+    {
+      "epoch": 19.87,
+      "grad_norm": 5.931815147399902,
+      "learning_rate": 5.71666992401354e-09,
+      "loss": 1.0706,
+      "step": 66410
+    },
+    {
+      "epoch": 19.87,
+      "grad_norm": 7.53218412399292,
+      "learning_rate": 5.591723820624894e-09,
+      "loss": 0.8242,
+      "step": 66415
+    },
+    {
+      "epoch": 19.87,
+      "grad_norm": 3.1983590126037598,
+      "learning_rate": 5.468158132426737e-09,
+      "loss": 0.9099,
+      "step": 66420
+    },
+    {
+      "epoch": 19.87,
+      "grad_norm": 4.075475692749023,
+      "learning_rate": 5.3459728662413886e-09,
+      "loss": 0.9736,
+      "step": 66425
+    },
+    {
+      "epoch": 19.88,
+      "grad_norm": 1.4461299180984497,
+      "learning_rate": 5.225168028819005e-09,
+      "loss": 0.944,
+      "step": 66430
+    },
+    {
+      "epoch": 19.88,
+      "grad_norm": 3.7994940280914307,
+      "learning_rate": 5.105743626829251e-09,
+      "loss": 0.9979,
+      "step": 66435
+    },
+    {
+      "epoch": 19.88,
+      "grad_norm": 2.046570301055908,
+      "learning_rate": 4.987699666869627e-09,
+      "loss": 1.1112,
+      "step": 66440
+    },
+    {
+      "epoch": 19.88,
+      "grad_norm": 3.123607873916626,
+      "learning_rate": 4.871036155454367e-09,
+      "loss": 0.9295,
+      "step": 66445
+    },
+    {
+      "epoch": 19.88,
+      "grad_norm": 3.029066801071167,
+      "learning_rate": 4.755753099033866e-09,
+      "loss": 0.9579,
+      "step": 66450
+    },
+    {
+      "epoch": 19.88,
+      "grad_norm": 2.6288695335388184,
+      "learning_rate": 4.641850503972478e-09,
+      "loss": 1.0604,
+      "step": 66455
+    },
+    {
+      "epoch": 19.88,
+      "grad_norm": 2.6345958709716797,
+      "learning_rate": 4.529328376559616e-09,
+      "loss": 1.0179,
+      "step": 66460
+    },
+    {
+      "epoch": 19.89,
+      "grad_norm": 2.6742000579833984,
+      "learning_rate": 4.418186723009754e-09,
+      "loss": 1.1658,
+      "step": 66465
+    },
+    {
+      "epoch": 19.89,
+      "grad_norm": 8.863636016845703,
+      "learning_rate": 4.3084255494652e-09,
+      "loss": 1.0605,
+      "step": 66470
+    },
+    {
+      "epoch": 19.89,
+      "grad_norm": 1.21725594997406,
+      "learning_rate": 4.200044861982222e-09,
+      "loss": 0.8926,
+      "step": 66475
+    },
+    {
+      "epoch": 19.89,
+      "grad_norm": 9.21455192565918,
+      "learning_rate": 4.093044666547696e-09,
+      "loss": 1.0839,
+      "step": 66480
+    },
+    {
+      "epoch": 19.89,
+      "grad_norm": 3.9148738384246826,
+      "learning_rate": 3.987424969073561e-09,
+      "loss": 1.0616,
+      "step": 66485
+    },
+    {
+      "epoch": 19.89,
+      "grad_norm": 3.932304859161377,
+      "learning_rate": 3.883185775394038e-09,
+      "loss": 0.8605,
+      "step": 66490
+    },
+    {
+      "epoch": 19.89,
+      "grad_norm": 2.094050168991089,
+      "learning_rate": 3.780327091262858e-09,
+      "loss": 1.0609,
+      "step": 66495
+    },
+    {
+      "epoch": 19.9,
+      "grad_norm": 2.5073344707489014,
+      "learning_rate": 3.6788489223615884e-09,
+      "loss": 0.935,
+      "step": 66500
+    },
+    {
+      "epoch": 19.9,
+      "grad_norm": 3.42728853225708,
+      "learning_rate": 3.5787512742940785e-09,
+      "loss": 0.9495,
+      "step": 66505
+    },
+    {
+      "epoch": 19.9,
+      "grad_norm": 2.900682210922241,
+      "learning_rate": 3.4800341525920156e-09,
+      "loss": 1.0571,
+      "step": 66510
+    },
+    {
+      "epoch": 19.9,
+      "grad_norm": 2.075742483139038,
+      "learning_rate": 3.3826975627038183e-09,
+      "loss": 0.8483,
+      "step": 66515
+    },
+    {
+      "epoch": 19.9,
+      "grad_norm": 1.8850741386413574,
+      "learning_rate": 3.2867415100085176e-09,
+      "loss": 0.9677,
+      "step": 66520
+    },
+    {
+      "epoch": 19.9,
+      "grad_norm": 5.589338779449463,
+      "learning_rate": 3.192165999801877e-09,
+      "loss": 1.1917,
+      "step": 66525
+    },
+    {
+      "epoch": 19.91,
+      "grad_norm": 2.845837116241455,
+      "learning_rate": 3.098971037310272e-09,
+      "loss": 0.9501,
+      "step": 66530
+    },
+    {
+      "epoch": 19.91,
+      "grad_norm": 2.0691795349121094,
+      "learning_rate": 3.0071566276795858e-09,
+      "loss": 0.9732,
+      "step": 66535
+    },
+    {
+      "epoch": 19.91,
+      "grad_norm": 2.152705430984497,
+      "learning_rate": 2.9167227759807623e-09,
+      "loss": 1.0665,
+      "step": 66540
+    },
+    {
+      "epoch": 19.91,
+      "grad_norm": 1.9462311267852783,
+      "learning_rate": 2.8276694872098053e-09,
+      "loss": 1.1731,
+      "step": 66545
+    },
+    {
+      "epoch": 19.91,
+      "grad_norm": 1.606532335281372,
+      "learning_rate": 2.739996766279451e-09,
+      "loss": 0.7424,
+      "step": 66550
+    },
+    {
+      "epoch": 19.91,
+      "grad_norm": 2.7857766151428223,
+      "learning_rate": 2.6537046180385992e-09,
+      "loss": 0.8624,
+      "step": 66555
+    },
+    {
+      "epoch": 19.91,
+      "grad_norm": 1.5321699380874634,
+      "learning_rate": 2.5687930472501066e-09,
+      "loss": 0.8455,
+      "step": 66560
+    },
+    {
+      "epoch": 19.92,
+      "grad_norm": 3.194620370864868,
+      "learning_rate": 2.485262058604665e-09,
+      "loss": 1.0387,
+      "step": 66565
+    },
+    {
+      "epoch": 19.92,
+      "grad_norm": 1.5754766464233398,
+      "learning_rate": 2.4031116567124756e-09,
+      "loss": 1.1122,
+      "step": 66570
+    },
+    {
+      "epoch": 19.92,
+      "grad_norm": 2.2193338871002197,
+      "learning_rate": 2.3223418461143507e-09,
+      "loss": 1.1581,
+      "step": 66575
+    },
+    {
+      "epoch": 19.92,
+      "grad_norm": 2.662733316421509,
+      "learning_rate": 2.242952631270612e-09,
+      "loss": 0.8239,
+      "step": 66580
+    },
+    {
+      "epoch": 19.92,
+      "grad_norm": 2.0737709999084473,
+      "learning_rate": 2.1649440165638635e-09,
+      "loss": 1.1056,
+      "step": 66585
+    },
+    {
+      "epoch": 19.92,
+      "grad_norm": 1.9437065124511719,
+      "learning_rate": 2.0883160063017713e-09,
+      "loss": 1.0208,
+      "step": 66590
+    },
+    {
+      "epoch": 19.92,
+      "grad_norm": 2.8126699924468994,
+      "learning_rate": 2.0130686047198367e-09,
+      "loss": 0.8979,
+      "step": 66595
+    },
+    {
+      "epoch": 19.93,
+      "grad_norm": 5.540438175201416,
+      "learning_rate": 1.9392018159730684e-09,
+      "loss": 0.9252,
+      "step": 66600
+    },
+    {
+      "epoch": 19.93,
+      "grad_norm": 2.1464717388153076,
+      "learning_rate": 1.8667156441387613e-09,
+      "loss": 1.1339,
+      "step": 66605
+    },
+    {
+      "epoch": 19.93,
+      "grad_norm": 1.9437625408172607,
+      "learning_rate": 1.7956100932220444e-09,
+      "loss": 0.9708,
+      "step": 66610
+    },
+    {
+      "epoch": 19.93,
+      "grad_norm": 3.8403539657592773,
+      "learning_rate": 1.7258851671503317e-09,
+      "loss": 0.9926,
+      "step": 66615
+    },
+    {
+      "epoch": 19.93,
+      "grad_norm": 2.2192623615264893,
+      "learning_rate": 1.657540869773322e-09,
+      "loss": 1.0566,
+      "step": 66620
+    },
+    {
+      "epoch": 19.93,
+      "grad_norm": 4.014240741729736,
+      "learning_rate": 1.5905772048629973e-09,
+      "loss": 0.8297,
+      "step": 66625
+    },
+    {
+      "epoch": 19.93,
+      "grad_norm": 4.3666181564331055,
+      "learning_rate": 1.5249941761247276e-09,
+      "loss": 0.9347,
+      "step": 66630
+    },
+    {
+      "epoch": 19.94,
+      "grad_norm": 1.6916534900665283,
+      "learning_rate": 1.4607917871750643e-09,
+      "loss": 0.9515,
+      "step": 66635
+    },
+    {
+      "epoch": 19.94,
+      "grad_norm": 1.9983826875686646,
+      "learning_rate": 1.3979700415611697e-09,
+      "loss": 0.9741,
+      "step": 66640
+    },
+    {
+      "epoch": 19.94,
+      "grad_norm": 2.9387166500091553,
+      "learning_rate": 1.3365289427524908e-09,
+      "loss": 0.8877,
+      "step": 66645
+    },
+    {
+      "epoch": 19.94,
+      "grad_norm": 2.9667611122131348,
+      "learning_rate": 1.2764684941435346e-09,
+      "loss": 0.8683,
+      "step": 66650
+    },
+    {
+      "epoch": 19.94,
+      "grad_norm": 1.2292882204055786,
+      "learning_rate": 1.2177886990510923e-09,
+      "loss": 0.8254,
+      "step": 66655
+    },
+    {
+      "epoch": 19.94,
+      "grad_norm": 1.751691222190857,
+      "learning_rate": 1.1604895607142397e-09,
+      "loss": 0.9371,
+      "step": 66660
+    },
+    {
+      "epoch": 19.95,
+      "grad_norm": 2.478400945663452,
+      "learning_rate": 1.1045710822971123e-09,
+      "loss": 0.981,
+      "step": 66665
+    },
+    {
+      "epoch": 19.95,
+      "grad_norm": 5.2517781257629395,
+      "learning_rate": 1.0500332668916813e-09,
+      "loss": 0.8857,
+      "step": 66670
+    },
+    {
+      "epoch": 19.95,
+      "grad_norm": 3.200549840927124,
+      "learning_rate": 9.968761175094267e-10,
+      "loss": 1.0439,
+      "step": 66675
+    },
+    {
+      "epoch": 19.95,
+      "grad_norm": 3.636068105697632,
+      "learning_rate": 9.45099637081337e-10,
+      "loss": 0.9732,
+      "step": 66680
+    },
+    {
+      "epoch": 19.95,
+      "grad_norm": 2.145059823989868,
+      "learning_rate": 8.947038284717879e-10,
+      "loss": 0.7454,
+      "step": 66685
+    },
+    {
+      "epoch": 19.95,
+      "grad_norm": 1.3678562641143799,
+      "learning_rate": 8.456886944646636e-10,
+      "loss": 1.0683,
+      "step": 66690
+    },
+    {
+      "epoch": 19.95,
+      "grad_norm": 2.720142364501953,
+      "learning_rate": 7.980542377633571e-10,
+      "loss": 0.8569,
+      "step": 66695
+    },
+    {
+      "epoch": 19.96,
+      "grad_norm": 5.969086647033691,
+      "learning_rate": 7.51800460999097e-10,
+      "loss": 0.9017,
+      "step": 66700
+    },
+    {
+      "epoch": 19.96,
+      "grad_norm": 1.7213473320007324,
+      "learning_rate": 7.069273667281717e-10,
+      "loss": 0.8733,
+      "step": 66705
+    },
+    {
+      "epoch": 19.96,
+      "grad_norm": 1.9708019495010376,
+      "learning_rate": 6.634349574291543e-10,
+      "loss": 0.978,
+      "step": 66710
+    },
+    {
+      "epoch": 19.96,
+      "grad_norm": 1.8042583465576172,
+      "learning_rate": 6.213232355029019e-10,
+      "loss": 1.2455,
+      "step": 66715
+    },
+    {
+      "epoch": 19.96,
+      "grad_norm": 1.776975393295288,
+      "learning_rate": 5.805922032753319e-10,
+      "loss": 0.7985,
+      "step": 66720
+    },
+    {
+      "epoch": 19.96,
+      "grad_norm": 2.806536912918091,
+      "learning_rate": 5.412418629946459e-10,
+      "loss": 0.9752,
+      "step": 66725
+    },
+    {
+      "epoch": 19.96,
+      "grad_norm": 1.0485270023345947,
+      "learning_rate": 5.032722168396564e-10,
+      "loss": 0.8385,
+      "step": 66730
+    },
+    {
+      "epoch": 19.97,
+      "grad_norm": 2.4013590812683105,
+      "learning_rate": 4.666832669003585e-10,
+      "loss": 0.8629,
+      "step": 66735
+    },
+    {
+      "epoch": 19.97,
+      "grad_norm": 3.498737335205078,
+      "learning_rate": 4.3147501520013347e-10,
+      "loss": 1.2686,
+      "step": 66740
+    },
+    {
+      "epoch": 19.97,
+      "grad_norm": 2.2071971893310547,
+      "learning_rate": 3.9764746368742277e-10,
+      "loss": 1.0284,
+      "step": 66745
+    },
+    {
+      "epoch": 19.97,
+      "grad_norm": 2.6617162227630615,
+      "learning_rate": 3.652006142246256e-10,
+      "loss": 0.9895,
+      "step": 66750
+    },
+    {
+      "epoch": 19.97,
+      "grad_norm": 2.3580636978149414,
+      "learning_rate": 3.3413446860475204e-10,
+      "loss": 1.0622,
+      "step": 66755
+    },
+    {
+      "epoch": 19.97,
+      "grad_norm": 2.603350877761841,
+      "learning_rate": 3.044490285458723e-10,
+      "loss": 0.9365,
+      "step": 66760
+    },
+    {
+      "epoch": 19.98,
+      "grad_norm": 2.9707422256469727,
+      "learning_rate": 2.7614429568834087e-10,
+      "loss": 0.9106,
+      "step": 66765
+    },
+    {
+      "epoch": 19.98,
+      "grad_norm": 1.205654501914978,
+      "learning_rate": 2.4922027159202113e-10,
+      "loss": 0.9287,
+      "step": 66770
+    },
+    {
+      "epoch": 19.98,
+      "grad_norm": 1.2524653673171997,
+      "learning_rate": 2.2367695774461185e-10,
+      "loss": 1.0583,
+      "step": 66775
+    },
+    {
+      "epoch": 19.98,
+      "grad_norm": 1.9343750476837158,
+      "learning_rate": 1.9951435555887187e-10,
+      "loss": 1.0717,
+      "step": 66780
+    },
+    {
+      "epoch": 19.98,
+      "grad_norm": 2.118802309036255,
+      "learning_rate": 1.7673246636984443e-10,
+      "loss": 0.962,
+      "step": 66785
+    },
+    {
+      "epoch": 19.98,
+      "grad_norm": 2.374420404434204,
+      "learning_rate": 1.5533129143208146e-10,
+      "loss": 0.9052,
+      "step": 66790
+    },
+    {
+      "epoch": 19.98,
+      "grad_norm": 3.1811416149139404,
+      "learning_rate": 1.3531083193074612e-10,
+      "loss": 0.8696,
+      "step": 66795
+    },
+    {
+      "epoch": 19.99,
+      "grad_norm": 1.0755863189697266,
+      "learning_rate": 1.166710889677347e-10,
+      "loss": 0.7908,
+      "step": 66800
+    },
+    {
+      "epoch": 19.99,
+      "grad_norm": 4.653294086456299,
+      "learning_rate": 9.941206357555467e-11,
+      "loss": 0.968,
+      "step": 66805
+    },
+    {
+      "epoch": 19.99,
+      "grad_norm": 1.384509563446045,
+      "learning_rate": 8.353375670899777e-11,
+      "loss": 1.0534,
+      "step": 66810
+    },
+    {
+      "epoch": 19.99,
+      "grad_norm": 6.829671859741211,
+      "learning_rate": 6.903616924236466e-11,
+      "loss": 0.934,
+      "step": 66815
+    },
+    {
+      "epoch": 19.99,
+      "grad_norm": 2.6626029014587402,
+      "learning_rate": 5.591930197501594e-11,
+      "loss": 0.9137,
+      "step": 66820
+    },
+    {
+      "epoch": 19.99,
+      "grad_norm": 3.337085723876953,
+      "learning_rate": 4.418315563414765e-11,
+      "loss": 0.9638,
+      "step": 66825
+    },
+    {
+      "epoch": 19.99,
+      "grad_norm": 2.169858694076538,
+      "learning_rate": 3.3827730869240294e-11,
+      "loss": 0.9035,
+      "step": 66830
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 3.450448513031006,
+      "learning_rate": 2.4853028246507594e-11,
+      "loss": 0.9036,
+      "step": 66835
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 2.451824188232422,
+      "learning_rate": 1.7259048268325474e-11,
+      "loss": 0.8429,
+      "step": 66840
+    },
+    {
+      "epoch": 20.0,
+      "step": 66840,
+      "total_flos": 3.489559903035851e+18,
+      "train_loss": 1.1193655486238139,
+      "train_runtime": 133767.8538,
+      "train_samples_per_second": 3.998,
+      "train_steps_per_second": 0.5
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 66840,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 100,
+  "total_flos": 3.489559903035851e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}