diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24764 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 17646,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 2496.8073030248033,
+      "learning_rate": 5.6657223796033996e-09,
+      "loss": 12.3922,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2798.9639739819045,
+      "learning_rate": 2.8328611898017002e-08,
+      "loss": 12.556,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3186.5811854816743,
+      "learning_rate": 5.6657223796034004e-08,
+      "loss": 12.6663,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2433.131292284797,
+      "learning_rate": 8.4985835694051e-08,
+      "loss": 11.8166,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 927.7244167671865,
+      "learning_rate": 1.1331444759206801e-07,
+      "loss": 9.872,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 598.7163941919265,
+      "learning_rate": 1.41643059490085e-07,
+      "loss": 8.1112,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 593.7817469340148,
+      "learning_rate": 1.69971671388102e-07,
+      "loss": 7.6257,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 361.1842229161203,
+      "learning_rate": 1.9830028328611898e-07,
+      "loss": 6.8663,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 261.41360349547375,
+      "learning_rate": 2.2662889518413602e-07,
+      "loss": 6.297,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 294.80563694492105,
+      "learning_rate": 2.54957507082153e-07,
+      "loss": 5.8564,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 223.89911973808375,
+      "learning_rate": 2.8328611898017e-07,
+      "loss": 5.5386,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 205.52507903905445,
+      "learning_rate": 3.1161473087818695e-07,
+      "loss": 5.232,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 115.75404026789134,
+      "learning_rate": 3.39943342776204e-07,
+      "loss": 5.065,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 76.96969560974317,
+      "learning_rate": 3.6827195467422096e-07,
+      "loss": 4.862,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 80.34469970951176,
+      "learning_rate": 3.9660056657223797e-07,
+      "loss": 4.7516,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 81.7595945293725,
+      "learning_rate": 4.24929178470255e-07,
+      "loss": 4.5732,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 63.15931120806828,
+      "learning_rate": 4.5325779036827203e-07,
+      "loss": 4.4904,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 56.380024867999694,
+      "learning_rate": 4.815864022662889e-07,
+      "loss": 4.3397,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 60.056632903837375,
+      "learning_rate": 5.09915014164306e-07,
+      "loss": 4.2419,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 57.72643638091866,
+      "learning_rate": 5.382436260623229e-07,
+      "loss": 4.2126,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 44.8303385559054,
+      "learning_rate": 5.6657223796034e-07,
+      "loss": 4.083,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 45.71148017336654,
+      "learning_rate": 5.949008498583571e-07,
+      "loss": 3.9693,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 46.76823280991884,
+      "learning_rate": 6.232294617563739e-07,
+      "loss": 3.8478,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 42.27034196603908,
+      "learning_rate": 6.51558073654391e-07,
+      "loss": 3.8355,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 43.786683236286535,
+      "learning_rate": 6.79886685552408e-07,
+      "loss": 3.7417,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 33.38556504010887,
+      "learning_rate": 7.08215297450425e-07,
+      "loss": 3.619,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 38.77121164219573,
+      "learning_rate": 7.365439093484419e-07,
+      "loss": 3.6023,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 35.85722948165567,
+      "learning_rate": 7.64872521246459e-07,
+      "loss": 3.5028,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.87382466876409,
+      "learning_rate": 7.932011331444759e-07,
+      "loss": 3.3987,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 28.499096674335632,
+      "learning_rate": 8.215297450424931e-07,
+      "loss": 3.3644,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 26.379175262470692,
+      "learning_rate": 8.4985835694051e-07,
+      "loss": 3.2703,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 29.865960399124674,
+      "learning_rate": 8.781869688385269e-07,
+      "loss": 3.2143,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.246604944858476,
+      "learning_rate": 9.065155807365441e-07,
+      "loss": 3.1724,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 26.803614997802278,
+      "learning_rate": 9.34844192634561e-07,
+      "loss": 3.1165,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.868744499172813,
+      "learning_rate": 9.631728045325779e-07,
+      "loss": 3.0226,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 28.072404376457335,
+      "learning_rate": 9.91501416430595e-07,
+      "loss": 2.974,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.13981569942378,
+      "learning_rate": 1.019830028328612e-06,
+      "loss": 2.8986,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.682144066119207,
+      "learning_rate": 1.048158640226629e-06,
+      "loss": 2.8929,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 26.393633161733234,
+      "learning_rate": 1.0764872521246459e-06,
+      "loss": 2.8201,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.165888181101067,
+      "learning_rate": 1.104815864022663e-06,
+      "loss": 2.7715,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.497827155588414,
+      "learning_rate": 1.13314447592068e-06,
+      "loss": 2.7423,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.577170865053922,
+      "learning_rate": 1.1614730878186968e-06,
+      "loss": 2.6814,
+      "step": 205
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.811068284154047,
+      "learning_rate": 1.1898016997167141e-06,
+      "loss": 2.681,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 23.625896687896763,
+      "learning_rate": 1.218130311614731e-06,
+      "loss": 2.6641,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.558263239239487,
+      "learning_rate": 1.2464589235127478e-06,
+      "loss": 2.6209,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 28.59558876783872,
+      "learning_rate": 1.274787535410765e-06,
+      "loss": 2.6035,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 22.99206869102384,
+      "learning_rate": 1.303116147308782e-06,
+      "loss": 2.5784,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 24.770207084904367,
+      "learning_rate": 1.331444759206799e-06,
+      "loss": 2.5403,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.943253914438525,
+      "learning_rate": 1.359773371104816e-06,
+      "loss": 2.4988,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 28.669675628003457,
+      "learning_rate": 1.388101983002833e-06,
+      "loss": 2.4725,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 32.51240263679756,
+      "learning_rate": 1.41643059490085e-06,
+      "loss": 2.4197,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 24.03710460536725,
+      "learning_rate": 1.444759206798867e-06,
+      "loss": 2.4541,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 30.538221432223352,
+      "learning_rate": 1.4730878186968839e-06,
+      "loss": 2.375,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 27.4383780573818,
+      "learning_rate": 1.5014164305949011e-06,
+      "loss": 2.4054,
+      "step": 265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 38.83226342367871,
+      "learning_rate": 1.529745042492918e-06,
+      "loss": 2.4019,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 22.04189249067676,
+      "learning_rate": 1.558073654390935e-06,
+      "loss": 2.3203,
+      "step": 275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.17958111536244,
+      "learning_rate": 1.5864022662889519e-06,
+      "loss": 2.3432,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 22.95410575681228,
+      "learning_rate": 1.614730878186969e-06,
+      "loss": 2.3145,
+      "step": 285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.93998688099864,
+      "learning_rate": 1.6430594900849862e-06,
+      "loss": 2.26,
+      "step": 290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 21.32107695775869,
+      "learning_rate": 1.671388101983003e-06,
+      "loss": 2.2498,
+      "step": 295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 24.70050367023582,
+      "learning_rate": 1.69971671388102e-06,
+      "loss": 2.2334,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 48.55117360385103,
+      "learning_rate": 1.728045325779037e-06,
+      "loss": 2.2476,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 27.055278892184113,
+      "learning_rate": 1.7563739376770538e-06,
+      "loss": 2.2039,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.84999881904028,
+      "learning_rate": 1.7847025495750709e-06,
+      "loss": 2.1813,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 40.40413661716398,
+      "learning_rate": 1.8130311614730881e-06,
+      "loss": 2.2304,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 32.46690888295986,
+      "learning_rate": 1.841359773371105e-06,
+      "loss": 2.2263,
+      "step": 325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 20.51738634970082,
+      "learning_rate": 1.869688385269122e-06,
+      "loss": 2.1234,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 35.222336018483524,
+      "learning_rate": 1.8980169971671389e-06,
+      "loss": 2.1573,
+      "step": 335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 37.7368422922636,
+      "learning_rate": 1.9263456090651557e-06,
+      "loss": 2.1452,
+      "step": 340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 21.56463831517406,
+      "learning_rate": 1.954674220963173e-06,
+      "loss": 2.128,
+      "step": 345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 23.972624335060562,
+      "learning_rate": 1.98300283286119e-06,
+      "loss": 2.1398,
+      "step": 350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 20.213325978350014,
+      "learning_rate": 2.011331444759207e-06,
+      "loss": 2.0917,
+      "step": 355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 32.978422655923445,
+      "learning_rate": 2.039660056657224e-06,
+      "loss": 2.0906,
+      "step": 360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 24.072172345830047,
+      "learning_rate": 2.067988668555241e-06,
+      "loss": 2.073,
+      "step": 365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.862216422542378,
+      "learning_rate": 2.096317280453258e-06,
+      "loss": 2.0942,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.782579148305224,
+      "learning_rate": 2.124645892351275e-06,
+      "loss": 2.0978,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 27.64608563553747,
+      "learning_rate": 2.1529745042492918e-06,
+      "loss": 2.0611,
+      "step": 380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 25.729919004188048,
+      "learning_rate": 2.181303116147309e-06,
+      "loss": 2.047,
+      "step": 385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 19.56297139440245,
+      "learning_rate": 2.209631728045326e-06,
+      "loss": 2.0339,
+      "step": 390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 19.52987241347125,
+      "learning_rate": 2.237960339943343e-06,
+      "loss": 2.0449,
+      "step": 395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.353166316715352,
+      "learning_rate": 2.26628895184136e-06,
+      "loss": 2.0223,
+      "step": 400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.933589191705828,
+      "learning_rate": 2.294617563739377e-06,
+      "loss": 2.0439,
+      "step": 405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.840396501445202,
+      "learning_rate": 2.3229461756373937e-06,
+      "loss": 2.0128,
+      "step": 410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 38.57601116485757,
+      "learning_rate": 2.3512747875354108e-06,
+      "loss": 2.0437,
+      "step": 415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 41.85780844791552,
+      "learning_rate": 2.3796033994334282e-06,
+      "loss": 1.9994,
+      "step": 420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 35.48875979085677,
+      "learning_rate": 2.407932011331445e-06,
+      "loss": 2.0033,
+      "step": 425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 43.38637323829772,
+      "learning_rate": 2.436260623229462e-06,
+      "loss": 2.005,
+      "step": 430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 19.900163750049977,
+      "learning_rate": 2.464589235127479e-06,
+      "loss": 1.9753,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 22.204455228895014,
+      "learning_rate": 2.4929178470254956e-06,
+      "loss": 1.9557,
+      "step": 440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.74787616134596,
+      "learning_rate": 2.521246458923513e-06,
+      "loss": 1.9367,
+      "step": 445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.03296681245724,
+      "learning_rate": 2.54957507082153e-06,
+      "loss": 1.9577,
+      "step": 450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.070555798078754,
+      "learning_rate": 2.577903682719547e-06,
+      "loss": 1.9425,
+      "step": 455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 18.095413622860136,
+      "learning_rate": 2.606232294617564e-06,
+      "loss": 1.9424,
+      "step": 460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 37.1892575296801,
+      "learning_rate": 2.634560906515581e-06,
+      "loss": 1.941,
+      "step": 465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 30.687534445988742,
+      "learning_rate": 2.662889518413598e-06,
+      "loss": 1.9047,
+      "step": 470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.424833044102115,
+      "learning_rate": 2.6912181303116146e-06,
+      "loss": 1.9292,
+      "step": 475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 27.360616215950174,
+      "learning_rate": 2.719546742209632e-06,
+      "loss": 1.9009,
+      "step": 480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 40.15039463581183,
+      "learning_rate": 2.747875354107649e-06,
+      "loss": 1.9329,
+      "step": 485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.5561162952693,
+      "learning_rate": 2.776203966005666e-06,
+      "loss": 1.8924,
+      "step": 490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 20.21398000813242,
+      "learning_rate": 2.804532577903683e-06,
+      "loss": 1.8809,
+      "step": 495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 32.098925635303246,
+      "learning_rate": 2.8328611898017e-06,
+      "loss": 1.9138,
+      "step": 500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 42.05244685784008,
+      "learning_rate": 2.861189801699717e-06,
+      "loss": 1.8856,
+      "step": 505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.005208485231954,
+      "learning_rate": 2.889518413597734e-06,
+      "loss": 1.841,
+      "step": 510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 28.615268279304342,
+      "learning_rate": 2.9178470254957506e-06,
+      "loss": 1.8342,
+      "step": 515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 36.647396681534886,
+      "learning_rate": 2.9461756373937677e-06,
+      "loss": 1.8696,
+      "step": 520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 52.74806377562607,
+      "learning_rate": 2.9745042492917848e-06,
+      "loss": 1.8655,
+      "step": 525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 35.82795457958816,
+      "learning_rate": 3.0028328611898022e-06,
+      "loss": 1.8683,
+      "step": 530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 29.29875595295583,
+      "learning_rate": 3.0311614730878193e-06,
+      "loss": 1.8225,
+      "step": 535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 31.058050277581376,
+      "learning_rate": 3.059490084985836e-06,
+      "loss": 1.8488,
+      "step": 540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 23.73823300434112,
+      "learning_rate": 3.087818696883853e-06,
+      "loss": 1.8479,
+      "step": 545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 19.57654252274149,
+      "learning_rate": 3.11614730878187e-06,
+      "loss": 1.8151,
+      "step": 550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.0532264932674,
+      "learning_rate": 3.144475920679887e-06,
+      "loss": 1.8559,
+      "step": 555
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.48136081087222,
+      "learning_rate": 3.1728045325779038e-06,
+      "loss": 1.7867,
+      "step": 560
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.273223063451194,
+      "learning_rate": 3.201133144475921e-06,
+      "loss": 1.8184,
+      "step": 565
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 22.305551298403266,
+      "learning_rate": 3.229461756373938e-06,
+      "loss": 1.792,
+      "step": 570
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 51.38398971351818,
+      "learning_rate": 3.257790368271955e-06,
+      "loss": 1.8154,
+      "step": 575
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 49.786223800122116,
+      "learning_rate": 3.2861189801699724e-06,
+      "loss": 1.8044,
+      "step": 580
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 47.56380663604067,
+      "learning_rate": 3.314447592067989e-06,
+      "loss": 1.8327,
+      "step": 585
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 44.47524479659074,
+      "learning_rate": 3.342776203966006e-06,
+      "loss": 1.8105,
+      "step": 590
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 33.604777311500975,
+      "learning_rate": 3.371104815864023e-06,
+      "loss": 1.7709,
+      "step": 595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 28.320348049863327,
+      "learning_rate": 3.39943342776204e-06,
+      "loss": 1.7832,
+      "step": 600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 27.15094992711375,
+      "learning_rate": 3.427762039660057e-06,
+      "loss": 1.7769,
+      "step": 605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.163403461295847,
+      "learning_rate": 3.456090651558074e-06,
+      "loss": 1.7995,
+      "step": 610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 27.076996397342533,
+      "learning_rate": 3.484419263456091e-06,
+      "loss": 1.7721,
+      "step": 615
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.08583043568208,
+      "learning_rate": 3.5127478753541076e-06,
+      "loss": 1.7494,
+      "step": 620
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 34.60488151573862,
+      "learning_rate": 3.5410764872521247e-06,
+      "loss": 1.7875,
+      "step": 625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 19.583304295440154,
+      "learning_rate": 3.5694050991501417e-06,
+      "loss": 1.7533,
+      "step": 630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 42.29524810809757,
+      "learning_rate": 3.597733711048159e-06,
+      "loss": 1.7644,
+      "step": 635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 66.58393026110944,
+      "learning_rate": 3.6260623229461763e-06,
+      "loss": 1.7385,
+      "step": 640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 31.076394544761435,
+      "learning_rate": 3.654390934844193e-06,
+      "loss": 1.742,
+      "step": 645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.91941602854681,
+      "learning_rate": 3.68271954674221e-06,
+      "loss": 1.7907,
+      "step": 650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.48663060720521,
+      "learning_rate": 3.711048158640227e-06,
+      "loss": 1.772,
+      "step": 655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 21.169491370728753,
+      "learning_rate": 3.739376770538244e-06,
+      "loss": 1.7613,
+      "step": 660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 19.394119690737643,
+      "learning_rate": 3.7677053824362607e-06,
+      "loss": 1.7357,
+      "step": 665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 36.97000360839211,
+      "learning_rate": 3.7960339943342778e-06,
+      "loss": 1.7568,
+      "step": 670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 41.170536011729254,
+      "learning_rate": 3.824362606232295e-06,
+      "loss": 1.7878,
+      "step": 675
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.91271833985899,
+      "learning_rate": 3.8526912181303115e-06,
+      "loss": 1.735,
+      "step": 680
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.177589104249861,
+      "learning_rate": 3.881019830028329e-06,
+      "loss": 1.7479,
+      "step": 685
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 53.90923998912517,
+      "learning_rate": 3.909348441926346e-06,
+      "loss": 1.7264,
+      "step": 690
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 79.43302487475911,
+      "learning_rate": 3.937677053824363e-06,
+      "loss": 1.6986,
+      "step": 695
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 20.1848106409289,
+      "learning_rate": 3.96600566572238e-06,
+      "loss": 1.7403,
+      "step": 700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.977411772592603,
+      "learning_rate": 3.994334277620397e-06,
+      "loss": 1.729,
+      "step": 705
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.60858292804487,
+      "learning_rate": 4.022662889518414e-06,
+      "loss": 1.7391,
+      "step": 710
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 25.667932494840965,
+      "learning_rate": 4.0509915014164304e-06,
+      "loss": 1.7108,
+      "step": 715
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 20.78540678599752,
+      "learning_rate": 4.079320113314448e-06,
+      "loss": 1.7221,
+      "step": 720
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 76.42698172331296,
+      "learning_rate": 4.1076487252124646e-06,
+      "loss": 1.7105,
+      "step": 725
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 71.16243308160482,
+      "learning_rate": 4.135977337110482e-06,
+      "loss": 1.727,
+      "step": 730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.40654783870013,
+      "learning_rate": 4.1643059490084995e-06,
+      "loss": 1.6936,
+      "step": 735
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 34.693252022614544,
+      "learning_rate": 4.192634560906516e-06,
+      "loss": 1.6621,
+      "step": 740
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 38.53652676282218,
+      "learning_rate": 4.220963172804533e-06,
+      "loss": 1.6904,
+      "step": 745
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 26.83929663571352,
+      "learning_rate": 4.24929178470255e-06,
+      "loss": 1.6837,
+      "step": 750
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 45.81507221553908,
+      "learning_rate": 4.277620396600567e-06,
+      "loss": 1.6989,
+      "step": 755
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 33.692133005547205,
+      "learning_rate": 4.3059490084985835e-06,
+      "loss": 1.6386,
+      "step": 760
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 19.38330427723925,
+      "learning_rate": 4.334277620396601e-06,
+      "loss": 1.6653,
+      "step": 765
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 30.678965153137693,
+      "learning_rate": 4.362606232294618e-06,
+      "loss": 1.6807,
+      "step": 770
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.165456765295332,
+      "learning_rate": 4.390934844192635e-06,
+      "loss": 1.6987,
+      "step": 775
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 32.29327910090702,
+      "learning_rate": 4.419263456090652e-06,
+      "loss": 1.6802,
+      "step": 780
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.16971480892796,
+      "learning_rate": 4.447592067988669e-06,
+      "loss": 1.6497,
+      "step": 785
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 26.707784735238036,
+      "learning_rate": 4.475920679886686e-06,
+      "loss": 1.6604,
+      "step": 790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 50.87756098034557,
+      "learning_rate": 4.504249291784703e-06,
+      "loss": 1.6326,
+      "step": 795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 52.23774283052041,
+      "learning_rate": 4.53257790368272e-06,
+      "loss": 1.6485,
+      "step": 800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 63.05813173390926,
+      "learning_rate": 4.560906515580737e-06,
+      "loss": 1.6834,
+      "step": 805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 22.075712655510806,
+      "learning_rate": 4.589235127478754e-06,
+      "loss": 1.6382,
+      "step": 810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 29.09471283050641,
+      "learning_rate": 4.617563739376771e-06,
+      "loss": 1.6572,
+      "step": 815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 43.606834380034115,
+      "learning_rate": 4.645892351274787e-06,
+      "loss": 1.6566,
+      "step": 820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 49.418261757308244,
+      "learning_rate": 4.674220963172805e-06,
+      "loss": 1.6501,
+      "step": 825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 60.58000269353636,
+      "learning_rate": 4.7025495750708215e-06,
+      "loss": 1.6284,
+      "step": 830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.768535580269724,
+      "learning_rate": 4.730878186968839e-06,
+      "loss": 1.6342,
+      "step": 835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.27132812623081,
+      "learning_rate": 4.7592067988668565e-06,
+      "loss": 1.6236,
+      "step": 840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.37282219238344,
+      "learning_rate": 4.787535410764873e-06,
+      "loss": 1.6277,
+      "step": 845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 46.068808485877284,
+      "learning_rate": 4.81586402266289e-06,
+      "loss": 1.6378,
+      "step": 850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 64.72578782324423,
+      "learning_rate": 4.844192634560907e-06,
+      "loss": 1.6242,
+      "step": 855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 36.406125553208405,
+      "learning_rate": 4.872521246458924e-06,
+      "loss": 1.625,
+      "step": 860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.157180999051647,
+      "learning_rate": 4.9008498583569405e-06,
+      "loss": 1.6419,
+      "step": 865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 31.267783331002605,
+      "learning_rate": 4.929178470254958e-06,
+      "loss": 1.6156,
+      "step": 870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 30.51950308301667,
+      "learning_rate": 4.957507082152975e-06,
+      "loss": 1.6127,
+      "step": 875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 59.65679660528968,
+      "learning_rate": 4.985835694050991e-06,
+      "loss": 1.5823,
+      "step": 880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 33.09588742022726,
+      "learning_rate": 5.014164305949009e-06,
+      "loss": 1.6119,
+      "step": 885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 17.46276943511441,
+      "learning_rate": 5.042492917847026e-06,
+      "loss": 1.5669,
+      "step": 890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 27.95986512759189,
+      "learning_rate": 5.070821529745043e-06,
+      "loss": 1.6021,
+      "step": 895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 40.53798724973572,
+      "learning_rate": 5.09915014164306e-06,
+      "loss": 1.5952,
+      "step": 900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 98.13247181294633,
+      "learning_rate": 5.127478753541076e-06,
+      "loss": 1.6185,
+      "step": 905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 52.03314426839591,
+      "learning_rate": 5.155807365439094e-06,
+      "loss": 1.6101,
+      "step": 910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 64.46486065152293,
+      "learning_rate": 5.184135977337111e-06,
+      "loss": 1.5932,
+      "step": 915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 55.85131329674641,
+      "learning_rate": 5.212464589235128e-06,
+      "loss": 1.5597,
+      "step": 920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 104.94386787582185,
+      "learning_rate": 5.240793201133145e-06,
+      "loss": 1.6062,
+      "step": 925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 55.90312620104168,
+      "learning_rate": 5.269121813031162e-06,
+      "loss": 1.6001,
+      "step": 930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 36.37767655193305,
+      "learning_rate": 5.297450424929179e-06,
+      "loss": 1.5993,
+      "step": 935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 39.26128766918641,
+      "learning_rate": 5.325779036827196e-06,
+      "loss": 1.5651,
+      "step": 940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 43.57675317461804,
+      "learning_rate": 5.3541076487252134e-06,
+      "loss": 1.5913,
+      "step": 945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 17.26426628907133,
+      "learning_rate": 5.382436260623229e-06,
+      "loss": 1.594,
+      "step": 950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.037423782357767,
+      "learning_rate": 5.410764872521247e-06,
+      "loss": 1.5744,
+      "step": 955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 130.6445045404753,
+      "learning_rate": 5.439093484419264e-06,
+      "loss": 1.6486,
+      "step": 960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 54.45429682389578,
+      "learning_rate": 5.467422096317281e-06,
+      "loss": 1.5734,
+      "step": 965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 42.69622349959172,
+      "learning_rate": 5.495750708215298e-06,
+      "loss": 1.6064,
+      "step": 970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 59.71378250423668,
+      "learning_rate": 5.524079320113315e-06,
+      "loss": 1.6193,
+      "step": 975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 62.55182445293084,
+      "learning_rate": 5.552407932011332e-06,
+      "loss": 1.5776,
+      "step": 980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 59.05878711283353,
+      "learning_rate": 5.580736543909348e-06,
+      "loss": 1.5794,
+      "step": 985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 28.560889477217298,
+      "learning_rate": 5.609065155807366e-06,
+      "loss": 1.5593,
+      "step": 990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 27.490073642677785,
+      "learning_rate": 5.637393767705382e-06,
+      "loss": 1.5749,
+      "step": 995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.950065631003085,
+      "learning_rate": 5.6657223796034e-06,
+      "loss": 1.5537,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 43.35553144079011,
+      "learning_rate": 5.6940509915014164e-06,
+      "loss": 1.5762,
+      "step": 1005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 28.678343487108393,
+      "learning_rate": 5.722379603399434e-06,
+      "loss": 1.5324,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 98.86163137432244,
+      "learning_rate": 5.750708215297451e-06,
+      "loss": 1.5333,
+      "step": 1015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 63.299358135948566,
+      "learning_rate": 5.779036827195468e-06,
+      "loss": 1.542,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.56384405996424,
+      "learning_rate": 5.8073654390934855e-06,
+      "loss": 1.5661,
+      "step": 1025
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 27.414836270647065,
+      "learning_rate": 5.835694050991501e-06,
+      "loss": 1.564,
+      "step": 1030
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.974654931798312,
+      "learning_rate": 5.864022662889519e-06,
+      "loss": 1.5267,
+      "step": 1035
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.14250216808397,
+      "learning_rate": 5.892351274787535e-06,
+      "loss": 1.5485,
+      "step": 1040
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.483781002585005,
+      "learning_rate": 5.920679886685553e-06,
+      "loss": 1.5312,
+      "step": 1045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.977930403123093,
+      "learning_rate": 5.9490084985835695e-06,
+      "loss": 1.5574,
+      "step": 1050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 35.868563902903155,
+      "learning_rate": 5.977337110481587e-06,
+      "loss": 1.5089,
+      "step": 1055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 35.70653189378658,
+      "learning_rate": 6.0056657223796045e-06,
+      "loss": 1.4986,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 83.85460986190164,
+      "learning_rate": 6.033994334277621e-06,
+      "loss": 1.5424,
+      "step": 1065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 52.01364066249039,
+      "learning_rate": 6.062322946175639e-06,
+      "loss": 1.5451,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 41.4992437255658,
+      "learning_rate": 6.090651558073654e-06,
+      "loss": 1.5275,
+      "step": 1075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 30.925485991933222,
+      "learning_rate": 6.118980169971672e-06,
+      "loss": 1.5082,
+      "step": 1080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 44.720242480528114,
+      "learning_rate": 6.1473087818696885e-06,
+      "loss": 1.5481,
+      "step": 1085
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 53.91170639021118,
+      "learning_rate": 6.175637393767706e-06,
+      "loss": 1.5137,
+      "step": 1090
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 78.19398697511646,
+      "learning_rate": 6.203966005665723e-06,
+      "loss": 1.5195,
+      "step": 1095
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 39.94674555010526,
+      "learning_rate": 6.23229461756374e-06,
+      "loss": 1.5273,
+      "step": 1100
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.855101814171903,
+      "learning_rate": 6.260623229461757e-06,
+      "loss": 1.5309,
+      "step": 1105
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 44.867726205072024,
+      "learning_rate": 6.288951841359774e-06,
+      "loss": 1.5242,
+      "step": 1110
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 29.131859743656516,
+      "learning_rate": 6.317280453257792e-06,
+      "loss": 1.5273,
+      "step": 1115
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.611183595956476,
+      "learning_rate": 6.3456090651558075e-06,
+      "loss": 1.5011,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 53.35713779003586,
+      "learning_rate": 6.373937677053825e-06,
+      "loss": 1.4962,
+      "step": 1125
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 44.931444706391076,
+      "learning_rate": 6.402266288951842e-06,
+      "loss": 1.511,
+      "step": 1130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 22.44919456560898,
+      "learning_rate": 6.430594900849859e-06,
+      "loss": 1.5138,
+      "step": 1135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 45.08279033177284,
+      "learning_rate": 6.458923512747876e-06,
+      "loss": 1.4963,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.52813913674402,
+      "learning_rate": 6.487252124645893e-06,
+      "loss": 1.5131,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 26.26804383614154,
+      "learning_rate": 6.51558073654391e-06,
+      "loss": 1.5118,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 38.38984295804615,
+      "learning_rate": 6.543909348441927e-06,
+      "loss": 1.5052,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 62.89163803243198,
+      "learning_rate": 6.572237960339945e-06,
+      "loss": 1.4908,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.43464761001089,
+      "learning_rate": 6.600566572237961e-06,
+      "loss": 1.4863,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 27.171308819606647,
+      "learning_rate": 6.628895184135978e-06,
+      "loss": 1.4866,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.858618384000161,
+      "learning_rate": 6.657223796033995e-06,
+      "loss": 1.4997,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 26.73384894098579,
+      "learning_rate": 6.685552407932012e-06,
+      "loss": 1.4724,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 21.318889353798717,
+      "learning_rate": 6.713881019830029e-06,
+      "loss": 1.4655,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 47.36154668479138,
+      "learning_rate": 6.742209631728046e-06,
+      "loss": 1.4677,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.312162357247926,
+      "learning_rate": 6.770538243626062e-06,
+      "loss": 1.487,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 28.90405260718444,
+      "learning_rate": 6.79886685552408e-06,
+      "loss": 1.4774,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 43.20968543413434,
+      "learning_rate": 6.827195467422096e-06,
+      "loss": 1.4592,
+      "step": 1205
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.41534301679069,
+      "learning_rate": 6.855524079320114e-06,
+      "loss": 1.4741,
+      "step": 1210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 39.00940865689802,
+      "learning_rate": 6.883852691218131e-06,
+      "loss": 1.476,
+      "step": 1215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.463235319018573,
+      "learning_rate": 6.912181303116148e-06,
+      "loss": 1.4989,
+      "step": 1220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.14813609317964,
+      "learning_rate": 6.940509915014165e-06,
+      "loss": 1.4772,
+      "step": 1225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 45.64032082296392,
+      "learning_rate": 6.968838526912182e-06,
+      "loss": 1.4859,
+      "step": 1230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.426783265938841,
+      "learning_rate": 6.997167138810199e-06,
+      "loss": 1.4732,
+      "step": 1235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 56.98259656288749,
+      "learning_rate": 7.025495750708215e-06,
+      "loss": 1.4717,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 73.22843429393423,
+      "learning_rate": 7.053824362606233e-06,
+      "loss": 1.4471,
+      "step": 1245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 25.573775068067942,
+      "learning_rate": 7.082152974504249e-06,
+      "loss": 1.4372,
+      "step": 1250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.443033292868687,
+      "learning_rate": 7.110481586402267e-06,
+      "loss": 1.4631,
+      "step": 1255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.668450041884629,
+      "learning_rate": 7.1388101983002834e-06,
+      "loss": 1.4368,
+      "step": 1260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 115.98207102359974,
+      "learning_rate": 7.167138810198301e-06,
+      "loss": 1.468,
+      "step": 1265
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 128.82110344511952,
+      "learning_rate": 7.195467422096318e-06,
+      "loss": 1.4917,
+      "step": 1270
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 31.41157580165412,
+      "learning_rate": 7.223796033994335e-06,
+      "loss": 1.4766,
+      "step": 1275
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 33.117081868197246,
+      "learning_rate": 7.2521246458923525e-06,
+      "loss": 1.4506,
+      "step": 1280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 33.6257926561511,
+      "learning_rate": 7.280453257790368e-06,
+      "loss": 1.4871,
+      "step": 1285
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.68056867252948,
+      "learning_rate": 7.308781869688386e-06,
+      "loss": 1.4495,
+      "step": 1290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 76.32432258998954,
+      "learning_rate": 7.337110481586402e-06,
+      "loss": 1.4476,
+      "step": 1295
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 73.51943278047416,
+      "learning_rate": 7.36543909348442e-06,
+      "loss": 1.4601,
+      "step": 1300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 35.92091908157174,
+      "learning_rate": 7.3937677053824365e-06,
+      "loss": 1.4612,
+      "step": 1305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 37.15133392906697,
+      "learning_rate": 7.422096317280454e-06,
+      "loss": 1.4608,
+      "step": 1310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 54.24887555884184,
+      "learning_rate": 7.4504249291784715e-06,
+      "loss": 1.4471,
+      "step": 1315
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 66.61085194888463,
+      "learning_rate": 7.478753541076488e-06,
+      "loss": 1.4306,
+      "step": 1320
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 67.55198055290484,
+      "learning_rate": 7.507082152974506e-06,
+      "loss": 1.4294,
+      "step": 1325
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 60.83568936526317,
+      "learning_rate": 7.535410764872521e-06,
+      "loss": 1.4426,
+      "step": 1330
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 64.35303855984785,
+      "learning_rate": 7.563739376770539e-06,
+      "loss": 1.4382,
+      "step": 1335
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 56.55071526619856,
+      "learning_rate": 7.5920679886685555e-06,
+      "loss": 1.4224,
+      "step": 1340
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 67.12284685088699,
+      "learning_rate": 7.620396600566573e-06,
+      "loss": 1.4321,
+      "step": 1345
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 48.395017715798886,
+      "learning_rate": 7.64872521246459e-06,
+      "loss": 1.4346,
+      "step": 1350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 38.088435110493364,
+      "learning_rate": 7.677053824362606e-06,
+      "loss": 1.434,
+      "step": 1355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.108791280938446,
+      "learning_rate": 7.705382436260623e-06,
+      "loss": 1.4467,
+      "step": 1360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 46.213258435941164,
+      "learning_rate": 7.733711048158641e-06,
+      "loss": 1.4726,
+      "step": 1365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 31.36163401132065,
+      "learning_rate": 7.762039660056658e-06,
+      "loss": 1.417,
+      "step": 1370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 21.57521345294901,
+      "learning_rate": 7.790368271954675e-06,
+      "loss": 1.4167,
+      "step": 1375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 25.684711958358132,
+      "learning_rate": 7.818696883852693e-06,
+      "loss": 1.4279,
+      "step": 1380
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 46.14025445942214,
+      "learning_rate": 7.847025495750708e-06,
+      "loss": 1.4367,
+      "step": 1385
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.837822707093785,
+      "learning_rate": 7.875354107648726e-06,
+      "loss": 1.4414,
+      "step": 1390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 25.32494554542278,
+      "learning_rate": 7.903682719546743e-06,
+      "loss": 1.4225,
+      "step": 1395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 22.096692157328395,
+      "learning_rate": 7.93201133144476e-06,
+      "loss": 1.4244,
+      "step": 1400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 23.085791295961783,
+      "learning_rate": 7.960339943342776e-06,
+      "loss": 1.4269,
+      "step": 1405
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 42.63437473811217,
+      "learning_rate": 7.988668555240794e-06,
+      "loss": 1.4216,
+      "step": 1410
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.581831841044835,
+      "learning_rate": 8.016997167138811e-06,
+      "loss": 1.4241,
+      "step": 1415
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 107.73424906023473,
+      "learning_rate": 8.045325779036828e-06,
+      "loss": 1.4305,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 102.80669799826326,
+      "learning_rate": 8.073654390934846e-06,
+      "loss": 1.4035,
+      "step": 1425
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 54.25200717068274,
+      "learning_rate": 8.101983002832861e-06,
+      "loss": 1.4146,
+      "step": 1430
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 70.28692545400818,
+      "learning_rate": 8.13031161473088e-06,
+      "loss": 1.4386,
+      "step": 1435
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 41.22226556175442,
+      "learning_rate": 8.158640226628896e-06,
+      "loss": 1.4142,
+      "step": 1440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 33.320500906922966,
+      "learning_rate": 8.186968838526912e-06,
+      "loss": 1.4063,
+      "step": 1445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.96946984665567,
+      "learning_rate": 8.215297450424929e-06,
+      "loss": 1.3873,
+      "step": 1450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 36.4133779105684,
+      "learning_rate": 8.243626062322947e-06,
+      "loss": 1.3998,
+      "step": 1455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.19446044964224,
+      "learning_rate": 8.271954674220964e-06,
+      "loss": 1.4123,
+      "step": 1460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.008839218394122,
+      "learning_rate": 8.30028328611898e-06,
+      "loss": 1.3916,
+      "step": 1465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.51211979323123,
+      "learning_rate": 8.328611898016999e-06,
+      "loss": 1.4045,
+      "step": 1470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 28.355806563412795,
+      "learning_rate": 8.356940509915014e-06,
+      "loss": 1.3989,
+      "step": 1475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 56.645997417423445,
+      "learning_rate": 8.385269121813032e-06,
+      "loss": 1.3892,
+      "step": 1480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 37.58733117483026,
+      "learning_rate": 8.413597733711049e-06,
+      "loss": 1.3983,
+      "step": 1485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 23.45675645372948,
+      "learning_rate": 8.441926345609066e-06,
+      "loss": 1.4338,
+      "step": 1490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 43.27659750878071,
+      "learning_rate": 8.470254957507082e-06,
+      "loss": 1.4018,
+      "step": 1495
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 26.338070551642947,
+      "learning_rate": 8.4985835694051e-06,
+      "loss": 1.3832,
+      "step": 1500
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 35.56121674532834,
+      "learning_rate": 8.526912181303117e-06,
+      "loss": 1.3637,
+      "step": 1505
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.858156584818255,
+      "learning_rate": 8.555240793201134e-06,
+      "loss": 1.3929,
+      "step": 1510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.860684083485195,
+      "learning_rate": 8.583569405099152e-06,
+      "loss": 1.3933,
+      "step": 1515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.67764945655357,
+      "learning_rate": 8.611898016997167e-06,
+      "loss": 1.3912,
+      "step": 1520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.069808432652248,
+      "learning_rate": 8.640226628895185e-06,
+      "loss": 1.3796,
+      "step": 1525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 65.37764259474062,
+      "learning_rate": 8.668555240793202e-06,
+      "loss": 1.3854,
+      "step": 1530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 67.37196707064975,
+      "learning_rate": 8.696883852691219e-06,
+      "loss": 1.3932,
+      "step": 1535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 26.53654854865448,
+      "learning_rate": 8.725212464589235e-06,
+      "loss": 1.3971,
+      "step": 1540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 120.33111697552754,
+      "learning_rate": 8.753541076487254e-06,
+      "loss": 1.3925,
+      "step": 1545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 26.033106458020956,
+      "learning_rate": 8.78186968838527e-06,
+      "loss": 1.3924,
+      "step": 1550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 127.2750601403478,
+      "learning_rate": 8.810198300283287e-06,
+      "loss": 1.4086,
+      "step": 1555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 97.35619667630274,
+      "learning_rate": 8.838526912181304e-06,
+      "loss": 1.4252,
+      "step": 1560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 41.5132287656772,
+      "learning_rate": 8.86685552407932e-06,
+      "loss": 1.4048,
+      "step": 1565
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.19314986884624,
+      "learning_rate": 8.895184135977339e-06,
+      "loss": 1.4018,
+      "step": 1570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.207399962119673,
+      "learning_rate": 8.923512747875355e-06,
+      "loss": 1.3928,
+      "step": 1575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.706236409485633,
+      "learning_rate": 8.951841359773372e-06,
+      "loss": 1.402,
+      "step": 1580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.15951416173616,
+      "learning_rate": 8.980169971671388e-06,
+      "loss": 1.3589,
+      "step": 1585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 19.959848444063436,
+      "learning_rate": 9.008498583569407e-06,
+      "loss": 1.3918,
+      "step": 1590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.03318135371038,
+      "learning_rate": 9.036827195467422e-06,
+      "loss": 1.3908,
+      "step": 1595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 59.051519885278694,
+      "learning_rate": 9.06515580736544e-06,
+      "loss": 1.3731,
+      "step": 1600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 52.467108769254416,
+      "learning_rate": 9.093484419263457e-06,
+      "loss": 1.3487,
+      "step": 1605
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 59.018311552807326,
+      "learning_rate": 9.121813031161473e-06,
+      "loss": 1.377,
+      "step": 1610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.959473760297636,
+      "learning_rate": 9.150141643059492e-06,
+      "loss": 1.4018,
+      "step": 1615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 35.688486380413146,
+      "learning_rate": 9.178470254957508e-06,
+      "loss": 1.3702,
+      "step": 1620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 17.702780585065685,
+      "learning_rate": 9.206798866855525e-06,
+      "loss": 1.3887,
+      "step": 1625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 44.95672001007261,
+      "learning_rate": 9.235127478753542e-06,
+      "loss": 1.3852,
+      "step": 1630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 23.805509750982903,
+      "learning_rate": 9.26345609065156e-06,
+      "loss": 1.3646,
+      "step": 1635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 39.33476110738001,
+      "learning_rate": 9.291784702549575e-06,
+      "loss": 1.3764,
+      "step": 1640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.540226055124654,
+      "learning_rate": 9.320113314447593e-06,
+      "loss": 1.3633,
+      "step": 1645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 60.937887345190425,
+      "learning_rate": 9.34844192634561e-06,
+      "loss": 1.3941,
+      "step": 1650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.183842224404986,
+      "learning_rate": 9.376770538243626e-06,
+      "loss": 1.3632,
+      "step": 1655
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.268626818286448,
+      "learning_rate": 9.405099150141643e-06,
+      "loss": 1.3632,
+      "step": 1660
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.977488338411256,
+      "learning_rate": 9.433427762039661e-06,
+      "loss": 1.3437,
+      "step": 1665
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 17.69425703983086,
+      "learning_rate": 9.461756373937678e-06,
+      "loss": 1.3407,
+      "step": 1670
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 50.81269776992468,
+      "learning_rate": 9.490084985835695e-06,
+      "loss": 1.3649,
+      "step": 1675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 52.987834036106406,
+      "learning_rate": 9.518413597733713e-06,
+      "loss": 1.368,
+      "step": 1680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 53.09975091644179,
+      "learning_rate": 9.546742209631728e-06,
+      "loss": 1.3753,
+      "step": 1685
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 65.42753190343696,
+      "learning_rate": 9.575070821529746e-06,
+      "loss": 1.3373,
+      "step": 1690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.50878001715189,
+      "learning_rate": 9.603399433427763e-06,
+      "loss": 1.3204,
+      "step": 1695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 51.46300322685196,
+      "learning_rate": 9.63172804532578e-06,
+      "loss": 1.3517,
+      "step": 1700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 50.08389878971014,
+      "learning_rate": 9.660056657223796e-06,
+      "loss": 1.3377,
+      "step": 1705
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 98.7501586904867,
+      "learning_rate": 9.688385269121814e-06,
+      "loss": 1.358,
+      "step": 1710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 25.256204978294605,
+      "learning_rate": 9.716713881019831e-06,
+      "loss": 1.3391,
+      "step": 1715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 96.05571532903846,
+      "learning_rate": 9.745042492917848e-06,
+      "loss": 1.3615,
+      "step": 1720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 26.755333256803034,
+      "learning_rate": 9.773371104815866e-06,
+      "loss": 1.3662,
+      "step": 1725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 25.837535923418997,
+      "learning_rate": 9.801699716713881e-06,
+      "loss": 1.38,
+      "step": 1730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 24.738456192454084,
+      "learning_rate": 9.8300283286119e-06,
+      "loss": 1.333,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 41.26197817069162,
+      "learning_rate": 9.858356940509916e-06,
+      "loss": 1.3491,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 51.99630683152431,
+      "learning_rate": 9.886685552407933e-06,
+      "loss": 1.3429,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 46.34281673405669,
+      "learning_rate": 9.91501416430595e-06,
+      "loss": 1.3646,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 22.857745511222603,
+      "learning_rate": 9.943342776203968e-06,
+      "loss": 1.3297,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.712863491665592,
+      "learning_rate": 9.971671388101982e-06,
+      "loss": 1.321,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.7261894236906,
+      "learning_rate": 1e-05,
+      "loss": 1.3444,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 32.60283769493224,
+      "learning_rate": 9.99999755418257e-06,
+      "loss": 1.3408,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.203512310970503,
+      "learning_rate": 9.999990216732668e-06,
+      "loss": 1.3446,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.883185421539476,
+      "learning_rate": 9.999977987657479e-06,
+      "loss": 1.3313,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 19.323697826239748,
+      "learning_rate": 9.99996086696896e-06,
+      "loss": 1.3503,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 27.72449057867512,
+      "learning_rate": 9.999938854683867e-06,
+      "loss": 1.3201,
+      "step": 1790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 35.22723944266936,
+      "learning_rate": 9.99991195082373e-06,
+      "loss": 1.3395,
+      "step": 1795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 36.374853638958115,
+      "learning_rate": 9.999880155414872e-06,
+      "loss": 1.34,
+      "step": 1800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 29.11298844229375,
+      "learning_rate": 9.9998434684884e-06,
+      "loss": 1.3075,
+      "step": 1805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.978651122935727,
+      "learning_rate": 9.999801890080203e-06,
+      "loss": 1.3248,
+      "step": 1810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 62.277032520069206,
+      "learning_rate": 9.999755420230964e-06,
+      "loss": 1.3174,
+      "step": 1815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 42.25047939113886,
+      "learning_rate": 9.999704058986139e-06,
+      "loss": 1.3196,
+      "step": 1820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 121.34773754544872,
+      "learning_rate": 9.99964780639598e-06,
+      "loss": 1.3449,
+      "step": 1825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 157.875932629027,
+      "learning_rate": 9.999586662515519e-06,
+      "loss": 1.3766,
+      "step": 1830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 67.21735658366185,
+      "learning_rate": 9.999520627404576e-06,
+      "loss": 1.3336,
+      "step": 1835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 40.32573656282714,
+      "learning_rate": 9.999449701127753e-06,
+      "loss": 1.3358,
+      "step": 1840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 16.578779780453953,
+      "learning_rate": 9.999373883754442e-06,
+      "loss": 1.3263,
+      "step": 1845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 45.13578422606087,
+      "learning_rate": 9.999293175358814e-06,
+      "loss": 1.3156,
+      "step": 1850
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 25.271194035753734,
+      "learning_rate": 9.99920757601983e-06,
+      "loss": 1.3158,
+      "step": 1855
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 39.94426166106633,
+      "learning_rate": 9.999117085821233e-06,
+      "loss": 1.323,
+      "step": 1860
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 30.457883465258536,
+      "learning_rate": 9.999021704851555e-06,
+      "loss": 1.3516,
+      "step": 1865
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 36.9726912711344,
+      "learning_rate": 9.998921433204106e-06,
+      "loss": 1.333,
+      "step": 1870
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 23.030928072688916,
+      "learning_rate": 9.998816270976985e-06,
+      "loss": 1.2991,
+      "step": 1875
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 47.126156393517775,
+      "learning_rate": 9.998706218273078e-06,
+      "loss": 1.2744,
+      "step": 1880
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 32.59637028245757,
+      "learning_rate": 9.998591275200049e-06,
+      "loss": 1.3096,
+      "step": 1885
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 23.94926147678152,
+      "learning_rate": 9.998471441870353e-06,
+      "loss": 1.2996,
+      "step": 1890
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 21.030622870893925,
+      "learning_rate": 9.998346718401222e-06,
+      "loss": 1.3209,
+      "step": 1895
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.715711453727344,
+      "learning_rate": 9.998217104914683e-06,
+      "loss": 1.3006,
+      "step": 1900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.08367409477122,
+      "learning_rate": 9.998082601537533e-06,
+      "loss": 1.3012,
+      "step": 1905
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.183410345778563,
+      "learning_rate": 9.997943208401365e-06,
+      "loss": 1.3134,
+      "step": 1910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.086410225689482,
+      "learning_rate": 9.99779892564255e-06,
+      "loss": 1.2904,
+      "step": 1915
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 37.16067338584074,
+      "learning_rate": 9.997649753402243e-06,
+      "loss": 1.3047,
+      "step": 1920
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 40.474695223414116,
+      "learning_rate": 9.997495691826382e-06,
+      "loss": 1.2886,
+      "step": 1925
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 53.203394356059604,
+      "learning_rate": 9.997336741065694e-06,
+      "loss": 1.2939,
+      "step": 1930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 43.19777784215109,
+      "learning_rate": 9.997172901275682e-06,
+      "loss": 1.2875,
+      "step": 1935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 65.5795185045484,
+      "learning_rate": 9.997004172616633e-06,
+      "loss": 1.3288,
+      "step": 1940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 50.496217939819,
+      "learning_rate": 9.996830555253622e-06,
+      "loss": 1.3044,
+      "step": 1945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.159352252412273,
+      "learning_rate": 9.9966520493565e-06,
+      "loss": 1.3244,
+      "step": 1950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.449411392080844,
+      "learning_rate": 9.99646865509991e-06,
+      "loss": 1.3067,
+      "step": 1955
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 30.85859582159372,
+      "learning_rate": 9.996280372663266e-06,
+      "loss": 1.2854,
+      "step": 1960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.954831400149821,
+      "learning_rate": 9.996087202230773e-06,
+      "loss": 1.2711,
+      "step": 1965
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.535102116165461,
+      "learning_rate": 9.995889143991412e-06,
+      "loss": 1.2758,
+      "step": 1970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.05835385978514,
+      "learning_rate": 9.995686198138951e-06,
+      "loss": 1.2838,
+      "step": 1975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 27.158362157021386,
+      "learning_rate": 9.995478364871937e-06,
+      "loss": 1.2848,
+      "step": 1980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.607677347887147,
+      "learning_rate": 9.995265644393698e-06,
+      "loss": 1.3077,
+      "step": 1985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.019777361051082,
+      "learning_rate": 9.995048036912345e-06,
+      "loss": 1.2911,
+      "step": 1990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 36.439429993916605,
+      "learning_rate": 9.99482554264077e-06,
+      "loss": 1.2909,
+      "step": 1995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.316454115512972,
+      "learning_rate": 9.994598161796643e-06,
+      "loss": 1.28,
+      "step": 2000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.451990819041521,
+      "learning_rate": 9.994365894602419e-06,
+      "loss": 1.2796,
+      "step": 2005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.039745793623432,
+      "learning_rate": 9.994128741285329e-06,
+      "loss": 1.2776,
+      "step": 2010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 34.91032967270379,
+      "learning_rate": 9.99388670207739e-06,
+      "loss": 1.2953,
+      "step": 2015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 30.505391467678226,
+      "learning_rate": 9.99363977721539e-06,
+      "loss": 1.2912,
+      "step": 2020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 70.60877015993087,
+      "learning_rate": 9.993387966940908e-06,
+      "loss": 1.3002,
+      "step": 2025
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 42.076694141065936,
+      "learning_rate": 9.993131271500293e-06,
+      "loss": 1.2742,
+      "step": 2030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 40.78685017799231,
+      "learning_rate": 9.992869691144679e-06,
+      "loss": 1.2764,
+      "step": 2035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 33.41521506875986,
+      "learning_rate": 9.992603226129978e-06,
+      "loss": 1.2883,
+      "step": 2040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.665923005413482,
+      "learning_rate": 9.992331876716877e-06,
+      "loss": 1.269,
+      "step": 2045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 35.13745352099027,
+      "learning_rate": 9.992055643170844e-06,
+      "loss": 1.2919,
+      "step": 2050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 93.57018198555726,
+      "learning_rate": 9.99177452576213e-06,
+      "loss": 1.2477,
+      "step": 2055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 73.14188594561246,
+      "learning_rate": 9.991488524765755e-06,
+      "loss": 1.2727,
+      "step": 2060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 60.954485489252725,
+      "learning_rate": 9.991197640461527e-06,
+      "loss": 1.2804,
+      "step": 2065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 49.16627669625653,
+      "learning_rate": 9.99090187313402e-06,
+      "loss": 1.2764,
+      "step": 2070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 71.11546717627556,
+      "learning_rate": 9.990601223072596e-06,
+      "loss": 1.2999,
+      "step": 2075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.860674350454783,
+      "learning_rate": 9.990295690571386e-06,
+      "loss": 1.2745,
+      "step": 2080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 30.0270032458091,
+      "learning_rate": 9.989985275929302e-06,
+      "loss": 1.2672,
+      "step": 2085
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 48.4403765685246,
+      "learning_rate": 9.98966997945003e-06,
+      "loss": 1.2713,
+      "step": 2090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.184895143802937,
+      "learning_rate": 9.989349801442034e-06,
+      "loss": 1.2447,
+      "step": 2095
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.326803432440844,
+      "learning_rate": 9.989024742218554e-06,
+      "loss": 1.2718,
+      "step": 2100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 22.212522451101893,
+      "learning_rate": 9.9886948020976e-06,
+      "loss": 1.2692,
+      "step": 2105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.955909550882618,
+      "learning_rate": 9.988359981401967e-06,
+      "loss": 1.2632,
+      "step": 2110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 54.16029859047456,
+      "learning_rate": 9.988020280459214e-06,
+      "loss": 1.2474,
+      "step": 2115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.99762681137852,
+      "learning_rate": 9.987675699601682e-06,
+      "loss": 1.2878,
+      "step": 2120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.986794193093745,
+      "learning_rate": 9.987326239166484e-06,
+      "loss": 1.2581,
+      "step": 2125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 41.18160853788803,
+      "learning_rate": 9.986971899495505e-06,
+      "loss": 1.2505,
+      "step": 2130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 22.73713386704341,
+      "learning_rate": 9.986612680935409e-06,
+      "loss": 1.2369,
+      "step": 2135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 39.94586605054518,
+      "learning_rate": 9.986248583837624e-06,
+      "loss": 1.2506,
+      "step": 2140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 68.00261940817953,
+      "learning_rate": 9.985879608558359e-06,
+      "loss": 1.2386,
+      "step": 2145
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 25.278427620899553,
+      "learning_rate": 9.98550575545859e-06,
+      "loss": 1.2315,
+      "step": 2150
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 58.48992267362782,
+      "learning_rate": 9.985127024904072e-06,
+      "loss": 1.2684,
+      "step": 2155
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 107.22278634371908,
+      "learning_rate": 9.984743417265323e-06,
+      "loss": 1.261,
+      "step": 2160
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 56.43898177359786,
+      "learning_rate": 9.984354932917639e-06,
+      "loss": 1.2578,
+      "step": 2165
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 30.02405317886174,
+      "learning_rate": 9.983961572241085e-06,
+      "loss": 1.221,
+      "step": 2170
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 21.331779746602443,
+      "learning_rate": 9.983563335620494e-06,
+      "loss": 1.2405,
+      "step": 2175
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 43.31147042485862,
+      "learning_rate": 9.983160223445476e-06,
+      "loss": 1.2366,
+      "step": 2180
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.029017686861732,
+      "learning_rate": 9.982752236110401e-06,
+      "loss": 1.2432,
+      "step": 2185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.894536079070498,
+      "learning_rate": 9.982339374014416e-06,
+      "loss": 1.2213,
+      "step": 2190
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 52.58513573855897,
+      "learning_rate": 9.981921637561438e-06,
+      "loss": 1.2446,
+      "step": 2195
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 148.5603971500644,
+      "learning_rate": 9.981499027160147e-06,
+      "loss": 1.2624,
+      "step": 2200
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 67.61072411459752,
+      "learning_rate": 9.981071543223992e-06,
+      "loss": 1.2677,
+      "step": 2205
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 54.65073271042123,
+      "learning_rate": 9.980639186171198e-06,
+      "loss": 1.2462,
+      "step": 2210
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 38.85335665696665,
+      "learning_rate": 9.980201956424748e-06,
+      "loss": 1.2708,
+      "step": 2215
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.33642147794773,
+      "learning_rate": 9.979759854412395e-06,
+      "loss": 1.2643,
+      "step": 2220
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.980103706134027,
+      "learning_rate": 9.97931288056666e-06,
+      "loss": 1.2406,
+      "step": 2225
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.425628724046566,
+      "learning_rate": 9.978861035324832e-06,
+      "loss": 1.217,
+      "step": 2230
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 46.170020521427176,
+      "learning_rate": 9.97840431912896e-06,
+      "loss": 1.24,
+      "step": 2235
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 20.23431894994783,
+      "learning_rate": 9.977942732425862e-06,
+      "loss": 1.2447,
+      "step": 2240
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 29.45427587924372,
+      "learning_rate": 9.977476275667123e-06,
+      "loss": 1.2285,
+      "step": 2245
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.476899810487403,
+      "learning_rate": 9.97700494930909e-06,
+      "loss": 1.1955,
+      "step": 2250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.56756736605288,
+      "learning_rate": 9.976528753812874e-06,
+      "loss": 1.2215,
+      "step": 2255
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.963469861084638,
+      "learning_rate": 9.976047689644345e-06,
+      "loss": 1.2168,
+      "step": 2260
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 37.7498655059393,
+      "learning_rate": 9.97556175727415e-06,
+      "loss": 1.233,
+      "step": 2265
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.34730041668875,
+      "learning_rate": 9.975070957177681e-06,
+      "loss": 1.2258,
+      "step": 2270
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 29.797887323807636,
+      "learning_rate": 9.974575289835107e-06,
+      "loss": 1.2487,
+      "step": 2275
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.355285431034126,
+      "learning_rate": 9.974074755731351e-06,
+      "loss": 1.2322,
+      "step": 2280
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.54182670812362,
+      "learning_rate": 9.973569355356099e-06,
+      "loss": 1.2535,
+      "step": 2285
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.753113355014868,
+      "learning_rate": 9.973059089203797e-06,
+      "loss": 1.2257,
+      "step": 2290
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.018327238321696,
+      "learning_rate": 9.972543957773653e-06,
+      "loss": 1.2411,
+      "step": 2295
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.70691445466828,
+      "learning_rate": 9.972023961569632e-06,
+      "loss": 1.226,
+      "step": 2300
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 40.42682718423413,
+      "learning_rate": 9.971499101100463e-06,
+      "loss": 1.2181,
+      "step": 2305
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 51.091493759604035,
+      "learning_rate": 9.97096937687963e-06,
+      "loss": 1.2247,
+      "step": 2310
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 48.195350237791374,
+      "learning_rate": 9.970434789425378e-06,
+      "loss": 1.2125,
+      "step": 2315
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.38647273683228,
+      "learning_rate": 9.969895339260706e-06,
+      "loss": 1.2069,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 30.49139360025075,
+      "learning_rate": 9.969351026913375e-06,
+      "loss": 1.2194,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.385511153165469,
+      "learning_rate": 9.968801852915897e-06,
+      "loss": 1.2209,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.600235098601413,
+      "learning_rate": 9.968247817805548e-06,
+      "loss": 1.233,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 25.51632451097952,
+      "learning_rate": 9.967688922124351e-06,
+      "loss": 1.2234,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.458533234384973,
+      "learning_rate": 9.967125166419092e-06,
+      "loss": 1.2337,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.137901496985272,
+      "learning_rate": 9.966556551241307e-06,
+      "loss": 1.2252,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.554021202381733,
+      "learning_rate": 9.965983077147287e-06,
+      "loss": 1.232,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 29.255477281363408,
+      "learning_rate": 9.965404744698081e-06,
+      "loss": 1.223,
+      "step": 2360
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 55.71066517356006,
+      "learning_rate": 9.964821554459482e-06,
+      "loss": 1.211,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 46.97293908156358,
+      "learning_rate": 9.964233507002044e-06,
+      "loss": 1.2306,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 29.82807675197631,
+      "learning_rate": 9.963640602901069e-06,
+      "loss": 1.1983,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 40.8106865600396,
+      "learning_rate": 9.96304284273661e-06,
+      "loss": 1.2362,
+      "step": 2380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 37.035382402104936,
+      "learning_rate": 9.962440227093474e-06,
+      "loss": 1.2233,
+      "step": 2385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 47.48410746333308,
+      "learning_rate": 9.961832756561213e-06,
+      "loss": 1.2354,
+      "step": 2390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.34844898900087,
+      "learning_rate": 9.961220431734137e-06,
+      "loss": 1.2215,
+      "step": 2395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 108.59661182177989,
+      "learning_rate": 9.960603253211295e-06,
+      "loss": 1.2327,
+      "step": 2400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 76.44086594519507,
+      "learning_rate": 9.95998122159649e-06,
+      "loss": 1.2229,
+      "step": 2405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 38.53725294367486,
+      "learning_rate": 9.959354337498274e-06,
+      "loss": 1.2166,
+      "step": 2410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.492069683488847,
+      "learning_rate": 9.958722601529945e-06,
+      "loss": 1.2349,
+      "step": 2415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 32.42759975420263,
+      "learning_rate": 9.958086014309545e-06,
+      "loss": 1.212,
+      "step": 2420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 19.57195181578406,
+      "learning_rate": 9.957444576459864e-06,
+      "loss": 1.2193,
+      "step": 2425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 26.68571032098248,
+      "learning_rate": 9.956798288608442e-06,
+      "loss": 1.2197,
+      "step": 2430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.360805598503173,
+      "learning_rate": 9.956147151387559e-06,
+      "loss": 1.233,
+      "step": 2435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.128646502976743,
+      "learning_rate": 9.955491165434236e-06,
+      "loss": 1.2141,
+      "step": 2440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 26.658087544079972,
+      "learning_rate": 9.954830331390245e-06,
+      "loss": 1.1987,
+      "step": 2445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.065272609028908,
+      "learning_rate": 9.954164649902097e-06,
+      "loss": 1.192,
+      "step": 2450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.279093623666416,
+      "learning_rate": 9.953494121621047e-06,
+      "loss": 1.2053,
+      "step": 2455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 24.282084124178198,
+      "learning_rate": 9.95281874720309e-06,
+      "loss": 1.1988,
+      "step": 2460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 44.50488285966649,
+      "learning_rate": 9.952138527308963e-06,
+      "loss": 1.208,
+      "step": 2465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 33.79862535916042,
+      "learning_rate": 9.951453462604143e-06,
+      "loss": 1.1916,
+      "step": 2470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 27.13073772350017,
+      "learning_rate": 9.950763553758848e-06,
+      "loss": 1.21,
+      "step": 2475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 49.80064601585301,
+      "learning_rate": 9.950068801448037e-06,
+      "loss": 1.2042,
+      "step": 2480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 66.21016098075907,
+      "learning_rate": 9.9493692063514e-06,
+      "loss": 1.2083,
+      "step": 2485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 26.587068635738756,
+      "learning_rate": 9.948664769153372e-06,
+      "loss": 1.2169,
+      "step": 2490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.940254770635514,
+      "learning_rate": 9.947955490543125e-06,
+      "loss": 1.197,
+      "step": 2495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.998827359689898,
+      "learning_rate": 9.94724137121456e-06,
+      "loss": 1.1862,
+      "step": 2500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 41.25760329284341,
+      "learning_rate": 9.946522411866325e-06,
+      "loss": 1.1847,
+      "step": 2505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 133.01481338781392,
+      "learning_rate": 9.945798613201794e-06,
+      "loss": 1.2163,
+      "step": 2510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 89.31490352272507,
+      "learning_rate": 9.94506997592908e-06,
+      "loss": 1.177,
+      "step": 2515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 30.656395841990992,
+      "learning_rate": 9.944336500761029e-06,
+      "loss": 1.1902,
+      "step": 2520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 40.71404633836761,
+      "learning_rate": 9.943598188415217e-06,
+      "loss": 1.1812,
+      "step": 2525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 44.483429975247155,
+      "learning_rate": 9.942855039613958e-06,
+      "loss": 1.2245,
+      "step": 2530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 59.00601624439954,
+      "learning_rate": 9.942107055084292e-06,
+      "loss": 1.1852,
+      "step": 2535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 48.899022645872606,
+      "learning_rate": 9.941354235557994e-06,
+      "loss": 1.1838,
+      "step": 2540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.207646488335833,
+      "learning_rate": 9.940596581771568e-06,
+      "loss": 1.2067,
+      "step": 2545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.145055187749728,
+      "learning_rate": 9.939834094466245e-06,
+      "loss": 1.2037,
+      "step": 2550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.3637903577066,
+      "learning_rate": 9.939066774387988e-06,
+      "loss": 1.2061,
+      "step": 2555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 23.27972184509164,
+      "learning_rate": 9.93829462228749e-06,
+      "loss": 1.1862,
+      "step": 2560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.504767354168685,
+      "learning_rate": 9.937517638920162e-06,
+      "loss": 1.1787,
+      "step": 2565
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.882534891848458,
+      "learning_rate": 9.936735825046154e-06,
+      "loss": 1.1721,
+      "step": 2570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 47.85662609246532,
+      "learning_rate": 9.93594918143033e-06,
+      "loss": 1.1926,
+      "step": 2575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 80.08678346889124,
+      "learning_rate": 9.935157708842288e-06,
+      "loss": 1.1823,
+      "step": 2580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 28.614889713910266,
+      "learning_rate": 9.934361408056348e-06,
+      "loss": 1.1815,
+      "step": 2585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 53.20030002856704,
+      "learning_rate": 9.933560279851549e-06,
+      "loss": 1.2112,
+      "step": 2590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 56.59970823556055,
+      "learning_rate": 9.93275432501166e-06,
+      "loss": 1.1882,
+      "step": 2595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 43.98644204083481,
+      "learning_rate": 9.931943544325166e-06,
+      "loss": 1.16,
+      "step": 2600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.92441460995244,
+      "learning_rate": 9.931127938585275e-06,
+      "loss": 1.2017,
+      "step": 2605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 73.0494618190435,
+      "learning_rate": 9.930307508589918e-06,
+      "loss": 1.1923,
+      "step": 2610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 66.41961786413435,
+      "learning_rate": 9.929482255141744e-06,
+      "loss": 1.1775,
+      "step": 2615
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.458880822215225,
+      "learning_rate": 9.928652179048119e-06,
+      "loss": 1.1606,
+      "step": 2620
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.88590996455315,
+      "learning_rate": 9.927817281121131e-06,
+      "loss": 1.1666,
+      "step": 2625
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 29.09234292628808,
+      "learning_rate": 9.926977562177582e-06,
+      "loss": 1.173,
+      "step": 2630
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 34.55403219807346,
+      "learning_rate": 9.926133023038988e-06,
+      "loss": 1.184,
+      "step": 2635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.720705771918542,
+      "learning_rate": 9.92528366453159e-06,
+      "loss": 1.1818,
+      "step": 2640
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.388664176736826,
+      "learning_rate": 9.924429487486339e-06,
+      "loss": 1.1724,
+      "step": 2645
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.130272229112684,
+      "learning_rate": 9.923570492738895e-06,
+      "loss": 1.1772,
+      "step": 2650
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.637369376811128,
+      "learning_rate": 9.922706681129634e-06,
+      "loss": 1.1873,
+      "step": 2655
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 22.481508332457953,
+      "learning_rate": 9.921838053503654e-06,
+      "loss": 1.1841,
+      "step": 2660
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 36.96789134182856,
+      "learning_rate": 9.92096461071075e-06,
+      "loss": 1.1646,
+      "step": 2665
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.20424827349328,
+      "learning_rate": 9.920086353605439e-06,
+      "loss": 1.1683,
+      "step": 2670
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 22.984932609600577,
+      "learning_rate": 9.919203283046942e-06,
+      "loss": 1.1832,
+      "step": 2675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.123529916442433,
+      "learning_rate": 9.918315399899188e-06,
+      "loss": 1.1692,
+      "step": 2680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.409240507891948,
+      "learning_rate": 9.91742270503082e-06,
+      "loss": 1.1767,
+      "step": 2685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.592745764863913,
+      "learning_rate": 9.916525199315186e-06,
+      "loss": 1.1824,
+      "step": 2690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 35.317900020424446,
+      "learning_rate": 9.91562288363034e-06,
+      "loss": 1.1766,
+      "step": 2695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.418561626974661,
+      "learning_rate": 9.91471575885904e-06,
+      "loss": 1.1722,
+      "step": 2700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.186458826087673,
+      "learning_rate": 9.91380382588875e-06,
+      "loss": 1.1511,
+      "step": 2705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.392876468515981,
+      "learning_rate": 9.912887085611642e-06,
+      "loss": 1.1582,
+      "step": 2710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.027146893897786,
+      "learning_rate": 9.911965538924584e-06,
+      "loss": 1.2027,
+      "step": 2715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.74651860266453,
+      "learning_rate": 9.911039186729154e-06,
+      "loss": 1.1886,
+      "step": 2720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.629032974183286,
+      "learning_rate": 9.910108029931622e-06,
+      "loss": 1.1849,
+      "step": 2725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.341384937928606,
+      "learning_rate": 9.90917206944297e-06,
+      "loss": 1.1755,
+      "step": 2730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 35.88736400267985,
+      "learning_rate": 9.908231306178869e-06,
+      "loss": 1.1932,
+      "step": 2735
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.536749553802284,
+      "learning_rate": 9.907285741059694e-06,
+      "loss": 1.1806,
+      "step": 2740
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.12273250818116,
+      "learning_rate": 9.906335375010518e-06,
+      "loss": 1.178,
+      "step": 2745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 28.476959265452525,
+      "learning_rate": 9.905380208961109e-06,
+      "loss": 1.1809,
+      "step": 2750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.670220119269143,
+      "learning_rate": 9.904420243845931e-06,
+      "loss": 1.1638,
+      "step": 2755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 66.52647296720023,
+      "learning_rate": 9.903455480604144e-06,
+      "loss": 1.1785,
+      "step": 2760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 65.14747317513056,
+      "learning_rate": 9.902485920179605e-06,
+      "loss": 1.1931,
+      "step": 2765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 35.62003805602506,
+      "learning_rate": 9.901511563520855e-06,
+      "loss": 1.1614,
+      "step": 2770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 39.14685246448072,
+      "learning_rate": 9.900532411581141e-06,
+      "loss": 1.1755,
+      "step": 2775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 89.34481105669401,
+      "learning_rate": 9.899548465318387e-06,
+      "loss": 1.1767,
+      "step": 2780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 52.59479557413615,
+      "learning_rate": 9.898559725695217e-06,
+      "loss": 1.1697,
+      "step": 2785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 35.276525624359245,
+      "learning_rate": 9.89756619367894e-06,
+      "loss": 1.1632,
+      "step": 2790
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.966535745499472,
+      "learning_rate": 9.89656787024156e-06,
+      "loss": 1.16,
+      "step": 2795
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.569210509013814,
+      "learning_rate": 9.895564756359758e-06,
+      "loss": 1.1749,
+      "step": 2800
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.739276355409753,
+      "learning_rate": 9.89455685301491e-06,
+      "loss": 1.1714,
+      "step": 2805
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 59.13841059108984,
+      "learning_rate": 9.893544161193077e-06,
+      "loss": 1.1816,
+      "step": 2810
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 24.5159909442423,
+      "learning_rate": 9.892526681884997e-06,
+      "loss": 1.1617,
+      "step": 2815
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 53.43412207505383,
+      "learning_rate": 9.891504416086102e-06,
+      "loss": 1.1698,
+      "step": 2820
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.028092300763062,
+      "learning_rate": 9.890477364796502e-06,
+      "loss": 1.1462,
+      "step": 2825
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.266849451796633,
+      "learning_rate": 9.889445529020989e-06,
+      "loss": 1.1739,
+      "step": 2830
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 33.256972593357,
+      "learning_rate": 9.888408909769032e-06,
+      "loss": 1.1638,
+      "step": 2835
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.240389968680208,
+      "learning_rate": 9.887367508054788e-06,
+      "loss": 1.1539,
+      "step": 2840
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 38.32149327567404,
+      "learning_rate": 9.886321324897088e-06,
+      "loss": 1.1453,
+      "step": 2845
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.92159888176316,
+      "learning_rate": 9.885270361319439e-06,
+      "loss": 1.129,
+      "step": 2850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 51.068604121950585,
+      "learning_rate": 9.884214618350028e-06,
+      "loss": 1.1633,
+      "step": 2855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 74.99666351301994,
+      "learning_rate": 9.883154097021717e-06,
+      "loss": 1.158,
+      "step": 2860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 30.31376407102048,
+      "learning_rate": 9.882088798372043e-06,
+      "loss": 1.1438,
+      "step": 2865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 27.670642676177405,
+      "learning_rate": 9.881018723443214e-06,
+      "loss": 1.1429,
+      "step": 2870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 23.64857996752343,
+      "learning_rate": 9.879943873282116e-06,
+      "loss": 1.1514,
+      "step": 2875
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.005128955967104,
+      "learning_rate": 9.878864248940304e-06,
+      "loss": 1.1595,
+      "step": 2880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.966016821913678,
+      "learning_rate": 9.877779851474003e-06,
+      "loss": 1.1417,
+      "step": 2885
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.446956221399299,
+      "learning_rate": 9.876690681944107e-06,
+      "loss": 1.1525,
+      "step": 2890
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.693284053412812,
+      "learning_rate": 9.87559674141618e-06,
+      "loss": 1.1077,
+      "step": 2895
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.179007907357477,
+      "learning_rate": 9.874498030960455e-06,
+      "loss": 1.1408,
+      "step": 2900
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 33.083858737896215,
+      "learning_rate": 9.87339455165183e-06,
+      "loss": 1.1454,
+      "step": 2905
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 79.29658401277891,
+      "learning_rate": 9.872286304569867e-06,
+      "loss": 1.1322,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.385160626134113,
+      "learning_rate": 9.871173290798795e-06,
+      "loss": 1.1611,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.773144522385966,
+      "learning_rate": 9.870055511427507e-06,
+      "loss": 1.1643,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 16.312322239246836,
+      "learning_rate": 9.868932967549554e-06,
+      "loss": 1.1377,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.14049593920567,
+      "learning_rate": 9.867805660263152e-06,
+      "loss": 1.1327,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 19.77386440226736,
+      "learning_rate": 9.866673590671176e-06,
+      "loss": 1.1358,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 30.710807933209413,
+      "learning_rate": 9.86553675988116e-06,
+      "loss": 1.1293,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.894878262098294,
+      "learning_rate": 9.864395169005297e-06,
+      "loss": 1.1428,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 60.79458334408716,
+      "learning_rate": 9.863248819160436e-06,
+      "loss": 1.1302,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 123.22513760731647,
+      "learning_rate": 9.862097711468082e-06,
+      "loss": 1.1408,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.980769182017495,
+      "learning_rate": 9.860941847054394e-06,
+      "loss": 1.176,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 77.79656950182576,
+      "learning_rate": 9.859781227050186e-06,
+      "loss": 1.1614,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 45.821558359257104,
+      "learning_rate": 9.858615852590923e-06,
+      "loss": 1.1402,
+      "step": 2970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.431782798306614,
+      "learning_rate": 9.857445724816723e-06,
+      "loss": 1.1479,
+      "step": 2975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 18.816255936865193,
+      "learning_rate": 9.856270844872355e-06,
+      "loss": 1.1414,
+      "step": 2980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.78799968669382,
+      "learning_rate": 9.855091213907233e-06,
+      "loss": 1.169,
+      "step": 2985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 20.618619597391554,
+      "learning_rate": 9.853906833075424e-06,
+      "loss": 1.1364,
+      "step": 2990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 29.602152847998962,
+      "learning_rate": 9.852717703535637e-06,
+      "loss": 1.1313,
+      "step": 2995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 20.41523192024887,
+      "learning_rate": 9.85152382645123e-06,
+      "loss": 1.13,
+      "step": 3000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 37.823239974778716,
+      "learning_rate": 9.85032520299021e-06,
+      "loss": 1.1508,
+      "step": 3005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 43.2473952972184,
+      "learning_rate": 9.849121834325216e-06,
+      "loss": 1.1316,
+      "step": 3010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 52.74014485034811,
+      "learning_rate": 9.847913721633541e-06,
+      "loss": 1.127,
+      "step": 3015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.225887741481749,
+      "learning_rate": 9.846700866097111e-06,
+      "loss": 1.1283,
+      "step": 3020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 22.830528962299184,
+      "learning_rate": 9.845483268902497e-06,
+      "loss": 1.1332,
+      "step": 3025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.962513124404307,
+      "learning_rate": 9.844260931240908e-06,
+      "loss": 1.1192,
+      "step": 3030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.075114486387971,
+      "learning_rate": 9.843033854308189e-06,
+      "loss": 1.1234,
+      "step": 3035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 43.92726259157927,
+      "learning_rate": 9.841802039304819e-06,
+      "loss": 1.1325,
+      "step": 3040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.602847875408965,
+      "learning_rate": 9.84056548743592e-06,
+      "loss": 1.0945,
+      "step": 3045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 72.84853493289299,
+      "learning_rate": 9.839324199911244e-06,
+      "loss": 1.1146,
+      "step": 3050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 64.59634241144452,
+      "learning_rate": 9.838078177945174e-06,
+      "loss": 1.1265,
+      "step": 3055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.606427097030371,
+      "learning_rate": 9.83682742275673e-06,
+      "loss": 1.1458,
+      "step": 3060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 22.74960141201855,
+      "learning_rate": 9.835571935569554e-06,
+      "loss": 1.1276,
+      "step": 3065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.872212344467762,
+      "learning_rate": 9.834311717611929e-06,
+      "loss": 1.1275,
+      "step": 3070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 26.848450408931033,
+      "learning_rate": 9.833046770116758e-06,
+      "loss": 1.1178,
+      "step": 3075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 39.22359304839263,
+      "learning_rate": 9.831777094321572e-06,
+      "loss": 1.1238,
+      "step": 3080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 38.21441517690827,
+      "learning_rate": 9.83050269146853e-06,
+      "loss": 1.13,
+      "step": 3085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 71.39821885265845,
+      "learning_rate": 9.829223562804415e-06,
+      "loss": 1.1276,
+      "step": 3090
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 69.47267188756749,
+      "learning_rate": 9.827939709580631e-06,
+      "loss": 1.1404,
+      "step": 3095
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 27.2120383573781,
+      "learning_rate": 9.82665113305321e-06,
+      "loss": 1.1328,
+      "step": 3100
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.3549164827145,
+      "learning_rate": 9.8253578344828e-06,
+      "loss": 1.0994,
+      "step": 3105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 26.975780544067167,
+      "learning_rate": 9.824059815134668e-06,
+      "loss": 1.114,
+      "step": 3110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.89034671963636,
+      "learning_rate": 9.822757076278701e-06,
+      "loss": 1.127,
+      "step": 3115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.784217590750867,
+      "learning_rate": 9.821449619189407e-06,
+      "loss": 1.1127,
+      "step": 3120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.672619503726025,
+      "learning_rate": 9.820137445145903e-06,
+      "loss": 1.1257,
+      "step": 3125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 33.283960901757666,
+      "learning_rate": 9.818820555431928e-06,
+      "loss": 1.1119,
+      "step": 3130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 20.205326596044547,
+      "learning_rate": 9.817498951335827e-06,
+      "loss": 1.1172,
+      "step": 3135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 38.647807030435,
+      "learning_rate": 9.81617263415056e-06,
+      "loss": 1.1332,
+      "step": 3140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.620577977257334,
+      "learning_rate": 9.814841605173704e-06,
+      "loss": 1.0901,
+      "step": 3145
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 30.05892374289853,
+      "learning_rate": 9.813505865707437e-06,
+      "loss": 1.1164,
+      "step": 3150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 22.96909488857937,
+      "learning_rate": 9.81216541705855e-06,
+      "loss": 1.1051,
+      "step": 3155
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 18.269660698493023,
+      "learning_rate": 9.810820260538441e-06,
+      "loss": 1.1375,
+      "step": 3160
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.772284348648533,
+      "learning_rate": 9.809470397463112e-06,
+      "loss": 1.116,
+      "step": 3165
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.855531811848072,
+      "learning_rate": 9.808115829153169e-06,
+      "loss": 1.1404,
+      "step": 3170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.71560575354875,
+      "learning_rate": 9.806756556933823e-06,
+      "loss": 1.1401,
+      "step": 3175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.457646473635371,
+      "learning_rate": 9.805392582134888e-06,
+      "loss": 1.0976,
+      "step": 3180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.021716252996788,
+      "learning_rate": 9.804023906090779e-06,
+      "loss": 1.1171,
+      "step": 3185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 36.17543609821767,
+      "learning_rate": 9.802650530140503e-06,
+      "loss": 1.1359,
+      "step": 3190
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.174434300819575,
+      "learning_rate": 9.801272455627678e-06,
+      "loss": 1.1354,
+      "step": 3195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.968977941832232,
+      "learning_rate": 9.799889683900506e-06,
+      "loss": 1.082,
+      "step": 3200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.179685514337823,
+      "learning_rate": 9.798502216311791e-06,
+      "loss": 1.1101,
+      "step": 3205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 28.261361854490232,
+      "learning_rate": 9.797110054218932e-06,
+      "loss": 1.1267,
+      "step": 3210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 24.91609567279778,
+      "learning_rate": 9.795713198983918e-06,
+      "loss": 1.1207,
+      "step": 3215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 19.450449096827676,
+      "learning_rate": 9.794311651973329e-06,
+      "loss": 1.1189,
+      "step": 3220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 34.838540645917654,
+      "learning_rate": 9.792905414558336e-06,
+      "loss": 1.1238,
+      "step": 3225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 49.79019197252163,
+      "learning_rate": 9.7914944881147e-06,
+      "loss": 1.1031,
+      "step": 3230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 22.569223081112888,
+      "learning_rate": 9.790078874022767e-06,
+      "loss": 1.103,
+      "step": 3235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 34.088080046583904,
+      "learning_rate": 9.788658573667475e-06,
+      "loss": 1.1132,
+      "step": 3240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 110.16064646913127,
+      "learning_rate": 9.787233588438336e-06,
+      "loss": 1.1,
+      "step": 3245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 40.23449989378385,
+      "learning_rate": 9.785803919729455e-06,
+      "loss": 1.1135,
+      "step": 3250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 44.46605455243022,
+      "learning_rate": 9.784369568939516e-06,
+      "loss": 1.1247,
+      "step": 3255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 21.131004713406863,
+      "learning_rate": 9.78293053747178e-06,
+      "loss": 1.1123,
+      "step": 3260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.71929947964882,
+      "learning_rate": 9.781486826734092e-06,
+      "loss": 1.0865,
+      "step": 3265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 42.99587715186835,
+      "learning_rate": 9.780038438138875e-06,
+      "loss": 1.1011,
+      "step": 3270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 39.95133149205219,
+      "learning_rate": 9.778585373103123e-06,
+      "loss": 1.0921,
+      "step": 3275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 26.10809015145351,
+      "learning_rate": 9.777127633048412e-06,
+      "loss": 1.1019,
+      "step": 3280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 34.02220191762274,
+      "learning_rate": 9.775665219400884e-06,
+      "loss": 1.1112,
+      "step": 3285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.365701088432422,
+      "learning_rate": 9.774198133591263e-06,
+      "loss": 1.0958,
+      "step": 3290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.289864979863733,
+      "learning_rate": 9.772726377054838e-06,
+      "loss": 1.0848,
+      "step": 3295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 14.09533795089284,
+      "learning_rate": 9.771249951231465e-06,
+      "loss": 1.1023,
+      "step": 3300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 33.23702455667827,
+      "learning_rate": 9.769768857565573e-06,
+      "loss": 1.1026,
+      "step": 3305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.308081889604273,
+      "learning_rate": 9.768283097506155e-06,
+      "loss": 1.079,
+      "step": 3310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 24.694123563015506,
+      "learning_rate": 9.766792672506771e-06,
+      "loss": 1.0742,
+      "step": 3315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 20.34539350546079,
+      "learning_rate": 9.765297584025542e-06,
+      "loss": 1.0914,
+      "step": 3320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 46.916385220387916,
+      "learning_rate": 9.763797833525157e-06,
+      "loss": 1.097,
+      "step": 3325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.347280588345283,
+      "learning_rate": 9.76229342247286e-06,
+      "loss": 1.0702,
+      "step": 3330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.490168781307537,
+      "learning_rate": 9.760784352340457e-06,
+      "loss": 1.1083,
+      "step": 3335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 35.84052955467092,
+      "learning_rate": 9.759270624604314e-06,
+      "loss": 1.0903,
+      "step": 3340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 30.14735937800149,
+      "learning_rate": 9.757752240745348e-06,
+      "loss": 1.0864,
+      "step": 3345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.482082933392752,
+      "learning_rate": 9.756229202249038e-06,
+      "loss": 1.1044,
+      "step": 3350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 24.69217569957398,
+      "learning_rate": 9.75470151060541e-06,
+      "loss": 1.0724,
+      "step": 3355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.182521065589473,
+      "learning_rate": 9.753169167309052e-06,
+      "loss": 1.1146,
+      "step": 3360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.069497891124406,
+      "learning_rate": 9.751632173859092e-06,
+      "loss": 1.0826,
+      "step": 3365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.510795081029487,
+      "learning_rate": 9.750090531759211e-06,
+      "loss": 1.0923,
+      "step": 3370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 25.081856809348373,
+      "learning_rate": 9.748544242517644e-06,
+      "loss": 1.0852,
+      "step": 3375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 16.06757949104813,
+      "learning_rate": 9.746993307647165e-06,
+      "loss": 1.1093,
+      "step": 3380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 35.25108551159975,
+      "learning_rate": 9.745437728665095e-06,
+      "loss": 1.0882,
+      "step": 3385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.884674188000995,
+      "learning_rate": 9.743877507093298e-06,
+      "loss": 1.0845,
+      "step": 3390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.939698638231848,
+      "learning_rate": 9.742312644458183e-06,
+      "loss": 1.1061,
+      "step": 3395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.83963911448851,
+      "learning_rate": 9.740743142290696e-06,
+      "loss": 1.0948,
+      "step": 3400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.542057789428704,
+      "learning_rate": 9.739169002126326e-06,
+      "loss": 1.0681,
+      "step": 3405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.85859530160301,
+      "learning_rate": 9.737590225505091e-06,
+      "loss": 1.0714,
+      "step": 3410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.529021625117974,
+      "learning_rate": 9.736006813971557e-06,
+      "loss": 1.083,
+      "step": 3415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 32.642327759447724,
+      "learning_rate": 9.734418769074813e-06,
+      "loss": 1.0949,
+      "step": 3420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 18.722861850410514,
+      "learning_rate": 9.732826092368491e-06,
+      "loss": 1.0958,
+      "step": 3425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 39.00659387481143,
+      "learning_rate": 9.731228785410746e-06,
+      "loss": 1.0861,
+      "step": 3430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.711037906612924,
+      "learning_rate": 9.729626849764266e-06,
+      "loss": 1.0772,
+      "step": 3435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.42121870550603,
+      "learning_rate": 9.728020286996273e-06,
+      "loss": 1.0815,
+      "step": 3440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 30.639900438665272,
+      "learning_rate": 9.726409098678505e-06,
+      "loss": 1.0801,
+      "step": 3445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.123471421452953,
+      "learning_rate": 9.724793286387233e-06,
+      "loss": 1.0938,
+      "step": 3450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 19.93670960977343,
+      "learning_rate": 9.72317285170325e-06,
+      "loss": 1.0521,
+      "step": 3455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 28.837291173282402,
+      "learning_rate": 9.721547796211872e-06,
+      "loss": 1.0937,
+      "step": 3460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 99.22043991359959,
+      "learning_rate": 9.719918121502933e-06,
+      "loss": 1.1099,
+      "step": 3465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 118.91259663380502,
+      "learning_rate": 9.718283829170788e-06,
+      "loss": 1.0882,
+      "step": 3470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 94.83197424147019,
+      "learning_rate": 9.71664492081431e-06,
+      "loss": 1.1115,
+      "step": 3475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 23.176334095455275,
+      "learning_rate": 9.715001398036884e-06,
+      "loss": 1.1077,
+      "step": 3480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 23.25526333684265,
+      "learning_rate": 9.713353262446419e-06,
+      "loss": 1.0704,
+      "step": 3485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 62.50190406613821,
+      "learning_rate": 9.711700515655327e-06,
+      "loss": 1.0833,
+      "step": 3490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 46.93771092847943,
+      "learning_rate": 9.710043159280532e-06,
+      "loss": 1.0825,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 29.238170589766323,
+      "learning_rate": 9.708381194943476e-06,
+      "loss": 1.0879,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.685350295706728,
+      "learning_rate": 9.706714624270097e-06,
+      "loss": 1.0783,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.303197061444635,
+      "learning_rate": 9.705043448890852e-06,
+      "loss": 1.0803,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 23.053480521822973,
+      "learning_rate": 9.703367670440695e-06,
+      "loss": 1.0666,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 16.52258620177513,
+      "learning_rate": 9.701687290559084e-06,
+      "loss": 1.082,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 22.692161284567145,
+      "learning_rate": 9.700002310889981e-06,
+      "loss": 1.0668,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 20.73627310180072,
+      "learning_rate": 9.698312733081847e-06,
+      "loss": 1.0806,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 31.383021093811415,
+      "learning_rate": 9.696618558787643e-06,
+      "loss": 1.0799,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 35.7883766937957,
+      "learning_rate": 9.69491978966482e-06,
+      "loss": 1.0579,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 22.439749721861627,
+      "learning_rate": 9.693216427375338e-06,
+      "loss": 1.055,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.528624040661086,
+      "learning_rate": 9.691508473585635e-06,
+      "loss": 1.0866,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.191808567072622,
+      "learning_rate": 9.689795929966653e-06,
+      "loss": 1.0625,
+      "step": 3555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 49.35630024440896,
+      "learning_rate": 9.688078798193816e-06,
+      "loss": 1.0858,
+      "step": 3560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 141.8741033629112,
+      "learning_rate": 9.686357079947042e-06,
+      "loss": 1.0819,
+      "step": 3565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 101.01456442603427,
+      "learning_rate": 9.684630776910734e-06,
+      "loss": 1.0799,
+      "step": 3570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 48.44038091165258,
+      "learning_rate": 9.682899890773782e-06,
+      "loss": 1.0784,
+      "step": 3575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 39.09727281088586,
+      "learning_rate": 9.68116442322956e-06,
+      "loss": 1.094,
+      "step": 3580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 58.65427281326046,
+      "learning_rate": 9.679424375975916e-06,
+      "loss": 1.0738,
+      "step": 3585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.717354874808702,
+      "learning_rate": 9.677679750715194e-06,
+      "loss": 1.0706,
+      "step": 3590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.32608732711157,
+      "learning_rate": 9.675930549154201e-06,
+      "loss": 1.0567,
+      "step": 3595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 20.979587808667443,
+      "learning_rate": 9.674176773004232e-06,
+      "loss": 1.0738,
+      "step": 3600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 19.984425443283612,
+      "learning_rate": 9.672418423981051e-06,
+      "loss": 1.0816,
+      "step": 3605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.322899089745414,
+      "learning_rate": 9.6706555038049e-06,
+      "loss": 1.0641,
+      "step": 3610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.27669372683442,
+      "learning_rate": 9.66888801420049e-06,
+      "loss": 1.0529,
+      "step": 3615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 26.73787927731739,
+      "learning_rate": 9.667115956897007e-06,
+      "loss": 1.0565,
+      "step": 3620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.332940164863567,
+      "learning_rate": 9.6653393336281e-06,
+      "loss": 1.0608,
+      "step": 3625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.957432148595675,
+      "learning_rate": 9.663558146131886e-06,
+      "loss": 1.062,
+      "step": 3630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.524600244030466,
+      "learning_rate": 9.66177239615095e-06,
+      "loss": 1.0416,
+      "step": 3635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 33.43797866232677,
+      "learning_rate": 9.65998208543234e-06,
+      "loss": 1.0665,
+      "step": 3640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 28.187859604767837,
+      "learning_rate": 9.658187215727567e-06,
+      "loss": 1.0512,
+      "step": 3645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 43.17834437252287,
+      "learning_rate": 9.656387788792594e-06,
+      "loss": 1.0667,
+      "step": 3650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.833570997925662,
+      "learning_rate": 9.654583806387855e-06,
+      "loss": 1.0474,
+      "step": 3655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 19.4936094387291,
+      "learning_rate": 9.652775270278236e-06,
+      "loss": 1.0629,
+      "step": 3660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.406721518276477,
+      "learning_rate": 9.65096218223307e-06,
+      "loss": 1.0415,
+      "step": 3665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.348020212680531,
+      "learning_rate": 9.649144544026154e-06,
+      "loss": 1.0609,
+      "step": 3670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 25.662165096305973,
+      "learning_rate": 9.647322357435734e-06,
+      "loss": 1.0547,
+      "step": 3675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 62.94870309796483,
+      "learning_rate": 9.6454956242445e-06,
+      "loss": 1.0514,
+      "step": 3680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 33.18978079391046,
+      "learning_rate": 9.643664346239598e-06,
+      "loss": 1.0716,
+      "step": 3685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 34.21728007762453,
+      "learning_rate": 9.641828525212616e-06,
+      "loss": 1.0459,
+      "step": 3690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 19.398489033262475,
+      "learning_rate": 9.639988162959586e-06,
+      "loss": 1.0614,
+      "step": 3695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 35.381394002160434,
+      "learning_rate": 9.638143261280983e-06,
+      "loss": 1.0515,
+      "step": 3700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 52.88442016004509,
+      "learning_rate": 9.636293821981728e-06,
+      "loss": 1.0715,
+      "step": 3705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 60.17940969814508,
+      "learning_rate": 9.634439846871173e-06,
+      "loss": 1.0605,
+      "step": 3710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 42.01453778390265,
+      "learning_rate": 9.632581337763115e-06,
+      "loss": 1.0632,
+      "step": 3715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.91412969693914,
+      "learning_rate": 9.630718296475782e-06,
+      "loss": 1.0546,
+      "step": 3720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.122096671164469,
+      "learning_rate": 9.628850724831838e-06,
+      "loss": 1.0404,
+      "step": 3725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.360961065245824,
+      "learning_rate": 9.626978624658378e-06,
+      "loss": 1.0499,
+      "step": 3730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 20.681549559872614,
+      "learning_rate": 9.625101997786929e-06,
+      "loss": 1.0515,
+      "step": 3735
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 34.311939706830195,
+      "learning_rate": 9.623220846053448e-06,
+      "loss": 1.0422,
+      "step": 3740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.754605131977751,
+      "learning_rate": 9.621335171298312e-06,
+      "loss": 1.0469,
+      "step": 3745
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.552801959273697,
+      "learning_rate": 9.619444975366328e-06,
+      "loss": 1.0359,
+      "step": 3750
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.2524305965504,
+      "learning_rate": 9.617550260106729e-06,
+      "loss": 1.0403,
+      "step": 3755
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.093338843745048,
+      "learning_rate": 9.615651027373163e-06,
+      "loss": 1.0454,
+      "step": 3760
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.439354245341704,
+      "learning_rate": 9.613747279023704e-06,
+      "loss": 1.0597,
+      "step": 3765
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 34.420764529314724,
+      "learning_rate": 9.611839016920837e-06,
+      "loss": 1.034,
+      "step": 3770
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.15542917484291,
+      "learning_rate": 9.609926242931467e-06,
+      "loss": 1.0317,
+      "step": 3775
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 20.92806883061577,
+      "learning_rate": 9.608008958926911e-06,
+      "loss": 1.0545,
+      "step": 3780
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 20.937194797150127,
+      "learning_rate": 9.606087166782904e-06,
+      "loss": 1.0473,
+      "step": 3785
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.847890544358966,
+      "learning_rate": 9.604160868379584e-06,
+      "loss": 1.0362,
+      "step": 3790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 61.39464963918651,
+      "learning_rate": 9.6022300656015e-06,
+      "loss": 1.0238,
+      "step": 3795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 101.58609678822324,
+      "learning_rate": 9.600294760337611e-06,
+      "loss": 1.0628,
+      "step": 3800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 48.990707631127584,
+      "learning_rate": 9.598354954481276e-06,
+      "loss": 1.0538,
+      "step": 3805
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 17.648537571948292,
+      "learning_rate": 9.596410649930262e-06,
+      "loss": 1.046,
+      "step": 3810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 21.385685025543232,
+      "learning_rate": 9.59446184858673e-06,
+      "loss": 1.0317,
+      "step": 3815
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 27.822628216288326,
+      "learning_rate": 9.592508552357251e-06,
+      "loss": 1.0599,
+      "step": 3820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 43.65608973671009,
+      "learning_rate": 9.590550763152781e-06,
+      "loss": 1.0216,
+      "step": 3825
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 68.66082796487945,
+      "learning_rate": 9.588588482888684e-06,
+      "loss": 1.0369,
+      "step": 3830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 44.87563845739073,
+      "learning_rate": 9.586621713484708e-06,
+      "loss": 1.0447,
+      "step": 3835
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 85.29718189990733,
+      "learning_rate": 9.584650456864997e-06,
+      "loss": 1.0345,
+      "step": 3840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 41.63515269827128,
+      "learning_rate": 9.582674714958088e-06,
+      "loss": 1.0539,
+      "step": 3845
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 44.2242489976644,
+      "learning_rate": 9.580694489696896e-06,
+      "loss": 1.0205,
+      "step": 3850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 36.041665972339715,
+      "learning_rate": 9.578709783018734e-06,
+      "loss": 1.0488,
+      "step": 3855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 67.57874599444452,
+      "learning_rate": 9.576720596865292e-06,
+      "loss": 1.0363,
+      "step": 3860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 38.21695327694072,
+      "learning_rate": 9.574726933182645e-06,
+      "loss": 1.0391,
+      "step": 3865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 13.661349716058288,
+      "learning_rate": 9.572728793921248e-06,
+      "loss": 1.0213,
+      "step": 3870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.13353362747518,
+      "learning_rate": 9.570726181035934e-06,
+      "loss": 1.0435,
+      "step": 3875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.334770151580553,
+      "learning_rate": 9.568719096485915e-06,
+      "loss": 1.0535,
+      "step": 3880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 13.117262631880102,
+      "learning_rate": 9.566707542234774e-06,
+      "loss": 1.0373,
+      "step": 3885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.325426340360542,
+      "learning_rate": 9.56469152025047e-06,
+      "loss": 1.02,
+      "step": 3890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 27.760103500384453,
+      "learning_rate": 9.562671032505328e-06,
+      "loss": 1.0613,
+      "step": 3895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 19.677606688797173,
+      "learning_rate": 9.560646080976052e-06,
+      "loss": 1.0374,
+      "step": 3900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.221222305008846,
+      "learning_rate": 9.558616667643703e-06,
+      "loss": 1.0383,
+      "step": 3905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.05047992856828,
+      "learning_rate": 9.55658279449371e-06,
+      "loss": 1.0332,
+      "step": 3910
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.783299101241285,
+      "learning_rate": 9.554544463515867e-06,
+      "loss": 1.0334,
+      "step": 3915
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.469272814234113,
+      "learning_rate": 9.552501676704328e-06,
+      "loss": 1.0243,
+      "step": 3920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.437389694030383,
+      "learning_rate": 9.55045443605761e-06,
+      "loss": 1.042,
+      "step": 3925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.818804679626847,
+      "learning_rate": 9.548402743578578e-06,
+      "loss": 1.0106,
+      "step": 3930
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 72.12551959106173,
+      "learning_rate": 9.54634660127446e-06,
+      "loss": 1.0335,
+      "step": 3935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 60.35525493588452,
+      "learning_rate": 9.544286011156838e-06,
+      "loss": 1.0206,
+      "step": 3940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 28.517071830491236,
+      "learning_rate": 9.542220975241641e-06,
+      "loss": 1.027,
+      "step": 3945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.41421474517318,
+      "learning_rate": 9.540151495549148e-06,
+      "loss": 1.0294,
+      "step": 3950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.95708153303163,
+      "learning_rate": 9.538077574103988e-06,
+      "loss": 1.0358,
+      "step": 3955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 21.632102740651966,
+      "learning_rate": 9.535999212935135e-06,
+      "loss": 1.0211,
+      "step": 3960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.86693762831278,
+      "learning_rate": 9.533916414075906e-06,
+      "loss": 1.0252,
+      "step": 3965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 21.262197447013794,
+      "learning_rate": 9.531829179563958e-06,
+      "loss": 1.039,
+      "step": 3970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 34.54442719901615,
+      "learning_rate": 9.529737511441288e-06,
+      "loss": 1.0335,
+      "step": 3975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 63.2216943225613,
+      "learning_rate": 9.527641411754234e-06,
+      "loss": 1.0274,
+      "step": 3980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 48.38593781451802,
+      "learning_rate": 9.525540882553465e-06,
+      "loss": 1.0166,
+      "step": 3985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 45.643379369645665,
+      "learning_rate": 9.523435925893986e-06,
+      "loss": 1.0045,
+      "step": 3990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 24.825075531481474,
+      "learning_rate": 9.521326543835135e-06,
+      "loss": 1.0216,
+      "step": 3995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 32.30923946943804,
+      "learning_rate": 9.519212738440572e-06,
+      "loss": 1.0252,
+      "step": 4000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 31.268891186427773,
+      "learning_rate": 9.517094511778294e-06,
+      "loss": 0.9941,
+      "step": 4005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 19.506914693015624,
+      "learning_rate": 9.514971865920618e-06,
+      "loss": 1.0087,
+      "step": 4010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.515776436928725,
+      "learning_rate": 9.512844802944186e-06,
+      "loss": 1.0195,
+      "step": 4015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.148277998784094,
+      "learning_rate": 9.510713324929962e-06,
+      "loss": 1.0374,
+      "step": 4020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.235328265558776,
+      "learning_rate": 9.508577433963227e-06,
+      "loss": 1.0438,
+      "step": 4025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.09979546945935,
+      "learning_rate": 9.506437132133581e-06,
+      "loss": 1.0227,
+      "step": 4030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.238589281663184,
+      "learning_rate": 9.50429242153494e-06,
+      "loss": 1.0176,
+      "step": 4035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.311313851556408,
+      "learning_rate": 9.50214330426553e-06,
+      "loss": 1.0284,
+      "step": 4040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.345687173293602,
+      "learning_rate": 9.499989782427893e-06,
+      "loss": 1.0142,
+      "step": 4045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.004499727601269,
+      "learning_rate": 9.497831858128876e-06,
+      "loss": 1.0229,
+      "step": 4050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.725097828165937,
+      "learning_rate": 9.495669533479634e-06,
+      "loss": 1.0155,
+      "step": 4055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.228255051625794,
+      "learning_rate": 9.49350281059563e-06,
+      "loss": 1.0027,
+      "step": 4060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.82098769307729,
+      "learning_rate": 9.491331691596625e-06,
+      "loss": 1.0003,
+      "step": 4065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 20.029534817911642,
+      "learning_rate": 9.489156178606684e-06,
+      "loss": 1.0191,
+      "step": 4070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.096382014592697,
+      "learning_rate": 9.48697627375417e-06,
+      "loss": 1.0358,
+      "step": 4075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.164168759887309,
+      "learning_rate": 9.484791979171744e-06,
+      "loss": 1.0141,
+      "step": 4080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.03131118489538,
+      "learning_rate": 9.482603296996358e-06,
+      "loss": 1.0161,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.75518249609745,
+      "learning_rate": 9.48041022936926e-06,
+      "loss": 1.0082,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.349151443252808,
+      "learning_rate": 9.478212778435987e-06,
+      "loss": 1.0038,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.74009872478776,
+      "learning_rate": 9.476010946346365e-06,
+      "loss": 1.0176,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.412538009143013,
+      "learning_rate": 9.473804735254507e-06,
+      "loss": 1.0209,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 55.590697499654354,
+      "learning_rate": 9.471594147318806e-06,
+      "loss": 1.0235,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.395513398092206,
+      "learning_rate": 9.469379184701942e-06,
+      "loss": 1.0252,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 82.57676781630296,
+      "learning_rate": 9.46715984957087e-06,
+      "loss": 1.0101,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 109.30761353748431,
+      "learning_rate": 9.464936144096828e-06,
+      "loss": 1.0,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 53.283011866055276,
+      "learning_rate": 9.462708070455327e-06,
+      "loss": 1.0277,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 53.341823780234854,
+      "learning_rate": 9.46047563082615e-06,
+      "loss": 1.0237,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.15169787672353,
+      "learning_rate": 9.458238827393353e-06,
+      "loss": 1.0182,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.281494842923637,
+      "learning_rate": 9.455997662345262e-06,
+      "loss": 0.9904,
+      "step": 4145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.863337348739847,
+      "learning_rate": 9.45375213787447e-06,
+      "loss": 1.0252,
+      "step": 4150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 19.464852452342406,
+      "learning_rate": 9.451502256177832e-06,
+      "loss": 1.008,
+      "step": 4155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 23.836143947796305,
+      "learning_rate": 9.44924801945647e-06,
+      "loss": 1.0041,
+      "step": 4160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 21.083807605337846,
+      "learning_rate": 9.446989429915763e-06,
+      "loss": 1.0062,
+      "step": 4165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 20.37749053282835,
+      "learning_rate": 9.44472648976535e-06,
+      "loss": 1.0121,
+      "step": 4170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.873447859539434,
+      "learning_rate": 9.442459201219127e-06,
+      "loss": 1.0051,
+      "step": 4175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 18.959460649769564,
+      "learning_rate": 9.440187566495246e-06,
+      "loss": 1.004,
+      "step": 4180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 27.467828324088764,
+      "learning_rate": 9.437911587816105e-06,
+      "loss": 0.9968,
+      "step": 4185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 43.20261939532219,
+      "learning_rate": 9.435631267408355e-06,
+      "loss": 1.0115,
+      "step": 4190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 26.800565504607206,
+      "learning_rate": 9.4333466075029e-06,
+      "loss": 1.0122,
+      "step": 4195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.94300587635886,
+      "learning_rate": 9.431057610334878e-06,
+      "loss": 1.0043,
+      "step": 4200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.614855349141937,
+      "learning_rate": 9.42876427814368e-06,
+      "loss": 1.0002,
+      "step": 4205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.144998257948764,
+      "learning_rate": 9.426466613172935e-06,
+      "loss": 0.9941,
+      "step": 4210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.103838790198429,
+      "learning_rate": 9.42416461767051e-06,
+      "loss": 0.9967,
+      "step": 4215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.103071437327314,
+      "learning_rate": 9.421858293888509e-06,
+      "loss": 0.9872,
+      "step": 4220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.4543054300558,
+      "learning_rate": 9.41954764408327e-06,
+      "loss": 1.0034,
+      "step": 4225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 29.419165904681627,
+      "learning_rate": 9.417232670515367e-06,
+      "loss": 0.9876,
+      "step": 4230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 24.500606114143153,
+      "learning_rate": 9.414913375449598e-06,
+      "loss": 1.004,
+      "step": 4235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.066193807337648,
+      "learning_rate": 9.41258976115499e-06,
+      "loss": 1.0127,
+      "step": 4240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 52.031867154118224,
+      "learning_rate": 9.410261829904805e-06,
+      "loss": 1.0046,
+      "step": 4245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 21.157296869356358,
+      "learning_rate": 9.407929583976514e-06,
+      "loss": 1.0206,
+      "step": 4250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.959852530800298,
+      "learning_rate": 9.40559302565182e-06,
+      "loss": 0.9852,
+      "step": 4255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.55092856854972,
+      "learning_rate": 9.40325215721664e-06,
+      "loss": 1.0152,
+      "step": 4260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.525060393654432,
+      "learning_rate": 9.400906980961106e-06,
+      "loss": 1.004,
+      "step": 4265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.212100273291716,
+      "learning_rate": 9.398557499179573e-06,
+      "loss": 0.9973,
+      "step": 4270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 22.460965399961488,
+      "learning_rate": 9.396203714170595e-06,
+      "loss": 0.996,
+      "step": 4275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.39405900296308,
+      "learning_rate": 9.393845628236949e-06,
+      "loss": 1.007,
+      "step": 4280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 51.01601071441724,
+      "learning_rate": 9.391483243685612e-06,
+      "loss": 0.9842,
+      "step": 4285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 56.88915495340164,
+      "learning_rate": 9.38911656282777e-06,
+      "loss": 1.0078,
+      "step": 4290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 65.33444023898811,
+      "learning_rate": 9.386745587978809e-06,
+      "loss": 0.9993,
+      "step": 4295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 35.49556367595794,
+      "learning_rate": 9.384370321458318e-06,
+      "loss": 0.986,
+      "step": 4300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 33.949914291057055,
+      "learning_rate": 9.381990765590086e-06,
+      "loss": 0.9923,
+      "step": 4305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 29.173102258533216,
+      "learning_rate": 9.379606922702092e-06,
+      "loss": 0.9875,
+      "step": 4310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.288434941882375,
+      "learning_rate": 9.377218795126519e-06,
+      "loss": 1.0079,
+      "step": 4315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 19.693191672912995,
+      "learning_rate": 9.374826385199735e-06,
+      "loss": 0.9909,
+      "step": 4320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.090041449361541,
+      "learning_rate": 9.372429695262297e-06,
+      "loss": 0.9855,
+      "step": 4325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.317039626693504,
+      "learning_rate": 9.370028727658956e-06,
+      "loss": 0.9726,
+      "step": 4330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.737797784798412,
+      "learning_rate": 9.367623484738639e-06,
+      "loss": 0.9939,
+      "step": 4335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 22.78204210712204,
+      "learning_rate": 9.365213968854463e-06,
+      "loss": 0.9837,
+      "step": 4340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 72.390320020571,
+      "learning_rate": 9.362800182363718e-06,
+      "loss": 0.9886,
+      "step": 4345
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 57.5047526181269,
+      "learning_rate": 9.36038212762788e-06,
+      "loss": 1.0029,
+      "step": 4350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 101.10985440634894,
+      "learning_rate": 9.3579598070126e-06,
+      "loss": 0.9989,
+      "step": 4355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 72.40360809441891,
+      "learning_rate": 9.355533222887693e-06,
+      "loss": 1.0273,
+      "step": 4360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 23.804178434830444,
+      "learning_rate": 9.353102377627155e-06,
+      "loss": 0.9941,
+      "step": 4365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 96.09502523304418,
+      "learning_rate": 9.350667273609148e-06,
+      "loss": 1.0098,
+      "step": 4370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 75.77778989570355,
+      "learning_rate": 9.348227913216e-06,
+      "loss": 0.9983,
+      "step": 4375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.44568445884761,
+      "learning_rate": 9.345784298834202e-06,
+      "loss": 0.9828,
+      "step": 4380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 41.25631175653349,
+      "learning_rate": 9.343336432854408e-06,
+      "loss": 0.9831,
+      "step": 4385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 35.698092682203594,
+      "learning_rate": 9.340884317671432e-06,
+      "loss": 0.998,
+      "step": 4390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 32.49391162293078,
+      "learning_rate": 9.338427955684243e-06,
+      "loss": 0.9806,
+      "step": 4395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.241743309089527,
+      "learning_rate": 9.335967349295967e-06,
+      "loss": 0.9885,
+      "step": 4400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 32.37182387434173,
+      "learning_rate": 9.333502500913882e-06,
+      "loss": 0.9873,
+      "step": 4405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 30.142157717062506,
+      "learning_rate": 9.331033412949417e-06,
+      "loss": 1.0035,
+      "step": 4410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.9808930058223,
+      "learning_rate": 9.328560087818143e-06,
+      "loss": 0.9826,
+      "step": 4415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.087635510049255,
+      "learning_rate": 9.326082527939786e-06,
+      "loss": 0.9674,
+      "step": 4420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.901907099337802,
+      "learning_rate": 9.323600735738207e-06,
+      "loss": 0.9783,
+      "step": 4425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 22.503501555175596,
+      "learning_rate": 9.321114713641409e-06,
+      "loss": 0.9824,
+      "step": 4430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 15.802700822262079,
+      "learning_rate": 9.318624464081535e-06,
+      "loss": 0.9759,
+      "step": 4435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.704909570321815,
+      "learning_rate": 9.316129989494866e-06,
+      "loss": 0.9643,
+      "step": 4440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.51245114231154,
+      "learning_rate": 9.313631292321812e-06,
+      "loss": 0.9793,
+      "step": 4445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.31093610700748,
+      "learning_rate": 9.311128375006915e-06,
+      "loss": 1.0008,
+      "step": 4450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.48618925404952,
+      "learning_rate": 9.308621239998847e-06,
+      "loss": 0.9894,
+      "step": 4455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.277491750024144,
+      "learning_rate": 9.306109889750405e-06,
+      "loss": 0.987,
+      "step": 4460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.651178877768626,
+      "learning_rate": 9.303594326718514e-06,
+      "loss": 0.9691,
+      "step": 4465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 22.96171498854302,
+      "learning_rate": 9.301074553364214e-06,
+      "loss": 0.9568,
+      "step": 4470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 21.281745464081407,
+      "learning_rate": 9.29855057215267e-06,
+      "loss": 0.9785,
+      "step": 4475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 45.786548050160974,
+      "learning_rate": 9.296022385553156e-06,
+      "loss": 0.9658,
+      "step": 4480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 28.449037927202944,
+      "learning_rate": 9.293489996039068e-06,
+      "loss": 0.9571,
+      "step": 4485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 25.5063776566906,
+      "learning_rate": 9.290953406087913e-06,
+      "loss": 0.9924,
+      "step": 4490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 23.902215031162577,
+      "learning_rate": 9.288412618181305e-06,
+      "loss": 0.9644,
+      "step": 4495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 28.321553047312477,
+      "learning_rate": 9.285867634804961e-06,
+      "loss": 0.9682,
+      "step": 4500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 19.414695666211387,
+      "learning_rate": 9.283318458448711e-06,
+      "loss": 0.9748,
+      "step": 4505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.265335693282779,
+      "learning_rate": 9.280765091606481e-06,
+      "loss": 0.9667,
+      "step": 4510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.589118747517364,
+      "learning_rate": 9.2782075367763e-06,
+      "loss": 0.9871,
+      "step": 4515
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.766284702178446,
+      "learning_rate": 9.275645796460292e-06,
+      "loss": 0.9696,
+      "step": 4520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 22.62383414229558,
+      "learning_rate": 9.273079873164676e-06,
+      "loss": 0.972,
+      "step": 4525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 22.746944385026122,
+      "learning_rate": 9.270509769399767e-06,
+      "loss": 0.9561,
+      "step": 4530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.635254266812355,
+      "learning_rate": 9.267935487679962e-06,
+      "loss": 0.9686,
+      "step": 4535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 26.008001828316356,
+      "learning_rate": 9.265357030523756e-06,
+      "loss": 0.9966,
+      "step": 4540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 19.690837446359836,
+      "learning_rate": 9.262774400453717e-06,
+      "loss": 0.9762,
+      "step": 4545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.165326461540976,
+      "learning_rate": 9.260187599996507e-06,
+      "loss": 0.9907,
+      "step": 4550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.63560083766881,
+      "learning_rate": 9.25759663168286e-06,
+      "loss": 0.9769,
+      "step": 4555
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 29.926072811215636,
+      "learning_rate": 9.255001498047592e-06,
+      "loss": 0.9694,
+      "step": 4560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 43.916438852666026,
+      "learning_rate": 9.252402201629588e-06,
+      "loss": 0.9616,
+      "step": 4565
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 36.62552865460111,
+      "learning_rate": 9.249798744971815e-06,
+      "loss": 0.9674,
+      "step": 4570
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 44.70586590779445,
+      "learning_rate": 9.2471911306213e-06,
+      "loss": 0.9651,
+      "step": 4575
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 24.188160765283822,
+      "learning_rate": 9.244579361129147e-06,
+      "loss": 0.9441,
+      "step": 4580
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.006816269337822,
+      "learning_rate": 9.241963439050519e-06,
+      "loss": 0.9683,
+      "step": 4585
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.293519001103515,
+      "learning_rate": 9.239343366944641e-06,
+      "loss": 0.967,
+      "step": 4590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.409536051527816,
+      "learning_rate": 9.236719147374801e-06,
+      "loss": 0.9567,
+      "step": 4595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.825285059132497,
+      "learning_rate": 9.234090782908346e-06,
+      "loss": 0.9572,
+      "step": 4600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.057182699834947,
+      "learning_rate": 9.231458276116676e-06,
+      "loss": 0.9747,
+      "step": 4605
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.464766579821713,
+      "learning_rate": 9.22882162957524e-06,
+      "loss": 0.949,
+      "step": 4610
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.69207643557753,
+      "learning_rate": 9.226180845863544e-06,
+      "loss": 0.9741,
+      "step": 4615
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 38.55352202928315,
+      "learning_rate": 9.223535927565135e-06,
+      "loss": 0.9529,
+      "step": 4620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 34.96091841468431,
+      "learning_rate": 9.220886877267609e-06,
+      "loss": 0.9746,
+      "step": 4625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 20.01930871822319,
+      "learning_rate": 9.218233697562604e-06,
+      "loss": 0.9626,
+      "step": 4630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.324063907979856,
+      "learning_rate": 9.215576391045797e-06,
+      "loss": 0.9602,
+      "step": 4635
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.150997568199399,
+      "learning_rate": 9.212914960316902e-06,
+      "loss": 0.9481,
+      "step": 4640
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 41.68771819761749,
+      "learning_rate": 9.21024940797967e-06,
+      "loss": 0.9771,
+      "step": 4645
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 73.61863057444658,
+      "learning_rate": 9.207579736641881e-06,
+      "loss": 0.9795,
+      "step": 4650
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 38.792496873295825,
+      "learning_rate": 9.204905948915345e-06,
+      "loss": 0.9535,
+      "step": 4655
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 51.91198314996973,
+      "learning_rate": 9.202228047415905e-06,
+      "loss": 0.9454,
+      "step": 4660
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 66.68820192554747,
+      "learning_rate": 9.199546034763423e-06,
+      "loss": 0.9643,
+      "step": 4665
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 115.50748156350188,
+      "learning_rate": 9.196859913581781e-06,
+      "loss": 0.978,
+      "step": 4670
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 50.953571534266246,
+      "learning_rate": 9.194169686498887e-06,
+      "loss": 0.9425,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 49.81249494752531,
+      "learning_rate": 9.191475356146661e-06,
+      "loss": 0.9647,
+      "step": 4680
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.125148068708931,
+      "learning_rate": 9.188776925161042e-06,
+      "loss": 0.9427,
+      "step": 4685
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 35.472294700071664,
+      "learning_rate": 9.186074396181974e-06,
+      "loss": 0.9474,
+      "step": 4690
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 31.34096469566438,
+      "learning_rate": 9.183367771853417e-06,
+      "loss": 0.9613,
+      "step": 4695
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.51902471391319,
+      "learning_rate": 9.180657054823334e-06,
+      "loss": 0.9462,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.096319871398451,
+      "learning_rate": 9.17794224774369e-06,
+      "loss": 0.956,
+      "step": 4705
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.97502452642927,
+      "learning_rate": 9.175223353270457e-06,
+      "loss": 0.9323,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.373732134631773,
+      "learning_rate": 9.172500374063603e-06,
+      "loss": 0.9411,
+      "step": 4715
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.572979968003773,
+      "learning_rate": 9.169773312787086e-06,
+      "loss": 0.9431,
+      "step": 4720
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 17.165158198545925,
+      "learning_rate": 9.167042172108874e-06,
+      "loss": 0.9598,
+      "step": 4725
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.683924164964983,
+      "learning_rate": 9.164306954700905e-06,
+      "loss": 0.9486,
+      "step": 4730
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.224964756040686,
+      "learning_rate": 9.161567663239126e-06,
+      "loss": 0.9391,
+      "step": 4735
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.376098770544099,
+      "learning_rate": 9.15882430040345e-06,
+      "loss": 0.963,
+      "step": 4740
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.908609034274736,
+      "learning_rate": 9.15607686887779e-06,
+      "loss": 0.9311,
+      "step": 4745
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.63840384715431,
+      "learning_rate": 9.153325371350028e-06,
+      "loss": 0.9697,
+      "step": 4750
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 22.946614579574426,
+      "learning_rate": 9.150569810512033e-06,
+      "loss": 0.96,
+      "step": 4755
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.167063738591903,
+      "learning_rate": 9.147810189059639e-06,
+      "loss": 0.9479,
+      "step": 4760
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 37.92914695340067,
+      "learning_rate": 9.145046509692661e-06,
+      "loss": 0.9423,
+      "step": 4765
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 24.365889248555195,
+      "learning_rate": 9.142278775114882e-06,
+      "loss": 0.9418,
+      "step": 4770
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 67.68436185583097,
+      "learning_rate": 9.139506988034049e-06,
+      "loss": 0.9537,
+      "step": 4775
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 39.491890864272584,
+      "learning_rate": 9.136731151161877e-06,
+      "loss": 0.9591,
+      "step": 4780
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 45.38530370615395,
+      "learning_rate": 9.133951267214043e-06,
+      "loss": 0.9355,
+      "step": 4785
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.378791634160802,
+      "learning_rate": 9.13116733891018e-06,
+      "loss": 0.9579,
+      "step": 4790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 45.370670351833056,
+      "learning_rate": 9.128379368973884e-06,
+      "loss": 0.9408,
+      "step": 4795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.49270893151392,
+      "learning_rate": 9.125587360132697e-06,
+      "loss": 0.9281,
+      "step": 4800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 43.64429040063124,
+      "learning_rate": 9.12279131511812e-06,
+      "loss": 0.9346,
+      "step": 4805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.215164955234982,
+      "learning_rate": 9.1199912366656e-06,
+      "loss": 0.9424,
+      "step": 4810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.742762494067678,
+      "learning_rate": 9.117187127514524e-06,
+      "loss": 0.9456,
+      "step": 4815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.727143118510911,
+      "learning_rate": 9.11437899040823e-06,
+      "loss": 0.9465,
+      "step": 4820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.832153090627127,
+      "learning_rate": 9.111566828093998e-06,
+      "loss": 0.9178,
+      "step": 4825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.630265380456894,
+      "learning_rate": 9.108750643323036e-06,
+      "loss": 0.9504,
+      "step": 4830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 19.444717713770846,
+      "learning_rate": 9.1059304388505e-06,
+      "loss": 0.9245,
+      "step": 4835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 24.329705280184676,
+      "learning_rate": 9.103106217435467e-06,
+      "loss": 0.9341,
+      "step": 4840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 27.83847152420756,
+      "learning_rate": 9.100277981840953e-06,
+      "loss": 0.9589,
+      "step": 4845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.50842107404206,
+      "learning_rate": 9.097445734833893e-06,
+      "loss": 0.9566,
+      "step": 4850
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 43.06023501265569,
+      "learning_rate": 9.094609479185153e-06,
+      "loss": 0.9532,
+      "step": 4855
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 50.720088892736904,
+      "learning_rate": 9.091769217669517e-06,
+      "loss": 0.9569,
+      "step": 4860
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 62.24897567865879,
+      "learning_rate": 9.088924953065691e-06,
+      "loss": 0.9266,
+      "step": 4865
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 57.80399144018989,
+      "learning_rate": 9.086076688156297e-06,
+      "loss": 0.9427,
+      "step": 4870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.46459487344967,
+      "learning_rate": 9.083224425727867e-06,
+      "loss": 0.9351,
+      "step": 4875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.698678860667545,
+      "learning_rate": 9.080368168570845e-06,
+      "loss": 0.9469,
+      "step": 4880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.687820627493341,
+      "learning_rate": 9.077507919479589e-06,
+      "loss": 0.9212,
+      "step": 4885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 45.913078252317284,
+      "learning_rate": 9.074643681252356e-06,
+      "loss": 0.9336,
+      "step": 4890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 32.80796660229274,
+      "learning_rate": 9.071775456691303e-06,
+      "loss": 0.9298,
+      "step": 4895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.983476920809734,
+      "learning_rate": 9.068903248602497e-06,
+      "loss": 0.9457,
+      "step": 4900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.211639785267453,
+      "learning_rate": 9.066027059795896e-06,
+      "loss": 0.9482,
+      "step": 4905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.51243616241861,
+      "learning_rate": 9.06314689308535e-06,
+      "loss": 0.934,
+      "step": 4910
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.320268531763826,
+      "learning_rate": 9.060262751288607e-06,
+      "loss": 0.9434,
+      "step": 4915
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.275224107377596,
+      "learning_rate": 9.057374637227299e-06,
+      "loss": 0.9162,
+      "step": 4920
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.384577769247377,
+      "learning_rate": 9.054482553726946e-06,
+      "loss": 0.9366,
+      "step": 4925
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.895425873893831,
+      "learning_rate": 9.051586503616952e-06,
+      "loss": 0.9241,
+      "step": 4930
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.604982979399926,
+      "learning_rate": 9.0486864897306e-06,
+      "loss": 0.9256,
+      "step": 4935
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 22.10558216735927,
+      "learning_rate": 9.045782514905052e-06,
+      "loss": 0.9175,
+      "step": 4940
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 21.66537436134228,
+      "learning_rate": 9.042874581981347e-06,
+      "loss": 0.9241,
+      "step": 4945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 28.46117267926381,
+      "learning_rate": 9.03996269380439e-06,
+      "loss": 0.9401,
+      "step": 4950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 42.00707684144836,
+      "learning_rate": 9.037046853222963e-06,
+      "loss": 0.932,
+      "step": 4955
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 64.63923023317382,
+      "learning_rate": 9.034127063089712e-06,
+      "loss": 0.9438,
+      "step": 4960
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 78.80763586939075,
+      "learning_rate": 9.031203326261144e-06,
+      "loss": 0.9452,
+      "step": 4965
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 75.57620927164031,
+      "learning_rate": 9.028275645597631e-06,
+      "loss": 0.9298,
+      "step": 4970
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.658113859599196,
+      "learning_rate": 9.0253440239634e-06,
+      "loss": 0.923,
+      "step": 4975
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 48.495850283320955,
+      "learning_rate": 9.022408464226541e-06,
+      "loss": 0.9365,
+      "step": 4980
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 37.806366636999435,
+      "learning_rate": 9.019468969258985e-06,
+      "loss": 0.9413,
+      "step": 4985
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 57.21603624921526,
+      "learning_rate": 9.01652554193652e-06,
+      "loss": 0.9333,
+      "step": 4990
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 40.6670860223311,
+      "learning_rate": 9.013578185138784e-06,
+      "loss": 0.9312,
+      "step": 4995
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.668175890753217,
+      "learning_rate": 9.010626901749254e-06,
+      "loss": 0.9295,
+      "step": 5000
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 40.73927855960309,
+      "learning_rate": 9.00767169465525e-06,
+      "loss": 0.9169,
+      "step": 5005
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.109205560365105,
+      "learning_rate": 9.004712566747929e-06,
+      "loss": 0.916,
+      "step": 5010
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.781500255230441,
+      "learning_rate": 9.001749520922289e-06,
+      "loss": 0.9198,
+      "step": 5015
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 25.48428047305956,
+      "learning_rate": 8.998782560077155e-06,
+      "loss": 0.9068,
+      "step": 5020
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 32.98125695456632,
+      "learning_rate": 8.995811687115186e-06,
+      "loss": 0.932,
+      "step": 5025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.97283215980044,
+      "learning_rate": 8.992836904942865e-06,
+      "loss": 0.9379,
+      "step": 5030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.253049260277942,
+      "learning_rate": 8.989858216470507e-06,
+      "loss": 0.918,
+      "step": 5035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.544214252920154,
+      "learning_rate": 8.986875624612236e-06,
+      "loss": 0.913,
+      "step": 5040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 22.04564956071582,
+      "learning_rate": 8.98388913228601e-06,
+      "loss": 0.9178,
+      "step": 5045
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 17.683536399577275,
+      "learning_rate": 8.980898742413587e-06,
+      "loss": 0.9304,
+      "step": 5050
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.161467334972684,
+      "learning_rate": 8.977904457920552e-06,
+      "loss": 0.9145,
+      "step": 5055
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.577177096548642,
+      "learning_rate": 8.974906281736291e-06,
+      "loss": 0.914,
+      "step": 5060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.327368511504124,
+      "learning_rate": 8.971904216794002e-06,
+      "loss": 0.9043,
+      "step": 5065
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 21.406488269474284,
+      "learning_rate": 8.968898266030688e-06,
+      "loss": 0.9202,
+      "step": 5070
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 31.269621520553603,
+      "learning_rate": 8.965888432387147e-06,
+      "loss": 0.9151,
+      "step": 5075
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.802503941578262,
+      "learning_rate": 8.962874718807984e-06,
+      "loss": 0.9097,
+      "step": 5080
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.848981198337093,
+      "learning_rate": 8.959857128241596e-06,
+      "loss": 0.9025,
+      "step": 5085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.072241314901154,
+      "learning_rate": 8.956835663640173e-06,
+      "loss": 0.9138,
+      "step": 5090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.332658464526737,
+      "learning_rate": 8.953810327959693e-06,
+      "loss": 0.9048,
+      "step": 5095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 21.132008030484357,
+      "learning_rate": 8.950781124159926e-06,
+      "loss": 0.9178,
+      "step": 5100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 39.50889372852151,
+      "learning_rate": 8.947748055204424e-06,
+      "loss": 0.9219,
+      "step": 5105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 21.346258805968034,
+      "learning_rate": 8.944711124060519e-06,
+      "loss": 0.9192,
+      "step": 5110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.112949405602714,
+      "learning_rate": 8.941670333699323e-06,
+      "loss": 0.9018,
+      "step": 5115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 35.026352865269004,
+      "learning_rate": 8.938625687095723e-06,
+      "loss": 0.9051,
+      "step": 5120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 21.948641624705495,
+      "learning_rate": 8.93557718722838e-06,
+      "loss": 0.9302,
+      "step": 5125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.460413331641375,
+      "learning_rate": 8.932524837079721e-06,
+      "loss": 0.9094,
+      "step": 5130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 33.16554501484199,
+      "learning_rate": 8.929468639635946e-06,
+      "loss": 0.9112,
+      "step": 5135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 55.507150253651496,
+      "learning_rate": 8.926408597887013e-06,
+      "loss": 0.9076,
+      "step": 5140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 25.672990324546266,
+      "learning_rate": 8.923344714826646e-06,
+      "loss": 0.9157,
+      "step": 5145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.314062818429154,
+      "learning_rate": 8.920276993452319e-06,
+      "loss": 0.9138,
+      "step": 5150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.4944502059902,
+      "learning_rate": 8.917205436765272e-06,
+      "loss": 0.9034,
+      "step": 5155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.607776879666043,
+      "learning_rate": 8.914130047770488e-06,
+      "loss": 0.9163,
+      "step": 5160
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.664686781195138,
+      "learning_rate": 8.911050829476707e-06,
+      "loss": 0.9142,
+      "step": 5165
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.415556266796443,
+      "learning_rate": 8.90796778489641e-06,
+      "loss": 0.9195,
+      "step": 5170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 20.298844004102687,
+      "learning_rate": 8.90488091704582e-06,
+      "loss": 0.9105,
+      "step": 5175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.737019741824055,
+      "learning_rate": 8.901790228944904e-06,
+      "loss": 0.9048,
+      "step": 5180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.18933269117288,
+      "learning_rate": 8.898695723617368e-06,
+      "loss": 0.9123,
+      "step": 5185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 27.076492095174448,
+      "learning_rate": 8.895597404090647e-06,
+      "loss": 0.9173,
+      "step": 5190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.017497657176467,
+      "learning_rate": 8.892495273395913e-06,
+      "loss": 0.9235,
+      "step": 5195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.241184424854241,
+      "learning_rate": 8.889389334568061e-06,
+      "loss": 0.8984,
+      "step": 5200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.896688029270276,
+      "learning_rate": 8.88627959064572e-06,
+      "loss": 0.8982,
+      "step": 5205
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.933863082457536,
+      "learning_rate": 8.883166044671232e-06,
+      "loss": 0.9027,
+      "step": 5210
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 21.301815661889073,
+      "learning_rate": 8.880048699690664e-06,
+      "loss": 0.8975,
+      "step": 5215
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.79520585776361,
+      "learning_rate": 8.876927558753798e-06,
+      "loss": 0.9131,
+      "step": 5220
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 69.54083736297005,
+      "learning_rate": 8.873802624914132e-06,
+      "loss": 0.8882,
+      "step": 5225
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 29.51640291474208,
+      "learning_rate": 8.870673901228874e-06,
+      "loss": 0.8986,
+      "step": 5230
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 48.5416443853786,
+      "learning_rate": 8.867541390758935e-06,
+      "loss": 0.8993,
+      "step": 5235
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 45.16022542591326,
+      "learning_rate": 8.864405096568937e-06,
+      "loss": 0.8991,
+      "step": 5240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 48.08298242632971,
+      "learning_rate": 8.861265021727202e-06,
+      "loss": 0.8939,
+      "step": 5245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.156779794459752,
+      "learning_rate": 8.858121169305747e-06,
+      "loss": 0.9005,
+      "step": 5250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.63891538193591,
+      "learning_rate": 8.854973542380289e-06,
+      "loss": 0.907,
+      "step": 5255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 48.7776713914096,
+      "learning_rate": 8.851822144030237e-06,
+      "loss": 0.9039,
+      "step": 5260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 55.13232465033205,
+      "learning_rate": 8.848666977338689e-06,
+      "loss": 0.8962,
+      "step": 5265
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 29.5918085171386,
+      "learning_rate": 8.84550804539243e-06,
+      "loss": 0.9048,
+      "step": 5270
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 25.783491763675134,
+      "learning_rate": 8.842345351281927e-06,
+      "loss": 0.9006,
+      "step": 5275
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 38.42385751738853,
+      "learning_rate": 8.839178898101327e-06,
+      "loss": 0.9112,
+      "step": 5280
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 17.363290467768564,
+      "learning_rate": 8.836008688948463e-06,
+      "loss": 0.9023,
+      "step": 5285
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.318139570581586,
+      "learning_rate": 8.832834726924832e-06,
+      "loss": 0.9123,
+      "step": 5290
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.274091473033106,
+      "learning_rate": 8.829657015135605e-06,
+      "loss": 0.8881,
+      "step": 5295
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.354741961314343,
+      "learning_rate": 8.826475556689627e-06,
+      "loss": 0.8992,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 22.730814539741452,
+      "learning_rate": 8.8232903546994e-06,
+      "loss": 0.8945,
+      "step": 5305
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.52278714976617,
+      "learning_rate": 8.820101412281098e-06,
+      "loss": 0.9051,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 21.82268336044992,
+      "learning_rate": 8.816908732554546e-06,
+      "loss": 0.8859,
+      "step": 5315
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.685162774578716,
+      "learning_rate": 8.81371231864323e-06,
+      "loss": 0.896,
+      "step": 5320
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.962361513069325,
+      "learning_rate": 8.810512173674288e-06,
+      "loss": 0.8869,
+      "step": 5325
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 17.735749882533618,
+      "learning_rate": 8.807308300778508e-06,
+      "loss": 0.9119,
+      "step": 5330
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 32.718411258492466,
+      "learning_rate": 8.804100703090324e-06,
+      "loss": 0.8902,
+      "step": 5335
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 17.748223147493437,
+      "learning_rate": 8.800889383747817e-06,
+      "loss": 0.8814,
+      "step": 5340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.8215112357413545,
+      "learning_rate": 8.797674345892707e-06,
+      "loss": 0.8917,
+      "step": 5345
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.491577453092495,
+      "learning_rate": 8.794455592670353e-06,
+      "loss": 0.8965,
+      "step": 5350
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 30.099277973463792,
+      "learning_rate": 8.791233127229746e-06,
+      "loss": 0.8815,
+      "step": 5355
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 21.471034023934507,
+      "learning_rate": 8.788006952723512e-06,
+      "loss": 0.9014,
+      "step": 5360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 21.232711801308177,
+      "learning_rate": 8.784777072307904e-06,
+      "loss": 0.892,
+      "step": 5365
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.428510179914167,
+      "learning_rate": 8.781543489142802e-06,
+      "loss": 0.872,
+      "step": 5370
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.45373486754463,
+      "learning_rate": 8.778306206391707e-06,
+      "loss": 0.8938,
+      "step": 5375
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.15825587665837,
+      "learning_rate": 8.775065227221742e-06,
+      "loss": 0.8897,
+      "step": 5380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 20.21159615999797,
+      "learning_rate": 8.77182055480364e-06,
+      "loss": 0.8693,
+      "step": 5385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 20.704802237923644,
+      "learning_rate": 8.768572192311757e-06,
+      "loss": 0.9013,
+      "step": 5390
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.60393581725405,
+      "learning_rate": 8.76532014292405e-06,
+      "loss": 0.8802,
+      "step": 5395
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 32.697742283265754,
+      "learning_rate": 8.762064409822086e-06,
+      "loss": 0.9065,
+      "step": 5400
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 34.84197654856875,
+      "learning_rate": 8.758804996191039e-06,
+      "loss": 0.8728,
+      "step": 5405
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.215924467751915,
+      "learning_rate": 8.75554190521968e-06,
+      "loss": 0.873,
+      "step": 5410
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.416431470943925,
+      "learning_rate": 8.752275140100379e-06,
+      "loss": 0.8839,
+      "step": 5415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.803589736885627,
+      "learning_rate": 8.749004704029101e-06,
+      "loss": 0.8829,
+      "step": 5420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.893782852593526,
+      "learning_rate": 8.745730600205402e-06,
+      "loss": 0.8834,
+      "step": 5425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.70609303531889,
+      "learning_rate": 8.742452831832424e-06,
+      "loss": 0.8986,
+      "step": 5430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 23.16719204178583,
+      "learning_rate": 8.7391714021169e-06,
+      "loss": 0.9119,
+      "step": 5435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.350841301575068,
+      "learning_rate": 8.735886314269136e-06,
+      "loss": 0.8937,
+      "step": 5440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.644162093171502,
+      "learning_rate": 8.732597571503028e-06,
+      "loss": 0.8847,
+      "step": 5445
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 26.775136944455053,
+      "learning_rate": 8.729305177036035e-06,
+      "loss": 0.8868,
+      "step": 5450
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.916084579118998,
+      "learning_rate": 8.726009134089202e-06,
+      "loss": 0.8925,
+      "step": 5455
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 25.97610450325379,
+      "learning_rate": 8.722709445887132e-06,
+      "loss": 0.888,
+      "step": 5460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.674169978898313,
+      "learning_rate": 8.719406115658002e-06,
+      "loss": 0.8882,
+      "step": 5465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.1421190490301,
+      "learning_rate": 8.716099146633548e-06,
+      "loss": 0.8817,
+      "step": 5470
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.709501995828935,
+      "learning_rate": 8.712788542049066e-06,
+      "loss": 0.8745,
+      "step": 5475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.302338450780343,
+      "learning_rate": 8.70947430514341e-06,
+      "loss": 0.8794,
+      "step": 5480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 25.359202453596357,
+      "learning_rate": 8.706156439158988e-06,
+      "loss": 0.8961,
+      "step": 5485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.57404538528837,
+      "learning_rate": 8.702834947341759e-06,
+      "loss": 0.8548,
+      "step": 5490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.621561636096867,
+      "learning_rate": 8.699509832941224e-06,
+      "loss": 0.8827,
+      "step": 5495
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.556414130907989,
+      "learning_rate": 8.696181099210436e-06,
+      "loss": 0.899,
+      "step": 5500
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 20.965519473228266,
+      "learning_rate": 8.692848749405985e-06,
+      "loss": 0.8852,
+      "step": 5505
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.702078440520662,
+      "learning_rate": 8.689512786787996e-06,
+      "loss": 0.8554,
+      "step": 5510
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.307984622928407,
+      "learning_rate": 8.686173214620134e-06,
+      "loss": 0.8872,
+      "step": 5515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 36.1312012987165,
+      "learning_rate": 8.68283003616959e-06,
+      "loss": 0.8548,
+      "step": 5520
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.263820091037312,
+      "learning_rate": 8.679483254707089e-06,
+      "loss": 0.8847,
+      "step": 5525
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.888523618921328,
+      "learning_rate": 8.676132873506873e-06,
+      "loss": 0.8723,
+      "step": 5530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.882036703727629,
+      "learning_rate": 8.672778895846715e-06,
+      "loss": 0.8834,
+      "step": 5535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 20.865318506881103,
+      "learning_rate": 8.669421325007897e-06,
+      "loss": 0.8802,
+      "step": 5540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.716621596417292,
+      "learning_rate": 8.666060164275224e-06,
+      "loss": 0.8843,
+      "step": 5545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.636304054762679,
+      "learning_rate": 8.66269541693701e-06,
+      "loss": 0.8835,
+      "step": 5550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.36733991467472,
+      "learning_rate": 8.659327086285079e-06,
+      "loss": 0.8919,
+      "step": 5555
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 33.84809764114348,
+      "learning_rate": 8.655955175614758e-06,
+      "loss": 0.8774,
+      "step": 5560
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 39.82957262336742,
+      "learning_rate": 8.65257968822488e-06,
+      "loss": 0.8783,
+      "step": 5565
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 67.01059353943249,
+      "learning_rate": 8.649200627417774e-06,
+      "loss": 0.8784,
+      "step": 5570
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 45.26710543417326,
+      "learning_rate": 8.645817996499264e-06,
+      "loss": 0.866,
+      "step": 5575
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 74.55636509273494,
+      "learning_rate": 8.642431798778675e-06,
+      "loss": 0.8906,
+      "step": 5580
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.565055424673996,
+      "learning_rate": 8.63904203756881e-06,
+      "loss": 0.8829,
+      "step": 5585
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 23.61630072608257,
+      "learning_rate": 8.635648716185964e-06,
+      "loss": 0.8635,
+      "step": 5590
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.35054245441266,
+      "learning_rate": 8.632251837949921e-06,
+      "loss": 0.8593,
+      "step": 5595
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.106255161947752,
+      "learning_rate": 8.628851406183932e-06,
+      "loss": 0.884,
+      "step": 5600
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.889113128468237,
+      "learning_rate": 8.625447424214734e-06,
+      "loss": 0.8813,
+      "step": 5605
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.865051244498067,
+      "learning_rate": 8.622039895372533e-06,
+      "loss": 0.8618,
+      "step": 5610
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.349655012279877,
+      "learning_rate": 8.618628822991009e-06,
+      "loss": 0.8655,
+      "step": 5615
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 23.20553059319047,
+      "learning_rate": 8.615214210407304e-06,
+      "loss": 0.8658,
+      "step": 5620
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 30.110412035643414,
+      "learning_rate": 8.611796060962025e-06,
+      "loss": 0.8619,
+      "step": 5625
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.021610994380586,
+      "learning_rate": 8.608374377999242e-06,
+      "loss": 0.8635,
+      "step": 5630
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.71017469605099,
+      "learning_rate": 8.604949164866478e-06,
+      "loss": 0.8625,
+      "step": 5635
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 16.572210414926698,
+      "learning_rate": 8.601520424914712e-06,
+      "loss": 0.8858,
+      "step": 5640
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 22.351789187509702,
+      "learning_rate": 8.598088161498372e-06,
+      "loss": 0.8571,
+      "step": 5645
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 36.86554438746097,
+      "learning_rate": 8.594652377975335e-06,
+      "loss": 0.8658,
+      "step": 5650
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 20.15626490218535,
+      "learning_rate": 8.591213077706918e-06,
+      "loss": 0.88,
+      "step": 5655
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.06297849796565,
+      "learning_rate": 8.587770264057887e-06,
+      "loss": 0.8498,
+      "step": 5660
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.00576879340255,
+      "learning_rate": 8.584323940396435e-06,
+      "loss": 0.8561,
+      "step": 5665
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 39.968686341276495,
+      "learning_rate": 8.580874110094193e-06,
+      "loss": 0.8756,
+      "step": 5670
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 26.27040895403256,
+      "learning_rate": 8.577420776526225e-06,
+      "loss": 0.8655,
+      "step": 5675
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.695621589411635,
+      "learning_rate": 8.57396394307102e-06,
+      "loss": 0.8633,
+      "step": 5680
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.951644924859142,
+      "learning_rate": 8.57050361311049e-06,
+      "loss": 0.8641,
+      "step": 5685
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.871755754368417,
+      "learning_rate": 8.567039790029972e-06,
+      "loss": 0.8441,
+      "step": 5690
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 18.86324287898087,
+      "learning_rate": 8.563572477218216e-06,
+      "loss": 0.8558,
+      "step": 5695
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.9497961055212,
+      "learning_rate": 8.560101678067385e-06,
+      "loss": 0.8626,
+      "step": 5700
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.64541028079422,
+      "learning_rate": 8.55662739597306e-06,
+      "loss": 0.8723,
+      "step": 5705
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.077404639704291,
+      "learning_rate": 8.553149634334221e-06,
+      "loss": 0.8498,
+      "step": 5710
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.095385130597634,
+      "learning_rate": 8.54966839655326e-06,
+      "loss": 0.8676,
+      "step": 5715
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.816198204667515,
+      "learning_rate": 8.546183686035963e-06,
+      "loss": 0.86,
+      "step": 5720
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.324141871325514,
+      "learning_rate": 8.542695506191516e-06,
+      "loss": 0.8453,
+      "step": 5725
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 22.931657761275982,
+      "learning_rate": 8.5392038604325e-06,
+      "loss": 0.8624,
+      "step": 5730
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 22.340937938883243,
+      "learning_rate": 8.535708752174887e-06,
+      "loss": 0.8584,
+      "step": 5735
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.47137594925963,
+      "learning_rate": 8.532210184838034e-06,
+      "loss": 0.8562,
+      "step": 5740
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.022989456393278,
+      "learning_rate": 8.528708161844688e-06,
+      "loss": 0.8517,
+      "step": 5745
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.389404070039904,
+      "learning_rate": 8.525202686620968e-06,
+      "loss": 0.8591,
+      "step": 5750
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.069368671975361,
+      "learning_rate": 8.521693762596376e-06,
+      "loss": 0.8572,
+      "step": 5755
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 31.320280806061785,
+      "learning_rate": 8.518181393203787e-06,
+      "loss": 0.868,
+      "step": 5760
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.119093866352447,
+      "learning_rate": 8.514665581879448e-06,
+      "loss": 0.8533,
+      "step": 5765
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.745899092682233,
+      "learning_rate": 8.511146332062971e-06,
+      "loss": 0.8492,
+      "step": 5770
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.102720102074791,
+      "learning_rate": 8.507623647197334e-06,
+      "loss": 0.8752,
+      "step": 5775
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.17053377877784,
+      "learning_rate": 8.504097530728875e-06,
+      "loss": 0.8411,
+      "step": 5780
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 28.606022516403232,
+      "learning_rate": 8.500567986107286e-06,
+      "loss": 0.8534,
+      "step": 5785
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 27.615463405818634,
+      "learning_rate": 8.497035016785617e-06,
+      "loss": 0.8369,
+      "step": 5790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 36.01795392268902,
+      "learning_rate": 8.49349862622027e-06,
+      "loss": 0.8488,
+      "step": 5795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.213418700089807,
+      "learning_rate": 8.489958817870987e-06,
+      "loss": 0.8535,
+      "step": 5800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.444031330332194,
+      "learning_rate": 8.486415595200862e-06,
+      "loss": 0.8453,
+      "step": 5805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.335189050348509,
+      "learning_rate": 8.482868961676321e-06,
+      "loss": 0.8573,
+      "step": 5810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.00363654906774,
+      "learning_rate": 8.479318920767133e-06,
+      "loss": 0.8515,
+      "step": 5815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.073163283981378,
+      "learning_rate": 8.4757654759464e-06,
+      "loss": 0.8542,
+      "step": 5820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 29.39872449053244,
+      "learning_rate": 8.472208630690553e-06,
+      "loss": 0.8415,
+      "step": 5825
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.515141299600042,
+      "learning_rate": 8.468648388479347e-06,
+      "loss": 0.8381,
+      "step": 5830
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 31.37491484904034,
+      "learning_rate": 8.465084752795867e-06,
+      "loss": 0.8603,
+      "step": 5835
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 58.3350303623845,
+      "learning_rate": 8.461517727126511e-06,
+      "loss": 0.836,
+      "step": 5840
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 29.165439866135134,
+      "learning_rate": 8.457947314960996e-06,
+      "loss": 0.858,
+      "step": 5845
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.285011618464788,
+      "learning_rate": 8.454373519792355e-06,
+      "loss": 0.8491,
+      "step": 5850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.832032196171967,
+      "learning_rate": 8.450796345116926e-06,
+      "loss": 0.8352,
+      "step": 5855
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.440740651835677,
+      "learning_rate": 8.447215794434356e-06,
+      "loss": 0.8607,
+      "step": 5860
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 53.62948227734646,
+      "learning_rate": 8.443631871247598e-06,
+      "loss": 0.8758,
+      "step": 5865
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 40.84158978189534,
+      "learning_rate": 8.440044579062894e-06,
+      "loss": 0.8418,
+      "step": 5870
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.478218209150542,
+      "learning_rate": 8.436453921389791e-06,
+      "loss": 0.8503,
+      "step": 5875
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.378497072180087,
+      "learning_rate": 8.43285990174113e-06,
+      "loss": 0.8671,
+      "step": 5880
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.7445269227027893,
+      "eval_runtime": 75.2523,
+      "eval_samples_per_second": 4.81,
+      "eval_steps_per_second": 0.611,
+      "step": 5882
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 16.074131915930433,
+      "learning_rate": 8.429262523633034e-06,
+      "loss": 0.7902,
+      "step": 5885
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.338805256843147,
+      "learning_rate": 8.425661790584916e-06,
+      "loss": 0.7448,
+      "step": 5890
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 21.507136158145123,
+      "learning_rate": 8.422057706119468e-06,
+      "loss": 0.7342,
+      "step": 5895
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 43.03840171835638,
+      "learning_rate": 8.418450273762665e-06,
+      "loss": 0.7323,
+      "step": 5900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 21.45456264756192,
+      "learning_rate": 8.41483949704376e-06,
+      "loss": 0.7333,
+      "step": 5905
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.793455098771023,
+      "learning_rate": 8.411225379495265e-06,
+      "loss": 0.7466,
+      "step": 5910
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.254994967182895,
+      "learning_rate": 8.407607924652971e-06,
+      "loss": 0.7464,
+      "step": 5915
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 18.7584036810001,
+      "learning_rate": 8.403987136055935e-06,
+      "loss": 0.7299,
+      "step": 5920
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.026506169069334,
+      "learning_rate": 8.40036301724647e-06,
+      "loss": 0.7196,
+      "step": 5925
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.346466183636792,
+      "learning_rate": 8.39673557177015e-06,
+      "loss": 0.7511,
+      "step": 5930
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 18.495927112645926,
+      "learning_rate": 8.3931048031758e-06,
+      "loss": 0.7254,
+      "step": 5935
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.709617231851123,
+      "learning_rate": 8.389470715015501e-06,
+      "loss": 0.7286,
+      "step": 5940
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.780660397689895,
+      "learning_rate": 8.385833310844582e-06,
+      "loss": 0.7139,
+      "step": 5945
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.032615580465922,
+      "learning_rate": 8.382192594221608e-06,
+      "loss": 0.749,
+      "step": 5950
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 25.25163294724016,
+      "learning_rate": 8.378548568708396e-06,
+      "loss": 0.7395,
+      "step": 5955
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 16.169665624561432,
+      "learning_rate": 8.374901237869989e-06,
+      "loss": 0.7303,
+      "step": 5960
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 26.34099151157238,
+      "learning_rate": 8.371250605274673e-06,
+      "loss": 0.7056,
+      "step": 5965
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 32.873244414003075,
+      "learning_rate": 8.367596674493959e-06,
+      "loss": 0.7414,
+      "step": 5970
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 30.87560803986548,
+      "learning_rate": 8.363939449102586e-06,
+      "loss": 0.7282,
+      "step": 5975
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 19.528015029331623,
+      "learning_rate": 8.360278932678515e-06,
+      "loss": 0.7305,
+      "step": 5980
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.445981084181716,
+      "learning_rate": 8.356615128802933e-06,
+      "loss": 0.7212,
+      "step": 5985
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 23.770775974779845,
+      "learning_rate": 8.352948041060234e-06,
+      "loss": 0.7378,
+      "step": 5990
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 19.14504601941375,
+      "learning_rate": 8.349277673038026e-06,
+      "loss": 0.7465,
+      "step": 5995
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 20.290936891320126,
+      "learning_rate": 8.345604028327134e-06,
+      "loss": 0.7314,
+      "step": 6000
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.7028244206432515,
+      "learning_rate": 8.341927110521583e-06,
+      "loss": 0.7185,
+      "step": 6005
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.604020058712843,
+      "learning_rate": 8.3382469232186e-06,
+      "loss": 0.7467,
+      "step": 6010
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 56.19937286099307,
+      "learning_rate": 8.334563470018611e-06,
+      "loss": 0.7333,
+      "step": 6015
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 26.771294283391278,
+      "learning_rate": 8.33087675452524e-06,
+      "loss": 0.7225,
+      "step": 6020
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.734466397931051,
+      "learning_rate": 8.327186780345295e-06,
+      "loss": 0.7247,
+      "step": 6025
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 33.72500753847307,
+      "learning_rate": 8.323493551088782e-06,
+      "loss": 0.7224,
+      "step": 6030
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.90816584525726,
+      "learning_rate": 8.319797070368885e-06,
+      "loss": 0.7334,
+      "step": 6035
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.719459783117856,
+      "learning_rate": 8.316097341801972e-06,
+      "loss": 0.732,
+      "step": 6040
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.500384938602528,
+      "learning_rate": 8.312394369007586e-06,
+      "loss": 0.7176,
+      "step": 6045
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.378871880562029,
+      "learning_rate": 8.308688155608446e-06,
+      "loss": 0.7264,
+      "step": 6050
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.840820667726737,
+      "learning_rate": 8.30497870523044e-06,
+      "loss": 0.7267,
+      "step": 6055
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 22.140046240955762,
+      "learning_rate": 8.301266021502622e-06,
+      "loss": 0.7344,
+      "step": 6060
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 31.539045412586958,
+      "learning_rate": 8.297550108057213e-06,
+      "loss": 0.7417,
+      "step": 6065
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.69385281998031,
+      "learning_rate": 8.293830968529592e-06,
+      "loss": 0.7461,
+      "step": 6070
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.582764103392428,
+      "learning_rate": 8.29010860655829e-06,
+      "loss": 0.7292,
+      "step": 6075
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 22.418600869543923,
+      "learning_rate": 8.286383025784997e-06,
+      "loss": 0.7311,
+      "step": 6080
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.838293990902837,
+      "learning_rate": 8.282654229854547e-06,
+      "loss": 0.7406,
+      "step": 6085
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.173142797379702,
+      "learning_rate": 8.278922222414924e-06,
+      "loss": 0.7331,
+      "step": 6090
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.858068006200584,
+      "learning_rate": 8.275187007117251e-06,
+      "loss": 0.6926,
+      "step": 6095
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 18.922773613667474,
+      "learning_rate": 8.27144858761579e-06,
+      "loss": 0.7242,
+      "step": 6100
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.763980438898749,
+      "learning_rate": 8.267706967567935e-06,
+      "loss": 0.7223,
+      "step": 6105
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.61708664822042,
+      "learning_rate": 8.263962150634215e-06,
+      "loss": 0.7126,
+      "step": 6110
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 14.596927035396877,
+      "learning_rate": 8.26021414047829e-06,
+      "loss": 0.7192,
+      "step": 6115
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.02870706457101,
+      "learning_rate": 8.256462940766932e-06,
+      "loss": 0.7286,
+      "step": 6120
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 36.59195200291681,
+      "learning_rate": 8.252708555170044e-06,
+      "loss": 0.7096,
+      "step": 6125
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 20.281469550974503,
+      "learning_rate": 8.248950987360645e-06,
+      "loss": 0.7167,
+      "step": 6130
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.97594999145645,
+      "learning_rate": 8.245190241014863e-06,
+      "loss": 0.7021,
+      "step": 6135
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.795171570462541,
+      "learning_rate": 8.241426319811938e-06,
+      "loss": 0.749,
+      "step": 6140
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.792290293045824,
+      "learning_rate": 8.237659227434213e-06,
+      "loss": 0.7278,
+      "step": 6145
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.983265764805912,
+      "learning_rate": 8.233888967567141e-06,
+      "loss": 0.7367,
+      "step": 6150
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 18.662478937697124,
+      "learning_rate": 8.230115543899265e-06,
+      "loss": 0.713,
+      "step": 6155
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 46.90862290358351,
+      "learning_rate": 8.226338960122229e-06,
+      "loss": 0.7281,
+      "step": 6160
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 34.89234533567352,
+      "learning_rate": 8.222559219930766e-06,
+      "loss": 0.7333,
+      "step": 6165
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 14.697726604622881,
+      "learning_rate": 8.218776327022696e-06,
+      "loss": 0.7464,
+      "step": 6170
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.747423802614167,
+      "learning_rate": 8.214990285098931e-06,
+      "loss": 0.7376,
+      "step": 6175
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.96902864680576,
+      "learning_rate": 8.211201097863452e-06,
+      "loss": 0.7287,
+      "step": 6180
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 13.636950347264827,
+      "learning_rate": 8.207408769023324e-06,
+      "loss": 0.7094,
+      "step": 6185
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.742565886062385,
+      "learning_rate": 8.203613302288689e-06,
+      "loss": 0.6964,
+      "step": 6190
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 24.16036680358423,
+      "learning_rate": 8.199814701372748e-06,
+      "loss": 0.7124,
+      "step": 6195
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 14.32299199634298,
+      "learning_rate": 8.19601296999178e-06,
+      "loss": 0.7255,
+      "step": 6200
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.085040599511274,
+      "learning_rate": 8.192208111865118e-06,
+      "loss": 0.7077,
+      "step": 6205
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.03400567894429,
+      "learning_rate": 8.188400130715159e-06,
+      "loss": 0.7118,
+      "step": 6210
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.854524776829947,
+      "learning_rate": 8.184589030267353e-06,
+      "loss": 0.7329,
+      "step": 6215
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.850945170965838,
+      "learning_rate": 8.180774814250204e-06,
+      "loss": 0.7138,
+      "step": 6220
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 24.114471404232695,
+      "learning_rate": 8.17695748639526e-06,
+      "loss": 0.7213,
+      "step": 6225
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 29.795134364882006,
+      "learning_rate": 8.173137050437118e-06,
+      "loss": 0.7127,
+      "step": 6230
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.154362218077843,
+      "learning_rate": 8.169313510113413e-06,
+      "loss": 0.7324,
+      "step": 6235
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 25.20196920703092,
+      "learning_rate": 8.165486869164814e-06,
+      "loss": 0.7124,
+      "step": 6240
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.811823800608863,
+      "learning_rate": 8.161657131335033e-06,
+      "loss": 0.7104,
+      "step": 6245
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 17.93546314393749,
+      "learning_rate": 8.157824300370801e-06,
+      "loss": 0.6941,
+      "step": 6250
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 33.16849346210743,
+      "learning_rate": 8.153988380021881e-06,
+      "loss": 0.7347,
+      "step": 6255
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 24.269651699560313,
+      "learning_rate": 8.150149374041061e-06,
+      "loss": 0.7379,
+      "step": 6260
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.381379260049865,
+      "learning_rate": 8.146307286184141e-06,
+      "loss": 0.726,
+      "step": 6265
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 12.343196756170697,
+      "learning_rate": 8.14246212020994e-06,
+      "loss": 0.7178,
+      "step": 6270
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.578717720187098,
+      "learning_rate": 8.138613879880284e-06,
+      "loss": 0.7137,
+      "step": 6275
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 31.370528526498276,
+      "learning_rate": 8.134762568960015e-06,
+      "loss": 0.7106,
+      "step": 6280
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.249042174913429,
+      "learning_rate": 8.130908191216974e-06,
+      "loss": 0.7089,
+      "step": 6285
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 23.809066454954635,
+      "learning_rate": 8.127050750422e-06,
+      "loss": 0.7045,
+      "step": 6290
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.854982995546571,
+      "learning_rate": 8.123190250348932e-06,
+      "loss": 0.7164,
+      "step": 6295
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 17.7265251355763,
+      "learning_rate": 8.119326694774602e-06,
+      "loss": 0.7012,
+      "step": 6300
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 17.579992645590757,
+      "learning_rate": 8.115460087478833e-06,
+      "loss": 0.7404,
+      "step": 6305
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 25.041819648349822,
+      "learning_rate": 8.11159043224443e-06,
+      "loss": 0.7015,
+      "step": 6310
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 24.949195747447597,
+      "learning_rate": 8.107717732857177e-06,
+      "loss": 0.6872,
+      "step": 6315
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 19.679411460396835,
+      "learning_rate": 8.103841993105843e-06,
+      "loss": 0.7072,
+      "step": 6320
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 20.012872260255527,
+      "learning_rate": 8.099963216782171e-06,
+      "loss": 0.7031,
+      "step": 6325
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 28.27758443192042,
+      "learning_rate": 8.09608140768087e-06,
+      "loss": 0.7259,
+      "step": 6330
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 24.772908515131697,
+      "learning_rate": 8.092196569599619e-06,
+      "loss": 0.7219,
+      "step": 6335
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.0503438246646,
+      "learning_rate": 8.08830870633906e-06,
+      "loss": 0.7279,
+      "step": 6340
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.318031610520302,
+      "learning_rate": 8.084417821702796e-06,
+      "loss": 0.7082,
+      "step": 6345
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 35.01375125895254,
+      "learning_rate": 8.080523919497381e-06,
+      "loss": 0.7282,
+      "step": 6350
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 20.21136986925559,
+      "learning_rate": 8.076627003532328e-06,
+      "loss": 0.7057,
+      "step": 6355
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.017121017069089,
+      "learning_rate": 8.072727077620092e-06,
+      "loss": 0.7136,
+      "step": 6360
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 27.32813819885369,
+      "learning_rate": 8.068824145576077e-06,
+      "loss": 0.7499,
+      "step": 6365
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 101.56111015961008,
+      "learning_rate": 8.064918211218628e-06,
+      "loss": 0.7211,
+      "step": 6370
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 15.904740460926877,
+      "learning_rate": 8.061009278369026e-06,
+      "loss": 0.7587,
+      "step": 6375
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.262944951104998,
+      "learning_rate": 8.05709735085148e-06,
+      "loss": 0.76,
+      "step": 6380
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 19.900012124876444,
+      "learning_rate": 8.053182432493141e-06,
+      "loss": 0.7362,
+      "step": 6385
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.601924165265274,
+      "learning_rate": 8.049264527124076e-06,
+      "loss": 0.7501,
+      "step": 6390
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 21.08025285119702,
+      "learning_rate": 8.045343638577278e-06,
+      "loss": 0.7563,
+      "step": 6395
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 20.646288499921937,
+      "learning_rate": 8.04141977068866e-06,
+      "loss": 0.7316,
+      "step": 6400
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.096270740741641,
+      "learning_rate": 8.037492927297044e-06,
+      "loss": 0.7381,
+      "step": 6405
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.580229233187346,
+      "learning_rate": 8.033563112244172e-06,
+      "loss": 0.7287,
+      "step": 6410
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 18.28636073508511,
+      "learning_rate": 8.02963032937468e-06,
+      "loss": 0.7352,
+      "step": 6415
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.346063474119852,
+      "learning_rate": 8.025694582536124e-06,
+      "loss": 0.7332,
+      "step": 6420
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.246364151412761,
+      "learning_rate": 8.021755875578945e-06,
+      "loss": 0.71,
+      "step": 6425
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.367025054003026,
+      "learning_rate": 8.017814212356492e-06,
+      "loss": 0.725,
+      "step": 6430
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.714057040582556,
+      "learning_rate": 8.013869596724994e-06,
+      "loss": 0.7379,
+      "step": 6435
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.593995045394287,
+      "learning_rate": 8.009922032543581e-06,
+      "loss": 0.7415,
+      "step": 6440
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 26.587621305434546,
+      "learning_rate": 8.005971523674257e-06,
+      "loss": 0.7186,
+      "step": 6445
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 27.706587902467838,
+      "learning_rate": 8.002018073981914e-06,
+      "loss": 0.7022,
+      "step": 6450
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 21.224412769515773,
+      "learning_rate": 7.998061687334318e-06,
+      "loss": 0.7056,
+      "step": 6455
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 14.643126663204766,
+      "learning_rate": 7.994102367602107e-06,
+      "loss": 0.7126,
+      "step": 6460
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.286326181980531,
+      "learning_rate": 7.990140118658792e-06,
+      "loss": 0.7317,
+      "step": 6465
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 26.733025687439078,
+      "learning_rate": 7.986174944380749e-06,
+      "loss": 0.7321,
+      "step": 6470
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 19.564773455771654,
+      "learning_rate": 7.982206848647212e-06,
+      "loss": 0.7268,
+      "step": 6475
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.448541273732022,
+      "learning_rate": 7.978235835340277e-06,
+      "loss": 0.744,
+      "step": 6480
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 36.37395815875439,
+      "learning_rate": 7.974261908344896e-06,
+      "loss": 0.6985,
+      "step": 6485
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 29.85056971539694,
+      "learning_rate": 7.970285071548868e-06,
+      "loss": 0.7214,
+      "step": 6490
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 22.471371403927865,
+      "learning_rate": 7.966305328842838e-06,
+      "loss": 0.6981,
+      "step": 6495
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 29.88202765503994,
+      "learning_rate": 7.962322684120295e-06,
+      "loss": 0.6977,
+      "step": 6500
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 17.33714463691604,
+      "learning_rate": 7.95833714127757e-06,
+      "loss": 0.7451,
+      "step": 6505
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.99959408887192,
+      "learning_rate": 7.954348704213825e-06,
+      "loss": 0.7204,
+      "step": 6510
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.049862478420733,
+      "learning_rate": 7.95035737683106e-06,
+      "loss": 0.7168,
+      "step": 6515
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.518802904427606,
+      "learning_rate": 7.946363163034092e-06,
+      "loss": 0.7031,
+      "step": 6520
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.273603677482322,
+      "learning_rate": 7.942366066730571e-06,
+      "loss": 0.7324,
+      "step": 6525
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 22.046757543422785,
+      "learning_rate": 7.938366091830967e-06,
+      "loss": 0.7265,
+      "step": 6530
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 33.12928987146557,
+      "learning_rate": 7.93436324224856e-06,
+      "loss": 0.7145,
+      "step": 6535
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.56350041819999,
+      "learning_rate": 7.930357521899444e-06,
+      "loss": 0.7149,
+      "step": 6540
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.501911722744474,
+      "learning_rate": 7.926348934702526e-06,
+      "loss": 0.6998,
+      "step": 6545
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.688733562201573,
+      "learning_rate": 7.922337484579516e-06,
+      "loss": 0.7107,
+      "step": 6550
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 21.058453052713165,
+      "learning_rate": 7.918323175454923e-06,
+      "loss": 0.7332,
+      "step": 6555
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 38.17079743809706,
+      "learning_rate": 7.914306011256051e-06,
+      "loss": 0.6928,
+      "step": 6560
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 67.58525110577213,
+      "learning_rate": 7.910285995913003e-06,
+      "loss": 0.7347,
+      "step": 6565
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 32.44302274542706,
+      "learning_rate": 7.90626313335867e-06,
+      "loss": 0.7164,
+      "step": 6570
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 20.402541349672042,
+      "learning_rate": 7.902237427528721e-06,
+      "loss": 0.7061,
+      "step": 6575
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.055494830630643,
+      "learning_rate": 7.89820888236162e-06,
+      "loss": 0.7074,
+      "step": 6580
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.66556604728358,
+      "learning_rate": 7.894177501798595e-06,
+      "loss": 0.7163,
+      "step": 6585
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.07830415040536,
+      "learning_rate": 7.890143289783658e-06,
+      "loss": 0.7098,
+      "step": 6590
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.736714344164882,
+      "learning_rate": 7.886106250263588e-06,
+      "loss": 0.7215,
+      "step": 6595
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.6673812587513,
+      "learning_rate": 7.882066387187926e-06,
+      "loss": 0.7087,
+      "step": 6600
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 27.162715320340062,
+      "learning_rate": 7.87802370450898e-06,
+      "loss": 0.703,
+      "step": 6605
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 17.595778017849216,
+      "learning_rate": 7.87397820618182e-06,
+      "loss": 0.732,
+      "step": 6610
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.497247310820503,
+      "learning_rate": 7.869929896164262e-06,
+      "loss": 0.6938,
+      "step": 6615
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 24.174550408631685,
+      "learning_rate": 7.865878778416879e-06,
+      "loss": 0.7326,
+      "step": 6620
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.682620191492965,
+      "learning_rate": 7.861824856902984e-06,
+      "loss": 0.681,
+      "step": 6625
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.042295074217735,
+      "learning_rate": 7.857768135588642e-06,
+      "loss": 0.7225,
+      "step": 6630
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 23.16669446157789,
+      "learning_rate": 7.853708618442654e-06,
+      "loss": 0.7164,
+      "step": 6635
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.921927166023332,
+      "learning_rate": 7.849646309436551e-06,
+      "loss": 0.6961,
+      "step": 6640
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.13787412696501,
+      "learning_rate": 7.845581212544605e-06,
+      "loss": 0.7325,
+      "step": 6645
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.317276381618884,
+      "learning_rate": 7.841513331743803e-06,
+      "loss": 0.7118,
+      "step": 6650
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.678166525281911,
+      "learning_rate": 7.837442671013868e-06,
+      "loss": 0.7186,
+      "step": 6655
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.039215360240258,
+      "learning_rate": 7.833369234337235e-06,
+      "loss": 0.7159,
+      "step": 6660
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.15585567306351,
+      "learning_rate": 7.829293025699056e-06,
+      "loss": 0.72,
+      "step": 6665
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.964575038033085,
+      "learning_rate": 7.8252140490872e-06,
+      "loss": 0.6964,
+      "step": 6670
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 31.392270596686796,
+      "learning_rate": 7.821132308492235e-06,
+      "loss": 0.725,
+      "step": 6675
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 22.243486431851114,
+      "learning_rate": 7.81704780790744e-06,
+      "loss": 0.7163,
+      "step": 6680
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 32.554006910354886,
+      "learning_rate": 7.812960551328792e-06,
+      "loss": 0.7215,
+      "step": 6685
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.860686728562028,
+      "learning_rate": 7.808870542754964e-06,
+      "loss": 0.7126,
+      "step": 6690
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.717346410772016,
+      "learning_rate": 7.804777786187324e-06,
+      "loss": 0.7107,
+      "step": 6695
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.488764788651848,
+      "learning_rate": 7.800682285629922e-06,
+      "loss": 0.7108,
+      "step": 6700
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 20.458748569916793,
+      "learning_rate": 7.796584045089499e-06,
+      "loss": 0.7195,
+      "step": 6705
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 15.795348738415434,
+      "learning_rate": 7.792483068575475e-06,
+      "loss": 0.7119,
+      "step": 6710
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 24.613194371981375,
+      "learning_rate": 7.788379360099944e-06,
+      "loss": 0.7097,
+      "step": 6715
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 16.381433477062384,
+      "learning_rate": 7.784272923677678e-06,
+      "loss": 0.6991,
+      "step": 6720
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.110548394689014,
+      "learning_rate": 7.78016376332611e-06,
+      "loss": 0.7046,
+      "step": 6725
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 15.048488403438107,
+      "learning_rate": 7.776051883065345e-06,
+      "loss": 0.7019,
+      "step": 6730
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 29.011311668426053,
+      "learning_rate": 7.771937286918147e-06,
+      "loss": 0.7244,
+      "step": 6735
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 28.93610413762296,
+      "learning_rate": 7.767819978909933e-06,
+      "loss": 0.6931,
+      "step": 6740
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.355017448756705,
+      "learning_rate": 7.763699963068782e-06,
+      "loss": 0.7062,
+      "step": 6745
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.110861457861793,
+      "learning_rate": 7.759577243425412e-06,
+      "loss": 0.6952,
+      "step": 6750
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.727622562247888,
+      "learning_rate": 7.755451824013194e-06,
+      "loss": 0.7046,
+      "step": 6755
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 25.70418981169325,
+      "learning_rate": 7.751323708868134e-06,
+      "loss": 0.683,
+      "step": 6760
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 25.78166214567401,
+      "learning_rate": 7.74719290202888e-06,
+      "loss": 0.6869,
+      "step": 6765
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.622497117699039,
+      "learning_rate": 7.74305940753671e-06,
+      "loss": 0.6902,
+      "step": 6770
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 39.78003913153285,
+      "learning_rate": 7.738923229435538e-06,
+      "loss": 0.7013,
+      "step": 6775
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 45.810278472194724,
+      "learning_rate": 7.734784371771894e-06,
+      "loss": 0.7171,
+      "step": 6780
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 23.054568253500427,
+      "learning_rate": 7.730642838594932e-06,
+      "loss": 0.692,
+      "step": 6785
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.059784782634388,
+      "learning_rate": 7.726498633956433e-06,
+      "loss": 0.6867,
+      "step": 6790
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.177875706336556,
+      "learning_rate": 7.72235176191078e-06,
+      "loss": 0.7012,
+      "step": 6795
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.4495938243906625,
+      "learning_rate": 7.718202226514968e-06,
+      "loss": 0.7038,
+      "step": 6800
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.44562019083214,
+      "learning_rate": 7.714050031828602e-06,
+      "loss": 0.6935,
+      "step": 6805
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.812788788197329,
+      "learning_rate": 7.709895181913887e-06,
+      "loss": 0.725,
+      "step": 6810
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.336833846004621,
+      "learning_rate": 7.705737680835623e-06,
+      "loss": 0.6925,
+      "step": 6815
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.987925326739377,
+      "learning_rate": 7.701577532661204e-06,
+      "loss": 0.6947,
+      "step": 6820
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.87671922318729,
+      "learning_rate": 7.697414741460616e-06,
+      "loss": 0.687,
+      "step": 6825
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.712200059941566,
+      "learning_rate": 7.693249311306433e-06,
+      "loss": 0.6904,
+      "step": 6830
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.774310159518635,
+      "learning_rate": 7.689081246273805e-06,
+      "loss": 0.6892,
+      "step": 6835
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.661289555094946,
+      "learning_rate": 7.684910550440462e-06,
+      "loss": 0.6999,
+      "step": 6840
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 33.664597069197804,
+      "learning_rate": 7.680737227886708e-06,
+      "loss": 0.7047,
+      "step": 6845
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 20.181955771404265,
+      "learning_rate": 7.67656128269542e-06,
+      "loss": 0.7261,
+      "step": 6850
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 30.26147343843953,
+      "learning_rate": 7.672382718952037e-06,
+      "loss": 0.7052,
+      "step": 6855
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.667253974389084,
+      "learning_rate": 7.668201540744556e-06,
+      "loss": 0.711,
+      "step": 6860
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 42.630822600609775,
+      "learning_rate": 7.664017752163542e-06,
+      "loss": 0.7036,
+      "step": 6865
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.345967377044545,
+      "learning_rate": 7.659831357302105e-06,
+      "loss": 0.6998,
+      "step": 6870
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.85298791802671,
+      "learning_rate": 7.655642360255911e-06,
+      "loss": 0.6963,
+      "step": 6875
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.02513471044103,
+      "learning_rate": 7.651450765123165e-06,
+      "loss": 0.6913,
+      "step": 6880
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 21.157869645277824,
+      "learning_rate": 7.647256576004619e-06,
+      "loss": 0.6693,
+      "step": 6885
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 22.99490929473168,
+      "learning_rate": 7.643059797003564e-06,
+      "loss": 0.6849,
+      "step": 6890
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 27.020744662631394,
+      "learning_rate": 7.638860432225818e-06,
+      "loss": 0.723,
+      "step": 6895
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.624030473466098,
+      "learning_rate": 7.634658485779736e-06,
+      "loss": 0.6946,
+      "step": 6900
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.65357246747531,
+      "learning_rate": 7.630453961776195e-06,
+      "loss": 0.6892,
+      "step": 6905
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.093852140081328,
+      "learning_rate": 7.6262468643285926e-06,
+      "loss": 0.6927,
+      "step": 6910
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.094869879189778,
+      "learning_rate": 7.622037197552846e-06,
+      "loss": 0.6913,
+      "step": 6915
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.165975514550444,
+      "learning_rate": 7.617824965567387e-06,
+      "loss": 0.7019,
+      "step": 6920
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 19.50880196493642,
+      "learning_rate": 7.613610172493156e-06,
+      "loss": 0.7077,
+      "step": 6925
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.672651849911297,
+      "learning_rate": 7.609392822453596e-06,
+      "loss": 0.7017,
+      "step": 6930
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.164537307950836,
+      "learning_rate": 7.605172919574657e-06,
+      "loss": 0.7015,
+      "step": 6935
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.451099827206772,
+      "learning_rate": 7.600950467984783e-06,
+      "loss": 0.6752,
+      "step": 6940
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.870539162624986,
+      "learning_rate": 7.5967254718149145e-06,
+      "loss": 0.6853,
+      "step": 6945
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.135503893996802,
+      "learning_rate": 7.592497935198474e-06,
+      "loss": 0.6946,
+      "step": 6950
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.045389341400304,
+      "learning_rate": 7.588267862271379e-06,
+      "loss": 0.7116,
+      "step": 6955
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.973021024474602,
+      "learning_rate": 7.584035257172022e-06,
+      "loss": 0.7072,
+      "step": 6960
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.220087665017225,
+      "learning_rate": 7.579800124041276e-06,
+      "loss": 0.6944,
+      "step": 6965
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 39.31130380483513,
+      "learning_rate": 7.575562467022484e-06,
+      "loss": 0.701,
+      "step": 6970
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 33.3590421783122,
+      "learning_rate": 7.571322290261462e-06,
+      "loss": 0.7133,
+      "step": 6975
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 43.2227104390999,
+      "learning_rate": 7.567079597906491e-06,
+      "loss": 0.7223,
+      "step": 6980
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 52.20311146641715,
+      "learning_rate": 7.5628343941083074e-06,
+      "loss": 0.7041,
+      "step": 6985
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 69.35024918310937,
+      "learning_rate": 7.5585866830201086e-06,
+      "loss": 0.6854,
+      "step": 6990
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 26.03096173960646,
+      "learning_rate": 7.554336468797549e-06,
+      "loss": 0.6967,
+      "step": 6995
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.747268228543884,
+      "learning_rate": 7.550083755598723e-06,
+      "loss": 0.6989,
+      "step": 7000
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 33.98837608030512,
+      "learning_rate": 7.5458285475841784e-06,
+      "loss": 0.6988,
+      "step": 7005
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 30.343042257903118,
+      "learning_rate": 7.541570848916898e-06,
+      "loss": 0.6924,
+      "step": 7010
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.01218820770515,
+      "learning_rate": 7.537310663762305e-06,
+      "loss": 0.6964,
+      "step": 7015
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.42358496227341,
+      "learning_rate": 7.533047996288252e-06,
+      "loss": 0.6913,
+      "step": 7020
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.964543921746909,
+      "learning_rate": 7.528782850665021e-06,
+      "loss": 0.7022,
+      "step": 7025
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 25.8037225474705,
+      "learning_rate": 7.524515231065321e-06,
+      "loss": 0.6785,
+      "step": 7030
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.350765998841295,
+      "learning_rate": 7.520245141664278e-06,
+      "loss": 0.6869,
+      "step": 7035
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.861735010461002,
+      "learning_rate": 7.515972586639435e-06,
+      "loss": 0.6799,
+      "step": 7040
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.877217521620906,
+      "learning_rate": 7.511697570170748e-06,
+      "loss": 0.7014,
+      "step": 7045
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.505481993998394,
+      "learning_rate": 7.507420096440583e-06,
+      "loss": 0.6984,
+      "step": 7050
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.778863037826613,
+      "learning_rate": 7.5031401696337066e-06,
+      "loss": 0.6979,
+      "step": 7055
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.471077209038233,
+      "learning_rate": 7.498857793937286e-06,
+      "loss": 0.6974,
+      "step": 7060
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 36.567832461479874,
+      "learning_rate": 7.494572973540886e-06,
+      "loss": 0.6899,
+      "step": 7065
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.43230031875812,
+      "learning_rate": 7.4902857126364605e-06,
+      "loss": 0.6821,
+      "step": 7070
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 24.623961517100078,
+      "learning_rate": 7.485996015418354e-06,
+      "loss": 0.6922,
+      "step": 7075
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.646515546307073,
+      "learning_rate": 7.481703886083291e-06,
+      "loss": 0.6745,
+      "step": 7080
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 31.396337867305935,
+      "learning_rate": 7.477409328830381e-06,
+      "loss": 0.6928,
+      "step": 7085
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 20.424188079541075,
+      "learning_rate": 7.473112347861103e-06,
+      "loss": 0.6722,
+      "step": 7090
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 26.00232330630306,
+      "learning_rate": 7.468812947379307e-06,
+      "loss": 0.6998,
+      "step": 7095
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 27.46957886840992,
+      "learning_rate": 7.464511131591216e-06,
+      "loss": 0.6846,
+      "step": 7100
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.132272963335847,
+      "learning_rate": 7.4602069047054105e-06,
+      "loss": 0.6982,
+      "step": 7105
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 22.348386685560573,
+      "learning_rate": 7.4559002709328335e-06,
+      "loss": 0.6851,
+      "step": 7110
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.814202412904162,
+      "learning_rate": 7.451591234486779e-06,
+      "loss": 0.7055,
+      "step": 7115
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.859238788081717,
+      "learning_rate": 7.447279799582895e-06,
+      "loss": 0.6921,
+      "step": 7120
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.588006740632137,
+      "learning_rate": 7.442965970439175e-06,
+      "loss": 0.6588,
+      "step": 7125
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.47004331535925,
+      "learning_rate": 7.438649751275952e-06,
+      "loss": 0.6859,
+      "step": 7130
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.497100321839717,
+      "learning_rate": 7.434331146315903e-06,
+      "loss": 0.7088,
+      "step": 7135
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.389468678247884,
+      "learning_rate": 7.430010159784032e-06,
+      "loss": 0.6818,
+      "step": 7140
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 35.137652159527434,
+      "learning_rate": 7.425686795907678e-06,
+      "loss": 0.6927,
+      "step": 7145
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 24.705252151791072,
+      "learning_rate": 7.4213610589165055e-06,
+      "loss": 0.7036,
+      "step": 7150
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.207253216482892,
+      "learning_rate": 7.4170329530425e-06,
+      "loss": 0.6846,
+      "step": 7155
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.846196135712745,
+      "learning_rate": 7.412702482519962e-06,
+      "loss": 0.6822,
+      "step": 7160
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.97292030080821,
+      "learning_rate": 7.40836965158551e-06,
+      "loss": 0.6978,
+      "step": 7165
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 21.994990939537736,
+      "learning_rate": 7.404034464478069e-06,
+      "loss": 0.7095,
+      "step": 7170
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.622262894881901,
+      "learning_rate": 7.399696925438868e-06,
+      "loss": 0.6591,
+      "step": 7175
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.138206921967175,
+      "learning_rate": 7.39535703871144e-06,
+      "loss": 0.6758,
+      "step": 7180
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.35115037571887,
+      "learning_rate": 7.391014808541611e-06,
+      "loss": 0.6926,
+      "step": 7185
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.995736156812953,
+      "learning_rate": 7.386670239177504e-06,
+      "loss": 0.6809,
+      "step": 7190
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.45647494417984,
+      "learning_rate": 7.382323334869529e-06,
+      "loss": 0.6822,
+      "step": 7195
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 24.7016547776482,
+      "learning_rate": 7.377974099870378e-06,
+      "loss": 0.6785,
+      "step": 7200
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 23.36867960192761,
+      "learning_rate": 7.373622538435024e-06,
+      "loss": 0.6808,
+      "step": 7205
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 32.65941997390678,
+      "learning_rate": 7.369268654820718e-06,
+      "loss": 0.6907,
+      "step": 7210
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 19.77258580914987,
+      "learning_rate": 7.3649124532869855e-06,
+      "loss": 0.702,
+      "step": 7215
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.899504710148166,
+      "learning_rate": 7.36055393809561e-06,
+      "loss": 0.684,
+      "step": 7220
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.847634728164552,
+      "learning_rate": 7.356193113510648e-06,
+      "loss": 0.687,
+      "step": 7225
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 20.93469198208261,
+      "learning_rate": 7.3518299837984095e-06,
+      "loss": 0.6607,
+      "step": 7230
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 44.311453507879676,
+      "learning_rate": 7.347464553227466e-06,
+      "loss": 0.6924,
+      "step": 7235
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.396465443143537,
+      "learning_rate": 7.343096826068631e-06,
+      "loss": 0.7023,
+      "step": 7240
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.15469275451865,
+      "learning_rate": 7.338726806594973e-06,
+      "loss": 0.684,
+      "step": 7245
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.59304430753221,
+      "learning_rate": 7.334354499081797e-06,
+      "loss": 0.7031,
+      "step": 7250
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 15.2512260575335,
+      "learning_rate": 7.329979907806655e-06,
+      "loss": 0.6959,
+      "step": 7255
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.326978416349075,
+      "learning_rate": 7.3256030370493216e-06,
+      "loss": 0.6902,
+      "step": 7260
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.110584046270967,
+      "learning_rate": 7.321223891091811e-06,
+      "loss": 0.6801,
+      "step": 7265
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.953504528002581,
+      "learning_rate": 7.316842474218357e-06,
+      "loss": 0.6851,
+      "step": 7270
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.575325837437997,
+      "learning_rate": 7.312458790715423e-06,
+      "loss": 0.6897,
+      "step": 7275
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 21.78543946475332,
+      "learning_rate": 7.308072844871679e-06,
+      "loss": 0.6691,
+      "step": 7280
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.868152052821847,
+      "learning_rate": 7.3036846409780175e-06,
+      "loss": 0.6942,
+      "step": 7285
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.721946329731745,
+      "learning_rate": 7.299294183327534e-06,
+      "loss": 0.6692,
+      "step": 7290
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.006162810748249,
+      "learning_rate": 7.294901476215537e-06,
+      "loss": 0.6773,
+      "step": 7295
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.387648611571954,
+      "learning_rate": 7.290506523939524e-06,
+      "loss": 0.6833,
+      "step": 7300
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.110771012553165,
+      "learning_rate": 7.286109330799198e-06,
+      "loss": 0.6709,
+      "step": 7305
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.516727078372885,
+      "learning_rate": 7.2817099010964545e-06,
+      "loss": 0.6775,
+      "step": 7310
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 38.802526080079346,
+      "learning_rate": 7.277308239135369e-06,
+      "loss": 0.7053,
+      "step": 7315
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 32.465664109302125,
+      "learning_rate": 7.27290434922221e-06,
+      "loss": 0.6796,
+      "step": 7320
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.845671379668499,
+      "learning_rate": 7.26849823566542e-06,
+      "loss": 0.7167,
+      "step": 7325
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.518712361079984,
+      "learning_rate": 7.264089902775618e-06,
+      "loss": 0.7029,
+      "step": 7330
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.250471903999252,
+      "learning_rate": 7.259679354865599e-06,
+      "loss": 0.6757,
+      "step": 7335
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.190619690981686,
+      "learning_rate": 7.255266596250316e-06,
+      "loss": 0.6919,
+      "step": 7340
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 25.288326771663677,
+      "learning_rate": 7.250851631246892e-06,
+      "loss": 0.6594,
+      "step": 7345
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.247667804652537,
+      "learning_rate": 7.246434464174607e-06,
+      "loss": 0.6532,
+      "step": 7350
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.682710149971358,
+      "learning_rate": 7.242015099354894e-06,
+      "loss": 0.6585,
+      "step": 7355
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 27.239771872923306,
+      "learning_rate": 7.2375935411113375e-06,
+      "loss": 0.6923,
+      "step": 7360
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.142724413466496,
+      "learning_rate": 7.2331697937696654e-06,
+      "loss": 0.6701,
+      "step": 7365
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 19.053316941159274,
+      "learning_rate": 7.2287438616577496e-06,
+      "loss": 0.6655,
+      "step": 7370
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.044116655990347,
+      "learning_rate": 7.2243157491056e-06,
+      "loss": 0.6751,
+      "step": 7375
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 23.211289117451855,
+      "learning_rate": 7.219885460445358e-06,
+      "loss": 0.6909,
+      "step": 7380
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.716287423668646,
+      "learning_rate": 7.215453000011295e-06,
+      "loss": 0.6851,
+      "step": 7385
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.994891603259087,
+      "learning_rate": 7.211018372139804e-06,
+      "loss": 0.6716,
+      "step": 7390
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.834328037352239,
+      "learning_rate": 7.2065815811694055e-06,
+      "loss": 0.6806,
+      "step": 7395
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.668936421997721,
+      "learning_rate": 7.202142631440728e-06,
+      "loss": 0.6895,
+      "step": 7400
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.712019148611033,
+      "learning_rate": 7.197701527296518e-06,
+      "loss": 0.6713,
+      "step": 7405
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.575949154287882,
+      "learning_rate": 7.193258273081626e-06,
+      "loss": 0.6635,
+      "step": 7410
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.801693793087312,
+      "learning_rate": 7.188812873143007e-06,
+      "loss": 0.6769,
+      "step": 7415
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.726495724148158,
+      "learning_rate": 7.184365331829719e-06,
+      "loss": 0.6822,
+      "step": 7420
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.831922759330421,
+      "learning_rate": 7.179915653492907e-06,
+      "loss": 0.6427,
+      "step": 7425
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.151607114346472,
+      "learning_rate": 7.175463842485815e-06,
+      "loss": 0.6901,
+      "step": 7430
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 34.95697762852432,
+      "learning_rate": 7.171009903163767e-06,
+      "loss": 0.6907,
+      "step": 7435
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 20.584668811325557,
+      "learning_rate": 7.1665538398841746e-06,
+      "loss": 0.6821,
+      "step": 7440
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.792833479274165,
+      "learning_rate": 7.162095657006523e-06,
+      "loss": 0.6822,
+      "step": 7445
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 32.94215443286715,
+      "learning_rate": 7.157635358892374e-06,
+      "loss": 0.6733,
+      "step": 7450
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 46.902507165291105,
+      "learning_rate": 7.153172949905357e-06,
+      "loss": 0.685,
+      "step": 7455
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 37.05914931544486,
+      "learning_rate": 7.1487084344111665e-06,
+      "loss": 0.7108,
+      "step": 7460
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.901869763596967,
+      "learning_rate": 7.144241816777559e-06,
+      "loss": 0.678,
+      "step": 7465
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 31.0258209958262,
+      "learning_rate": 7.139773101374346e-06,
+      "loss": 0.6655,
+      "step": 7470
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.571638831792662,
+      "learning_rate": 7.135302292573392e-06,
+      "loss": 0.6815,
+      "step": 7475
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.721170462197327,
+      "learning_rate": 7.130829394748613e-06,
+      "loss": 0.6796,
+      "step": 7480
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 28.939363425798188,
+      "learning_rate": 7.126354412275963e-06,
+      "loss": 0.6667,
+      "step": 7485
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 24.334434659753086,
+      "learning_rate": 7.121877349533438e-06,
+      "loss": 0.6793,
+      "step": 7490
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.334053700215673,
+      "learning_rate": 7.117398210901071e-06,
+      "loss": 0.6796,
+      "step": 7495
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.658342801586011,
+      "learning_rate": 7.112917000760923e-06,
+      "loss": 0.668,
+      "step": 7500
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.31412533720228,
+      "learning_rate": 7.108433723497081e-06,
+      "loss": 0.6665,
+      "step": 7505
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.405988971024291,
+      "learning_rate": 7.103948383495659e-06,
+      "loss": 0.6748,
+      "step": 7510
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.334245035577876,
+      "learning_rate": 7.099460985144784e-06,
+      "loss": 0.682,
+      "step": 7515
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 35.91501841354662,
+      "learning_rate": 7.094971532834601e-06,
+      "loss": 0.6485,
+      "step": 7520
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.892892396192009,
+      "learning_rate": 7.090480030957261e-06,
+      "loss": 0.6903,
+      "step": 7525
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 17.90987688639934,
+      "learning_rate": 7.0859864839069205e-06,
+      "loss": 0.6636,
+      "step": 7530
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 21.575729746016865,
+      "learning_rate": 7.081490896079738e-06,
+      "loss": 0.6779,
+      "step": 7535
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.673927795129549,
+      "learning_rate": 7.076993271873871e-06,
+      "loss": 0.6817,
+      "step": 7540
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.597632317299354,
+      "learning_rate": 7.072493615689464e-06,
+      "loss": 0.6743,
+      "step": 7545
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 20.80844236847429,
+      "learning_rate": 7.067991931928653e-06,
+      "loss": 0.6773,
+      "step": 7550
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.088359347091162,
+      "learning_rate": 7.063488224995555e-06,
+      "loss": 0.6706,
+      "step": 7555
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 14.031454429113667,
+      "learning_rate": 7.058982499296271e-06,
+      "loss": 0.6812,
+      "step": 7560
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.189777793794025,
+      "learning_rate": 7.0544747592388705e-06,
+      "loss": 0.677,
+      "step": 7565
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 14.163766472519798,
+      "learning_rate": 7.049965009233399e-06,
+      "loss": 0.6629,
+      "step": 7570
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.698400255214532,
+      "learning_rate": 7.045453253691868e-06,
+      "loss": 0.6697,
+      "step": 7575
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.874532923503898,
+      "learning_rate": 7.040939497028247e-06,
+      "loss": 0.6755,
+      "step": 7580
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.421821705632198,
+      "learning_rate": 7.0364237436584685e-06,
+      "loss": 0.6592,
+      "step": 7585
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.231467938299327,
+      "learning_rate": 7.031905998000414e-06,
+      "loss": 0.662,
+      "step": 7590
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 36.750873525169936,
+      "learning_rate": 7.027386264473914e-06,
+      "loss": 0.6562,
+      "step": 7595
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.722367709743043,
+      "learning_rate": 7.022864547500751e-06,
+      "loss": 0.6704,
+      "step": 7600
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.028304612662503,
+      "learning_rate": 7.018340851504637e-06,
+      "loss": 0.6637,
+      "step": 7605
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.942067762843156,
+      "learning_rate": 7.01381518091123e-06,
+      "loss": 0.6681,
+      "step": 7610
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.57676824119045,
+      "learning_rate": 7.009287540148113e-06,
+      "loss": 0.6918,
+      "step": 7615
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.920375170456783,
+      "learning_rate": 7.004757933644801e-06,
+      "loss": 0.6575,
+      "step": 7620
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.322859449563097,
+      "learning_rate": 7.000226365832729e-06,
+      "loss": 0.682,
+      "step": 7625
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.312229271567947,
+      "learning_rate": 6.995692841145253e-06,
+      "loss": 0.6701,
+      "step": 7630
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.707392654136837,
+      "learning_rate": 6.991157364017642e-06,
+      "loss": 0.6654,
+      "step": 7635
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.796894963851315,
+      "learning_rate": 6.986619938887076e-06,
+      "loss": 0.6578,
+      "step": 7640
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.853471127640145,
+      "learning_rate": 6.982080570192638e-06,
+      "loss": 0.6677,
+      "step": 7645
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.12977247638146,
+      "learning_rate": 6.977539262375318e-06,
+      "loss": 0.658,
+      "step": 7650
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.699221620073573,
+      "learning_rate": 6.972996019877998e-06,
+      "loss": 0.659,
+      "step": 7655
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.424426213591365,
+      "learning_rate": 6.968450847145456e-06,
+      "loss": 0.6668,
+      "step": 7660
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 18.343320233271434,
+      "learning_rate": 6.963903748624356e-06,
+      "loss": 0.6363,
+      "step": 7665
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 27.69523332992964,
+      "learning_rate": 6.959354728763247e-06,
+      "loss": 0.6543,
+      "step": 7670
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.276146037274146,
+      "learning_rate": 6.954803792012559e-06,
+      "loss": 0.6855,
+      "step": 7675
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.700634905867604,
+      "learning_rate": 6.950250942824595e-06,
+      "loss": 0.6564,
+      "step": 7680
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.506750873615168,
+      "learning_rate": 6.945696185653532e-06,
+      "loss": 0.6768,
+      "step": 7685
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.791087185631472,
+      "learning_rate": 6.941139524955409e-06,
+      "loss": 0.6658,
+      "step": 7690
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.266362817079182,
+      "learning_rate": 6.936580965188133e-06,
+      "loss": 0.6703,
+      "step": 7695
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.05943328414273,
+      "learning_rate": 6.9320205108114634e-06,
+      "loss": 0.634,
+      "step": 7700
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 34.555429787519394,
+      "learning_rate": 6.927458166287017e-06,
+      "loss": 0.6676,
+      "step": 7705
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.106677063243689,
+      "learning_rate": 6.9228939360782585e-06,
+      "loss": 0.6558,
+      "step": 7710
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.555116707313184,
+      "learning_rate": 6.918327824650497e-06,
+      "loss": 0.6678,
+      "step": 7715
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 34.94775840826066,
+      "learning_rate": 6.913759836470884e-06,
+      "loss": 0.6676,
+      "step": 7720
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 23.503473494897815,
+      "learning_rate": 6.9091899760084025e-06,
+      "loss": 0.6712,
+      "step": 7725
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.198449503546897,
+      "learning_rate": 6.904618247733874e-06,
+      "loss": 0.6821,
+      "step": 7730
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.423391165147742,
+      "learning_rate": 6.90004465611994e-06,
+      "loss": 0.665,
+      "step": 7735
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 27.492408282325805,
+      "learning_rate": 6.895469205641071e-06,
+      "loss": 0.6656,
+      "step": 7740
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 27.131001091554225,
+      "learning_rate": 6.890891900773552e-06,
+      "loss": 0.6466,
+      "step": 7745
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.343118919497925,
+      "learning_rate": 6.886312745995485e-06,
+      "loss": 0.6613,
+      "step": 7750
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 26.29602942837788,
+      "learning_rate": 6.881731745786779e-06,
+      "loss": 0.6744,
+      "step": 7755
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 34.93533020196708,
+      "learning_rate": 6.877148904629154e-06,
+      "loss": 0.6661,
+      "step": 7760
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 20.015281493550773,
+      "learning_rate": 6.872564227006122e-06,
+      "loss": 0.6672,
+      "step": 7765
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.05475189958386,
+      "learning_rate": 6.867977717403e-06,
+      "loss": 0.6693,
+      "step": 7770
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.722271714945407,
+      "learning_rate": 6.863389380306894e-06,
+      "loss": 0.6785,
+      "step": 7775
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.594100948343948,
+      "learning_rate": 6.858799220206698e-06,
+      "loss": 0.6808,
+      "step": 7780
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.256070128775296,
+      "learning_rate": 6.854207241593086e-06,
+      "loss": 0.6684,
+      "step": 7785
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.284928024061565,
+      "learning_rate": 6.849613448958518e-06,
+      "loss": 0.6584,
+      "step": 7790
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.343801575420452,
+      "learning_rate": 6.845017846797224e-06,
+      "loss": 0.6453,
+      "step": 7795
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.988882531705171,
+      "learning_rate": 6.840420439605207e-06,
+      "loss": 0.6343,
+      "step": 7800
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 20.704821166327548,
+      "learning_rate": 6.835821231880233e-06,
+      "loss": 0.6438,
+      "step": 7805
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 23.222455538039853,
+      "learning_rate": 6.831220228121831e-06,
+      "loss": 0.6546,
+      "step": 7810
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 21.479424891039326,
+      "learning_rate": 6.826617432831286e-06,
+      "loss": 0.6563,
+      "step": 7815
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 42.80175925549784,
+      "learning_rate": 6.822012850511641e-06,
+      "loss": 0.6724,
+      "step": 7820
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 27.776123859447747,
+      "learning_rate": 6.81740648566768e-06,
+      "loss": 0.6648,
+      "step": 7825
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.208497254204323,
+      "learning_rate": 6.812798342805933e-06,
+      "loss": 0.6484,
+      "step": 7830
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 31.762321890202262,
+      "learning_rate": 6.808188426434672e-06,
+      "loss": 0.6459,
+      "step": 7835
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.06389345005782,
+      "learning_rate": 6.803576741063903e-06,
+      "loss": 0.6704,
+      "step": 7840
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.02164928639619,
+      "learning_rate": 6.79896329120536e-06,
+      "loss": 0.6617,
+      "step": 7845
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.709489451722483,
+      "learning_rate": 6.794348081372507e-06,
+      "loss": 0.6528,
+      "step": 7850
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.238437551559613,
+      "learning_rate": 6.789731116080529e-06,
+      "loss": 0.6407,
+      "step": 7855
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.206729482619727,
+      "learning_rate": 6.785112399846328e-06,
+      "loss": 0.6625,
+      "step": 7860
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.776138225675623,
+      "learning_rate": 6.780491937188514e-06,
+      "loss": 0.6543,
+      "step": 7865
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.706557883721104,
+      "learning_rate": 6.775869732627417e-06,
+      "loss": 0.6576,
+      "step": 7870
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.764273172448627,
+      "learning_rate": 6.771245790685059e-06,
+      "loss": 0.6767,
+      "step": 7875
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 52.38525602824595,
+      "learning_rate": 6.766620115885172e-06,
+      "loss": 0.6719,
+      "step": 7880
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.77477399033817,
+      "learning_rate": 6.761992712753173e-06,
+      "loss": 0.6544,
+      "step": 7885
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.49369698024647,
+      "learning_rate": 6.757363585816178e-06,
+      "loss": 0.6428,
+      "step": 7890
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.283967365331387,
+      "learning_rate": 6.7527327396029875e-06,
+      "loss": 0.6457,
+      "step": 7895
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.407807058273839,
+      "learning_rate": 6.748100178644082e-06,
+      "loss": 0.6507,
+      "step": 7900
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.881801966972798,
+      "learning_rate": 6.743465907471623e-06,
+      "loss": 0.663,
+      "step": 7905
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.894217087735242,
+      "learning_rate": 6.738829930619438e-06,
+      "loss": 0.6609,
+      "step": 7910
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.343860422804275,
+      "learning_rate": 6.734192252623034e-06,
+      "loss": 0.6484,
+      "step": 7915
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.817117023965405,
+      "learning_rate": 6.729552878019574e-06,
+      "loss": 0.6381,
+      "step": 7920
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.166428445424156,
+      "learning_rate": 6.724911811347883e-06,
+      "loss": 0.6493,
+      "step": 7925
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.180628433066271,
+      "learning_rate": 6.7202690571484406e-06,
+      "loss": 0.6647,
+      "step": 7930
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.866832953011489,
+      "learning_rate": 6.71562461996338e-06,
+      "loss": 0.6332,
+      "step": 7935
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 21.691113499249862,
+      "learning_rate": 6.71097850433648e-06,
+      "loss": 0.6726,
+      "step": 7940
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.55997210610206,
+      "learning_rate": 6.706330714813161e-06,
+      "loss": 0.6354,
+      "step": 7945
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 19.705598228965442,
+      "learning_rate": 6.701681255940478e-06,
+      "loss": 0.6406,
+      "step": 7950
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.295922874790215,
+      "learning_rate": 6.697030132267124e-06,
+      "loss": 0.6578,
+      "step": 7955
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 16.841094205265872,
+      "learning_rate": 6.692377348343419e-06,
+      "loss": 0.6552,
+      "step": 7960
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 22.95613214066316,
+      "learning_rate": 6.687722908721308e-06,
+      "loss": 0.6734,
+      "step": 7965
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 27.35634967120633,
+      "learning_rate": 6.683066817954353e-06,
+      "loss": 0.6497,
+      "step": 7970
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.417662535768118,
+      "learning_rate": 6.678409080597732e-06,
+      "loss": 0.6628,
+      "step": 7975
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.305778063010779,
+      "learning_rate": 6.673749701208239e-06,
+      "loss": 0.6278,
+      "step": 7980
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.067243154378666,
+      "learning_rate": 6.669088684344266e-06,
+      "loss": 0.6304,
+      "step": 7985
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.237713121087854,
+      "learning_rate": 6.664426034565814e-06,
+      "loss": 0.6533,
+      "step": 7990
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.6883233127354975,
+      "learning_rate": 6.6597617564344796e-06,
+      "loss": 0.6452,
+      "step": 7995
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.400448497043079,
+      "learning_rate": 6.6550958545134515e-06,
+      "loss": 0.6489,
+      "step": 8000
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.767147219100383,
+      "learning_rate": 6.6504283333675065e-06,
+      "loss": 0.6449,
+      "step": 8005
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.370384698947925,
+      "learning_rate": 6.645759197563008e-06,
+      "loss": 0.6429,
+      "step": 8010
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.061534266870285,
+      "learning_rate": 6.641088451667894e-06,
+      "loss": 0.662,
+      "step": 8015
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 21.18127236595691,
+      "learning_rate": 6.636416100251687e-06,
+      "loss": 0.6504,
+      "step": 8020
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.535465862975343,
+      "learning_rate": 6.631742147885468e-06,
+      "loss": 0.6772,
+      "step": 8025
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.964640726111356,
+      "learning_rate": 6.627066599141895e-06,
+      "loss": 0.6597,
+      "step": 8030
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.845507943939163,
+      "learning_rate": 6.622389458595182e-06,
+      "loss": 0.6469,
+      "step": 8035
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 16.080552123187886,
+      "learning_rate": 6.617710730821103e-06,
+      "loss": 0.6429,
+      "step": 8040
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 21.511449344925648,
+      "learning_rate": 6.6130304203969805e-06,
+      "loss": 0.6482,
+      "step": 8045
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.772854516081404,
+      "learning_rate": 6.608348531901692e-06,
+      "loss": 0.6619,
+      "step": 8050
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.625015490224003,
+      "learning_rate": 6.603665069915654e-06,
+      "loss": 0.6568,
+      "step": 8055
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 25.861303879556942,
+      "learning_rate": 6.5989800390208226e-06,
+      "loss": 0.6386,
+      "step": 8060
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.4888690477709,
+      "learning_rate": 6.59429344380069e-06,
+      "loss": 0.6436,
+      "step": 8065
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.131438242239497,
+      "learning_rate": 6.5896052888402805e-06,
+      "loss": 0.6476,
+      "step": 8070
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.257155345532185,
+      "learning_rate": 6.584915578726141e-06,
+      "loss": 0.6393,
+      "step": 8075
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.259442077739987,
+      "learning_rate": 6.5802243180463425e-06,
+      "loss": 0.6219,
+      "step": 8080
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.655363886815646,
+      "learning_rate": 6.575531511390469e-06,
+      "loss": 0.6438,
+      "step": 8085
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.89276103817776,
+      "learning_rate": 6.570837163349624e-06,
+      "loss": 0.6257,
+      "step": 8090
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.058137605602374,
+      "learning_rate": 6.566141278516413e-06,
+      "loss": 0.6516,
+      "step": 8095
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 48.71355207579282,
+      "learning_rate": 6.561443861484946e-06,
+      "loss": 0.6402,
+      "step": 8100
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 28.380628719561553,
+      "learning_rate": 6.5567449168508346e-06,
+      "loss": 0.654,
+      "step": 8105
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.411789938460549,
+      "learning_rate": 6.552044449211181e-06,
+      "loss": 0.6552,
+      "step": 8110
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 22.60346491316076,
+      "learning_rate": 6.54734246316458e-06,
+      "loss": 0.6531,
+      "step": 8115
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.151539620095777,
+      "learning_rate": 6.542638963311112e-06,
+      "loss": 0.649,
+      "step": 8120
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 29.874279752302918,
+      "learning_rate": 6.537933954252338e-06,
+      "loss": 0.6307,
+      "step": 8125
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 55.371100766921124,
+      "learning_rate": 6.533227440591294e-06,
+      "loss": 0.6719,
+      "step": 8130
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.23924904787888,
+      "learning_rate": 6.5285194269324895e-06,
+      "loss": 0.6554,
+      "step": 8135
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.542786323664106,
+      "learning_rate": 6.523809917881902e-06,
+      "loss": 0.6363,
+      "step": 8140
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.138369074140726,
+      "learning_rate": 6.5190989180469736e-06,
+      "loss": 0.6489,
+      "step": 8145
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 20.644221672684967,
+      "learning_rate": 6.514386432036598e-06,
+      "loss": 0.6514,
+      "step": 8150
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.53784502326818,
+      "learning_rate": 6.5096724644611296e-06,
+      "loss": 0.6505,
+      "step": 8155
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.832374181994506,
+      "learning_rate": 6.5049570199323686e-06,
+      "loss": 0.6431,
+      "step": 8160
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.328834025436,
+      "learning_rate": 6.500240103063564e-06,
+      "loss": 0.6243,
+      "step": 8165
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.916018663546673,
+      "learning_rate": 6.495521718469402e-06,
+      "loss": 0.6457,
+      "step": 8170
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 23.457435094958125,
+      "learning_rate": 6.490801870766004e-06,
+      "loss": 0.6568,
+      "step": 8175
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 35.150856775834455,
+      "learning_rate": 6.486080564570925e-06,
+      "loss": 0.6513,
+      "step": 8180
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.170893686186264,
+      "learning_rate": 6.481357804503147e-06,
+      "loss": 0.6362,
+      "step": 8185
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.889101730266102,
+      "learning_rate": 6.476633595183073e-06,
+      "loss": 0.6487,
+      "step": 8190
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.13019074213934,
+      "learning_rate": 6.471907941232525e-06,
+      "loss": 0.6449,
+      "step": 8195
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.939034783675478,
+      "learning_rate": 6.467180847274737e-06,
+      "loss": 0.6366,
+      "step": 8200
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.291379823253529,
+      "learning_rate": 6.462452317934352e-06,
+      "loss": 0.6348,
+      "step": 8205
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.535236612949914,
+      "learning_rate": 6.45772235783742e-06,
+      "loss": 0.6284,
+      "step": 8210
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.745474723331273,
+      "learning_rate": 6.452990971611384e-06,
+      "loss": 0.6245,
+      "step": 8215
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 29.761555067812456,
+      "learning_rate": 6.448258163885092e-06,
+      "loss": 0.6626,
+      "step": 8220
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.001809069262323,
+      "learning_rate": 6.443523939288776e-06,
+      "loss": 0.6446,
+      "step": 8225
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.06191759505353,
+      "learning_rate": 6.438788302454053e-06,
+      "loss": 0.6637,
+      "step": 8230
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.716149188351201,
+      "learning_rate": 6.434051258013928e-06,
+      "loss": 0.6606,
+      "step": 8235
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.812890077780472,
+      "learning_rate": 6.429312810602777e-06,
+      "loss": 0.6329,
+      "step": 8240
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.844538478736041,
+      "learning_rate": 6.424572964856351e-06,
+      "loss": 0.6463,
+      "step": 8245
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.81354924758096,
+      "learning_rate": 6.4198317254117695e-06,
+      "loss": 0.6431,
+      "step": 8250
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.492158449379378,
+      "learning_rate": 6.4150890969075145e-06,
+      "loss": 0.6432,
+      "step": 8255
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.759512578968948,
+      "learning_rate": 6.410345083983427e-06,
+      "loss": 0.6303,
+      "step": 8260
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 13.119349300664416,
+      "learning_rate": 6.405599691280706e-06,
+      "loss": 0.6543,
+      "step": 8265
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 33.850079790425276,
+      "learning_rate": 6.400852923441892e-06,
+      "loss": 0.6455,
+      "step": 8270
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 27.973834250364234,
+      "learning_rate": 6.3961047851108795e-06,
+      "loss": 0.6439,
+      "step": 8275
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.791402441709256,
+      "learning_rate": 6.391355280932898e-06,
+      "loss": 0.6526,
+      "step": 8280
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.8898792407939835,
+      "learning_rate": 6.386604415554521e-06,
+      "loss": 0.6257,
+      "step": 8285
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.302907884289335,
+      "learning_rate": 6.381852193623641e-06,
+      "loss": 0.6526,
+      "step": 8290
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 24.989620308397974,
+      "learning_rate": 6.377098619789489e-06,
+      "loss": 0.6402,
+      "step": 8295
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 27.03639392548826,
+      "learning_rate": 6.372343698702612e-06,
+      "loss": 0.6241,
+      "step": 8300
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 26.418474973434172,
+      "learning_rate": 6.367587435014882e-06,
+      "loss": 0.6298,
+      "step": 8305
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.467358507902656,
+      "learning_rate": 6.362829833379475e-06,
+      "loss": 0.6539,
+      "step": 8310
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.390598419534992,
+      "learning_rate": 6.358070898450884e-06,
+      "loss": 0.6575,
+      "step": 8315
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 22.18745112063696,
+      "learning_rate": 6.353310634884901e-06,
+      "loss": 0.6289,
+      "step": 8320
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 18.267245620225978,
+      "learning_rate": 6.348549047338626e-06,
+      "loss": 0.637,
+      "step": 8325
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.441981411328962,
+      "learning_rate": 6.343786140470441e-06,
+      "loss": 0.6309,
+      "step": 8330
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.388295631417312,
+      "learning_rate": 6.339021918940031e-06,
+      "loss": 0.6482,
+      "step": 8335
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 20.34027340053568,
+      "learning_rate": 6.33425638740836e-06,
+      "loss": 0.6316,
+      "step": 8340
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 18.72616964408831,
+      "learning_rate": 6.3294895505376784e-06,
+      "loss": 0.6445,
+      "step": 8345
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.599848289761585,
+      "learning_rate": 6.324721412991509e-06,
+      "loss": 0.6373,
+      "step": 8350
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.225968332536441,
+      "learning_rate": 6.31995197943465e-06,
+      "loss": 0.6426,
+      "step": 8355
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.216949694680935,
+      "learning_rate": 6.315181254533167e-06,
+      "loss": 0.631,
+      "step": 8360
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.470168620240177,
+      "learning_rate": 6.310409242954392e-06,
+      "loss": 0.6312,
+      "step": 8365
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.414773113372265,
+      "learning_rate": 6.305635949366906e-06,
+      "loss": 0.6385,
+      "step": 8370
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 34.07839646444887,
+      "learning_rate": 6.300861378440557e-06,
+      "loss": 0.6402,
+      "step": 8375
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.71560255107337,
+      "learning_rate": 6.296085534846433e-06,
+      "loss": 0.6279,
+      "step": 8380
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 37.75055046604799,
+      "learning_rate": 6.291308423256873e-06,
+      "loss": 0.622,
+      "step": 8385
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 31.051698238500787,
+      "learning_rate": 6.286530048345452e-06,
+      "loss": 0.6399,
+      "step": 8390
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.138592809970431,
+      "learning_rate": 6.281750414786983e-06,
+      "loss": 0.6423,
+      "step": 8395
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 29.79002494254909,
+      "learning_rate": 6.276969527257513e-06,
+      "loss": 0.6361,
+      "step": 8400
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 28.102533538762536,
+      "learning_rate": 6.272187390434311e-06,
+      "loss": 0.626,
+      "step": 8405
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 31.97698553168282,
+      "learning_rate": 6.26740400899587e-06,
+      "loss": 0.6422,
+      "step": 8410
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.465448441293175,
+      "learning_rate": 6.262619387621902e-06,
+      "loss": 0.6367,
+      "step": 8415
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.061458649998584,
+      "learning_rate": 6.257833530993332e-06,
+      "loss": 0.6335,
+      "step": 8420
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.88280726412742,
+      "learning_rate": 6.2530464437922936e-06,
+      "loss": 0.6312,
+      "step": 8425
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.12066800542379,
+      "learning_rate": 6.2482581307021195e-06,
+      "loss": 0.6124,
+      "step": 8430
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 16.94764296608338,
+      "learning_rate": 6.243468596407348e-06,
+      "loss": 0.615,
+      "step": 8435
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.307650693950313,
+      "learning_rate": 6.238677845593709e-06,
+      "loss": 0.629,
+      "step": 8440
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.164608367344734,
+      "learning_rate": 6.233885882948124e-06,
+      "loss": 0.6238,
+      "step": 8445
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.618845576266027,
+      "learning_rate": 6.229092713158699e-06,
+      "loss": 0.6307,
+      "step": 8450
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.999939335710288,
+      "learning_rate": 6.22429834091472e-06,
+      "loss": 0.6252,
+      "step": 8455
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 27.277384481971897,
+      "learning_rate": 6.219502770906652e-06,
+      "loss": 0.6255,
+      "step": 8460
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 26.658657226008486,
+      "learning_rate": 6.214706007826133e-06,
+      "loss": 0.6336,
+      "step": 8465
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.790938205285673,
+      "learning_rate": 6.20990805636596e-06,
+      "loss": 0.6216,
+      "step": 8470
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.181913156116149,
+      "learning_rate": 6.205108921220102e-06,
+      "loss": 0.6092,
+      "step": 8475
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.651943934098993,
+      "learning_rate": 6.200308607083683e-06,
+      "loss": 0.6349,
+      "step": 8480
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.036649904845357,
+      "learning_rate": 6.195507118652977e-06,
+      "loss": 0.6251,
+      "step": 8485
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.146870640781684,
+      "learning_rate": 6.190704460625412e-06,
+      "loss": 0.6372,
+      "step": 8490
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.004606776153171,
+      "learning_rate": 6.185900637699555e-06,
+      "loss": 0.6379,
+      "step": 8495
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.261848092144794,
+      "learning_rate": 6.18109565457512e-06,
+      "loss": 0.6273,
+      "step": 8500
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.517734637753321,
+      "learning_rate": 6.176289515952949e-06,
+      "loss": 0.6235,
+      "step": 8505
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.182064084964512,
+      "learning_rate": 6.171482226535016e-06,
+      "loss": 0.6338,
+      "step": 8510
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 27.011998601299425,
+      "learning_rate": 6.1666737910244234e-06,
+      "loss": 0.6279,
+      "step": 8515
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.461115547233499,
+      "learning_rate": 6.161864214125393e-06,
+      "loss": 0.6187,
+      "step": 8520
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 26.957756432468457,
+      "learning_rate": 6.157053500543265e-06,
+      "loss": 0.6401,
+      "step": 8525
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 14.81850691520542,
+      "learning_rate": 6.152241654984488e-06,
+      "loss": 0.6462,
+      "step": 8530
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 19.24439589408677,
+      "learning_rate": 6.147428682156621e-06,
+      "loss": 0.6354,
+      "step": 8535
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 36.13098754949811,
+      "learning_rate": 6.142614586768325e-06,
+      "loss": 0.6291,
+      "step": 8540
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.549996730002588,
+      "learning_rate": 6.137799373529361e-06,
+      "loss": 0.6183,
+      "step": 8545
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 19.749594900839334,
+      "learning_rate": 6.132983047150579e-06,
+      "loss": 0.6197,
+      "step": 8550
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 17.905561914987118,
+      "learning_rate": 6.128165612343923e-06,
+      "loss": 0.6175,
+      "step": 8555
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.52868932543812,
+      "learning_rate": 6.1233470738224185e-06,
+      "loss": 0.6218,
+      "step": 8560
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.267449464182201,
+      "learning_rate": 6.118527436300175e-06,
+      "loss": 0.6316,
+      "step": 8565
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.904992821265832,
+      "learning_rate": 6.11370670449237e-06,
+      "loss": 0.639,
+      "step": 8570
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.478604606870359,
+      "learning_rate": 6.108884883115256e-06,
+      "loss": 0.6413,
+      "step": 8575
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 19.41975081931992,
+      "learning_rate": 6.1040619768861505e-06,
+      "loss": 0.6376,
+      "step": 8580
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 28.473262465248553,
+      "learning_rate": 6.099237990523437e-06,
+      "loss": 0.619,
+      "step": 8585
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.790384150641972,
+      "learning_rate": 6.094412928746546e-06,
+      "loss": 0.6234,
+      "step": 8590
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 19.06048377191451,
+      "learning_rate": 6.089586796275968e-06,
+      "loss": 0.6226,
+      "step": 8595
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 23.277298880330747,
+      "learning_rate": 6.084759597833239e-06,
+      "loss": 0.6264,
+      "step": 8600
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.554662803269148,
+      "learning_rate": 6.079931338140936e-06,
+      "loss": 0.6305,
+      "step": 8605
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 19.270349394789058,
+      "learning_rate": 6.075102021922677e-06,
+      "loss": 0.6439,
+      "step": 8610
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 67.3469154291316,
+      "learning_rate": 6.070271653903112e-06,
+      "loss": 0.6062,
+      "step": 8615
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 27.770308384712457,
+      "learning_rate": 6.06544023880792e-06,
+      "loss": 0.6329,
+      "step": 8620
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 25.407479290475464,
+      "learning_rate": 6.060607781363807e-06,
+      "loss": 0.6315,
+      "step": 8625
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.509769672420044,
+      "learning_rate": 6.055774286298492e-06,
+      "loss": 0.6365,
+      "step": 8630
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 27.6938281196245,
+      "learning_rate": 6.050939758340716e-06,
+      "loss": 0.6473,
+      "step": 8635
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.082490505267428,
+      "learning_rate": 6.046104202220228e-06,
+      "loss": 0.614,
+      "step": 8640
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.77340992162146,
+      "learning_rate": 6.041267622667784e-06,
+      "loss": 0.6326,
+      "step": 8645
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.152095095597254,
+      "learning_rate": 6.0364300244151385e-06,
+      "loss": 0.6233,
+      "step": 8650
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 23.517070545128238,
+      "learning_rate": 6.031591412195046e-06,
+      "loss": 0.621,
+      "step": 8655
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.288564339786074,
+      "learning_rate": 6.02675179074125e-06,
+      "loss": 0.6504,
+      "step": 8660
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.512244767455876,
+      "learning_rate": 6.021911164788483e-06,
+      "loss": 0.6419,
+      "step": 8665
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.139035823232772,
+      "learning_rate": 6.0170695390724595e-06,
+      "loss": 0.6154,
+      "step": 8670
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 30.611473228247544,
+      "learning_rate": 6.012226918329874e-06,
+      "loss": 0.6163,
+      "step": 8675
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 23.228208297353323,
+      "learning_rate": 6.007383307298391e-06,
+      "loss": 0.6235,
+      "step": 8680
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.16179449932119,
+      "learning_rate": 6.002538710716649e-06,
+      "loss": 0.6376,
+      "step": 8685
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.613436166968452,
+      "learning_rate": 5.997693133324244e-06,
+      "loss": 0.6184,
+      "step": 8690
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.000653168372537,
+      "learning_rate": 5.992846579861737e-06,
+      "loss": 0.6334,
+      "step": 8695
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 21.01047179127842,
+      "learning_rate": 5.98799905507064e-06,
+      "loss": 0.6309,
+      "step": 8700
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.397516068200028,
+      "learning_rate": 5.98315056369342e-06,
+      "loss": 0.6194,
+      "step": 8705
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.549455568952771,
+      "learning_rate": 5.978301110473486e-06,
+      "loss": 0.6266,
+      "step": 8710
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.407050992301864,
+      "learning_rate": 5.973450700155187e-06,
+      "loss": 0.6169,
+      "step": 8715
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.507586579212115,
+      "learning_rate": 5.968599337483814e-06,
+      "loss": 0.6138,
+      "step": 8720
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.604919897561592,
+      "learning_rate": 5.963747027205583e-06,
+      "loss": 0.6308,
+      "step": 8725
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.18989665536131,
+      "learning_rate": 5.95889377406764e-06,
+      "loss": 0.6248,
+      "step": 8730
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.554625835371105,
+      "learning_rate": 5.954039582818053e-06,
+      "loss": 0.6174,
+      "step": 8735
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 20.92839999126903,
+      "learning_rate": 5.949184458205811e-06,
+      "loss": 0.6188,
+      "step": 8740
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.267413663263957,
+      "learning_rate": 5.944328404980813e-06,
+      "loss": 0.6414,
+      "step": 8745
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 23.29476861797064,
+      "learning_rate": 5.939471427893862e-06,
+      "loss": 0.6219,
+      "step": 8750
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 22.204976774549824,
+      "learning_rate": 5.934613531696677e-06,
+      "loss": 0.6137,
+      "step": 8755
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 38.20480863930992,
+      "learning_rate": 5.929754721141863e-06,
+      "loss": 0.6239,
+      "step": 8760
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.401029185094114,
+      "learning_rate": 5.924895000982929e-06,
+      "loss": 0.6106,
+      "step": 8765
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 27.96896375943163,
+      "learning_rate": 5.920034375974267e-06,
+      "loss": 0.625,
+      "step": 8770
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.712006261527762,
+      "learning_rate": 5.91517285087116e-06,
+      "loss": 0.6301,
+      "step": 8775
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.332878892952024,
+      "learning_rate": 5.910310430429768e-06,
+      "loss": 0.6277,
+      "step": 8780
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.420058266879154,
+      "learning_rate": 5.905447119407132e-06,
+      "loss": 0.6302,
+      "step": 8785
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 20.417067516686085,
+      "learning_rate": 5.900582922561155e-06,
+      "loss": 0.6167,
+      "step": 8790
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.297247346641591,
+      "learning_rate": 5.895717844650613e-06,
+      "loss": 0.6135,
+      "step": 8795
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 31.71322909627282,
+      "learning_rate": 5.890851890435144e-06,
+      "loss": 0.6064,
+      "step": 8800
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.536266619232359,
+      "learning_rate": 5.885985064675243e-06,
+      "loss": 0.609,
+      "step": 8805
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 40.11580915341827,
+      "learning_rate": 5.881117372132257e-06,
+      "loss": 0.6287,
+      "step": 8810
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.751306111236599,
+      "learning_rate": 5.876248817568379e-06,
+      "loss": 0.6161,
+      "step": 8815
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.192346826103622,
+      "learning_rate": 5.871379405746647e-06,
+      "loss": 0.629,
+      "step": 8820
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 18.071210970161694,
+      "learning_rate": 5.8665091414309395e-06,
+      "loss": 0.6142,
+      "step": 8825
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.558916531059064,
+      "learning_rate": 5.861638029385969e-06,
+      "loss": 0.6144,
+      "step": 8830
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 38.805705537435244,
+      "learning_rate": 5.856766074377273e-06,
+      "loss": 0.5987,
+      "step": 8835
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 29.412936674979335,
+      "learning_rate": 5.851893281171217e-06,
+      "loss": 0.6259,
+      "step": 8840
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.380415844971367,
+      "learning_rate": 5.8470196545349865e-06,
+      "loss": 0.6221,
+      "step": 8845
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 21.738500373870316,
+      "learning_rate": 5.842145199236583e-06,
+      "loss": 0.6163,
+      "step": 8850
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 44.689830654455925,
+      "learning_rate": 5.837269920044815e-06,
+      "loss": 0.6088,
+      "step": 8855
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 22.913294256785697,
+      "learning_rate": 5.832393821729301e-06,
+      "loss": 0.6165,
+      "step": 8860
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 22.51389955214667,
+      "learning_rate": 5.827516909060459e-06,
+      "loss": 0.5958,
+      "step": 8865
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 21.03446683194453,
+      "learning_rate": 5.8226391868095064e-06,
+      "loss": 0.6123,
+      "step": 8870
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.297799547031005,
+      "learning_rate": 5.817760659748448e-06,
+      "loss": 0.6207,
+      "step": 8875
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 18.64278180009031,
+      "learning_rate": 5.812881332650079e-06,
+      "loss": 0.6187,
+      "step": 8880
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 18.235311973907905,
+      "learning_rate": 5.808001210287978e-06,
+      "loss": 0.6083,
+      "step": 8885
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.63252366698247,
+      "learning_rate": 5.803120297436498e-06,
+      "loss": 0.6127,
+      "step": 8890
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 21.192227497855644,
+      "learning_rate": 5.7982385988707705e-06,
+      "loss": 0.6142,
+      "step": 8895
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 19.526216402550123,
+      "learning_rate": 5.793356119366689e-06,
+      "loss": 0.6367,
+      "step": 8900
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.321169117285131,
+      "learning_rate": 5.788472863700918e-06,
+      "loss": 0.6125,
+      "step": 8905
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.52084507199846,
+      "learning_rate": 5.7835888366508785e-06,
+      "loss": 0.6203,
+      "step": 8910
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.1271021467657425,
+      "learning_rate": 5.778704042994744e-06,
+      "loss": 0.6008,
+      "step": 8915
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.254171708462301,
+      "learning_rate": 5.77381848751144e-06,
+      "loss": 0.6138,
+      "step": 8920
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.320024335417965,
+      "learning_rate": 5.76893217498064e-06,
+      "loss": 0.618,
+      "step": 8925
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.131875846221742,
+      "learning_rate": 5.764045110182752e-06,
+      "loss": 0.6103,
+      "step": 8930
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.229969236321889,
+      "learning_rate": 5.759157297898924e-06,
+      "loss": 0.6286,
+      "step": 8935
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.203086200392013,
+      "learning_rate": 5.754268742911037e-06,
+      "loss": 0.6078,
+      "step": 8940
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.337291102785596,
+      "learning_rate": 5.749379450001693e-06,
+      "loss": 0.6291,
+      "step": 8945
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.130884826076313,
+      "learning_rate": 5.74448942395422e-06,
+      "loss": 0.615,
+      "step": 8950
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.161302708063478,
+      "learning_rate": 5.739598669552664e-06,
+      "loss": 0.6025,
+      "step": 8955
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.614457781504585,
+      "learning_rate": 5.73470719158178e-06,
+      "loss": 0.5931,
+      "step": 8960
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.727423296416706,
+      "learning_rate": 5.729814994827034e-06,
+      "loss": 0.6256,
+      "step": 8965
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.883727699858422,
+      "learning_rate": 5.724922084074595e-06,
+      "loss": 0.6195,
+      "step": 8970
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.987766905601958,
+      "learning_rate": 5.720028464111326e-06,
+      "loss": 0.6148,
+      "step": 8975
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.839778190277152,
+      "learning_rate": 5.715134139724792e-06,
+      "loss": 0.5994,
+      "step": 8980
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.502188480226746,
+      "learning_rate": 5.710239115703238e-06,
+      "loss": 0.5999,
+      "step": 8985
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 16.162420148029522,
+      "learning_rate": 5.705343396835602e-06,
+      "loss": 0.6065,
+      "step": 8990
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 16.526083835319188,
+      "learning_rate": 5.7004469879114955e-06,
+      "loss": 0.6127,
+      "step": 8995
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.161343365344532,
+      "learning_rate": 5.6955498937212074e-06,
+      "loss": 0.6183,
+      "step": 9000
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.014971463269138,
+      "learning_rate": 5.6906521190556976e-06,
+      "loss": 0.6128,
+      "step": 9005
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.317961414655818,
+      "learning_rate": 5.685753668706591e-06,
+      "loss": 0.6106,
+      "step": 9010
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.993800769863125,
+      "learning_rate": 5.680854547466174e-06,
+      "loss": 0.6173,
+      "step": 9015
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 18.032312729410066,
+      "learning_rate": 5.67595476012739e-06,
+      "loss": 0.6204,
+      "step": 9020
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.913492906194817,
+      "learning_rate": 5.671054311483833e-06,
+      "loss": 0.5956,
+      "step": 9025
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.011346666019541,
+      "learning_rate": 5.66615320632974e-06,
+      "loss": 0.6032,
+      "step": 9030
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.120145904095579,
+      "learning_rate": 5.66125144946e-06,
+      "loss": 0.6298,
+      "step": 9035
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.41274470876883,
+      "learning_rate": 5.6563490456701296e-06,
+      "loss": 0.587,
+      "step": 9040
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.915343026406369,
+      "learning_rate": 5.6514459997562855e-06,
+      "loss": 0.6216,
+      "step": 9045
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.895671288900355,
+      "learning_rate": 5.64654231651525e-06,
+      "loss": 0.6183,
+      "step": 9050
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 23.881102306393533,
+      "learning_rate": 5.641638000744425e-06,
+      "loss": 0.6228,
+      "step": 9055
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 20.546332683774185,
+      "learning_rate": 5.636733057241839e-06,
+      "loss": 0.6191,
+      "step": 9060
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.134020839761392,
+      "learning_rate": 5.631827490806128e-06,
+      "loss": 0.5941,
+      "step": 9065
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.355455888132816,
+      "learning_rate": 5.626921306236541e-06,
+      "loss": 0.5877,
+      "step": 9070
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 30.435020932897146,
+      "learning_rate": 5.622014508332932e-06,
+      "loss": 0.6148,
+      "step": 9075
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.35724195784234,
+      "learning_rate": 5.617107101895751e-06,
+      "loss": 0.6164,
+      "step": 9080
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.154665214652813,
+      "learning_rate": 5.6121990917260474e-06,
+      "loss": 0.6191,
+      "step": 9085
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.512763141117652,
+      "learning_rate": 5.607290482625461e-06,
+      "loss": 0.6041,
+      "step": 9090
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.269055262135025,
+      "learning_rate": 5.602381279396213e-06,
+      "loss": 0.6069,
+      "step": 9095
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.536300771361045,
+      "learning_rate": 5.5974714868411125e-06,
+      "loss": 0.6053,
+      "step": 9100
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.77546270410342,
+      "learning_rate": 5.592561109763542e-06,
+      "loss": 0.6037,
+      "step": 9105
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.052644342202596,
+      "learning_rate": 5.587650152967454e-06,
+      "loss": 0.609,
+      "step": 9110
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.336728354774031,
+      "learning_rate": 5.582738621257372e-06,
+      "loss": 0.6046,
+      "step": 9115
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.189860501039402,
+      "learning_rate": 5.57782651943838e-06,
+      "loss": 0.6294,
+      "step": 9120
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.104760202629917,
+      "learning_rate": 5.572913852316116e-06,
+      "loss": 0.6019,
+      "step": 9125
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.372469587828709,
+      "learning_rate": 5.56800062469678e-06,
+      "loss": 0.5982,
+      "step": 9130
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 26.10833699620634,
+      "learning_rate": 5.563086841387111e-06,
+      "loss": 0.5945,
+      "step": 9135
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.931729991395199,
+      "learning_rate": 5.558172507194397e-06,
+      "loss": 0.6159,
+      "step": 9140
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.704437918283011,
+      "learning_rate": 5.5532576269264635e-06,
+      "loss": 0.6167,
+      "step": 9145
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.00692493270044,
+      "learning_rate": 5.5483422053916735e-06,
+      "loss": 0.5915,
+      "step": 9150
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.21951340821045,
+      "learning_rate": 5.543426247398912e-06,
+      "loss": 0.5967,
+      "step": 9155
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.8956292819034815,
+      "learning_rate": 5.538509757757594e-06,
+      "loss": 0.6109,
+      "step": 9160
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.693620673196478,
+      "learning_rate": 5.533592741277658e-06,
+      "loss": 0.6159,
+      "step": 9165
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.139425257113912,
+      "learning_rate": 5.528675202769549e-06,
+      "loss": 0.5993,
+      "step": 9170
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.863801913245817,
+      "learning_rate": 5.52375714704423e-06,
+      "loss": 0.6,
+      "step": 9175
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.185617291207205,
+      "learning_rate": 5.518838578913167e-06,
+      "loss": 0.6063,
+      "step": 9180
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.2860494408934295,
+      "learning_rate": 5.513919503188328e-06,
+      "loss": 0.591,
+      "step": 9185
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 34.38305181082742,
+      "learning_rate": 5.508999924682178e-06,
+      "loss": 0.6076,
+      "step": 9190
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 22.92435759386956,
+      "learning_rate": 5.504079848207671e-06,
+      "loss": 0.6167,
+      "step": 9195
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.760529664653179,
+      "learning_rate": 5.499159278578253e-06,
+      "loss": 0.6014,
+      "step": 9200
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.628714969215856,
+      "learning_rate": 5.4942382206078495e-06,
+      "loss": 0.5977,
+      "step": 9205
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.656187450738024,
+      "learning_rate": 5.489316679110864e-06,
+      "loss": 0.5956,
+      "step": 9210
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.2821838796463245,
+      "learning_rate": 5.484394658902173e-06,
+      "loss": 0.6061,
+      "step": 9215
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 19.831834598197847,
+      "learning_rate": 5.479472164797124e-06,
+      "loss": 0.5959,
+      "step": 9220
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.214689267998624,
+      "learning_rate": 5.474549201611521e-06,
+      "loss": 0.6063,
+      "step": 9225
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.16812247258966,
+      "learning_rate": 5.4696257741616366e-06,
+      "loss": 0.5894,
+      "step": 9230
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.923420658187072,
+      "learning_rate": 5.464701887264188e-06,
+      "loss": 0.6208,
+      "step": 9235
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.284791283442523,
+      "learning_rate": 5.45977754573635e-06,
+      "loss": 0.6002,
+      "step": 9240
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.352219537142107,
+      "learning_rate": 5.454852754395738e-06,
+      "loss": 0.5926,
+      "step": 9245
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.119697378052444,
+      "learning_rate": 5.449927518060407e-06,
+      "loss": 0.5869,
+      "step": 9250
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.266374499933223,
+      "learning_rate": 5.44500184154885e-06,
+      "loss": 0.5999,
+      "step": 9255
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.297673633134746,
+      "learning_rate": 5.44007572967999e-06,
+      "loss": 0.5976,
+      "step": 9260
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 21.96269904481905,
+      "learning_rate": 5.435149187273172e-06,
+      "loss": 0.6065,
+      "step": 9265
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.492653965184815,
+      "learning_rate": 5.430222219148168e-06,
+      "loss": 0.5974,
+      "step": 9270
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 23.22740569704164,
+      "learning_rate": 5.4252948301251615e-06,
+      "loss": 0.5934,
+      "step": 9275
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.173788220397405,
+      "learning_rate": 5.420367025024753e-06,
+      "loss": 0.6119,
+      "step": 9280
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.937326719215955,
+      "learning_rate": 5.415438808667944e-06,
+      "loss": 0.6035,
+      "step": 9285
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 25.668499087828007,
+      "learning_rate": 5.410510185876146e-06,
+      "loss": 0.6221,
+      "step": 9290
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.964394347664827,
+      "learning_rate": 5.405581161471157e-06,
+      "loss": 0.6056,
+      "step": 9295
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.995289947861624,
+      "learning_rate": 5.40065174027518e-06,
+      "loss": 0.5984,
+      "step": 9300
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.214810192286005,
+      "learning_rate": 5.3957219271108e-06,
+      "loss": 0.5886,
+      "step": 9305
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.989591849187818,
+      "learning_rate": 5.390791726800983e-06,
+      "loss": 0.6138,
+      "step": 9310
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.808671808767826,
+      "learning_rate": 5.385861144169081e-06,
+      "loss": 0.6074,
+      "step": 9315
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.561526838673997,
+      "learning_rate": 5.3809301840388126e-06,
+      "loss": 0.5923,
+      "step": 9320
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.999545774089387,
+      "learning_rate": 5.375998851234272e-06,
+      "loss": 0.6118,
+      "step": 9325
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 18.190251238079696,
+      "learning_rate": 5.371067150579912e-06,
+      "loss": 0.5854,
+      "step": 9330
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.054453094014715,
+      "learning_rate": 5.366135086900552e-06,
+      "loss": 0.5986,
+      "step": 9335
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 18.541785491546758,
+      "learning_rate": 5.361202665021359e-06,
+      "loss": 0.6017,
+      "step": 9340
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.566548150106643,
+      "learning_rate": 5.356269889767857e-06,
+      "loss": 0.5875,
+      "step": 9345
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.093235129596696,
+      "learning_rate": 5.351336765965913e-06,
+      "loss": 0.6059,
+      "step": 9350
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 16.523942808346206,
+      "learning_rate": 5.3464032984417345e-06,
+      "loss": 0.5845,
+      "step": 9355
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.225232390971318,
+      "learning_rate": 5.341469492021866e-06,
+      "loss": 0.5858,
+      "step": 9360
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.56203769259536,
+      "learning_rate": 5.336535351533182e-06,
+      "loss": 0.5957,
+      "step": 9365
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.809870152961832,
+      "learning_rate": 5.331600881802887e-06,
+      "loss": 0.5988,
+      "step": 9370
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.809022292311157,
+      "learning_rate": 5.326666087658505e-06,
+      "loss": 0.607,
+      "step": 9375
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.171141932866165,
+      "learning_rate": 5.321730973927879e-06,
+      "loss": 0.6137,
+      "step": 9380
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.533028033705287,
+      "learning_rate": 5.316795545439162e-06,
+      "loss": 0.5943,
+      "step": 9385
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.502217924497344,
+      "learning_rate": 5.31185980702082e-06,
+      "loss": 0.5949,
+      "step": 9390
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.901020453643195,
+      "learning_rate": 5.306923763501616e-06,
+      "loss": 0.6005,
+      "step": 9395
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 19.593915315383,
+      "learning_rate": 5.301987419710617e-06,
+      "loss": 0.6089,
+      "step": 9400
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.3117371932334,
+      "learning_rate": 5.297050780477179e-06,
+      "loss": 0.5964,
+      "step": 9405
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.826971591181108,
+      "learning_rate": 5.29211385063095e-06,
+      "loss": 0.592,
+      "step": 9410
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 31.45574299091673,
+      "learning_rate": 5.287176635001863e-06,
+      "loss": 0.6082,
+      "step": 9415
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.088086236450218,
+      "learning_rate": 5.282239138420127e-06,
+      "loss": 0.5993,
+      "step": 9420
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 42.54559890377032,
+      "learning_rate": 5.277301365716228e-06,
+      "loss": 0.6037,
+      "step": 9425
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.003789314693376,
+      "learning_rate": 5.272363321720926e-06,
+      "loss": 0.6034,
+      "step": 9430
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 36.70258270216636,
+      "learning_rate": 5.267425011265239e-06,
+      "loss": 0.6,
+      "step": 9435
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 20.528336862629228,
+      "learning_rate": 5.26248643918045e-06,
+      "loss": 0.5939,
+      "step": 9440
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 26.2131550907509,
+      "learning_rate": 5.2575476102980995e-06,
+      "loss": 0.5955,
+      "step": 9445
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 29.015949407973547,
+      "learning_rate": 5.252608529449973e-06,
+      "loss": 0.5867,
+      "step": 9450
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.981498137225529,
+      "learning_rate": 5.2476692014681095e-06,
+      "loss": 0.6001,
+      "step": 9455
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 28.895944965854035,
+      "learning_rate": 5.242729631184786e-06,
+      "loss": 0.6044,
+      "step": 9460
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.041760611777507,
+      "learning_rate": 5.237789823432517e-06,
+      "loss": 0.5943,
+      "step": 9465
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.521764242332235,
+      "learning_rate": 5.232849783044052e-06,
+      "loss": 0.5826,
+      "step": 9470
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.398708372866475,
+      "learning_rate": 5.227909514852361e-06,
+      "loss": 0.607,
+      "step": 9475
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.215730658188253,
+      "learning_rate": 5.222969023690645e-06,
+      "loss": 0.5931,
+      "step": 9480
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.599945335944277,
+      "learning_rate": 5.218028314392318e-06,
+      "loss": 0.6104,
+      "step": 9485
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.166393268974675,
+      "learning_rate": 5.213087391791013e-06,
+      "loss": 0.5799,
+      "step": 9490
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.588862984883596,
+      "learning_rate": 5.208146260720565e-06,
+      "loss": 0.5951,
+      "step": 9495
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.894396498559342,
+      "learning_rate": 5.203204926015014e-06,
+      "loss": 0.5937,
+      "step": 9500
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 27.202565212697937,
+      "learning_rate": 5.1982633925086035e-06,
+      "loss": 0.5935,
+      "step": 9505
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 23.281425496642882,
+      "learning_rate": 5.1933216650357685e-06,
+      "loss": 0.5886,
+      "step": 9510
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 22.44444843338854,
+      "learning_rate": 5.188379748431135e-06,
+      "loss": 0.6094,
+      "step": 9515
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.051494013021621,
+      "learning_rate": 5.1834376475295126e-06,
+      "loss": 0.5863,
+      "step": 9520
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 13.514138256595807,
+      "learning_rate": 5.17849536716589e-06,
+      "loss": 0.5763,
+      "step": 9525
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.214611501380019,
+      "learning_rate": 5.173552912175437e-06,
+      "loss": 0.5877,
+      "step": 9530
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.452926704996535,
+      "learning_rate": 5.168610287393489e-06,
+      "loss": 0.5908,
+      "step": 9535
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.43734189421902,
+      "learning_rate": 5.163667497655549e-06,
+      "loss": 0.5944,
+      "step": 9540
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.50824964629719,
+      "learning_rate": 5.15872454779728e-06,
+      "loss": 0.5913,
+      "step": 9545
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.772975708340726,
+      "learning_rate": 5.153781442654505e-06,
+      "loss": 0.5899,
+      "step": 9550
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.791507395852812,
+      "learning_rate": 5.148838187063199e-06,
+      "loss": 0.5664,
+      "step": 9555
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.19245147769233,
+      "learning_rate": 5.143894785859478e-06,
+      "loss": 0.579,
+      "step": 9560
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.188990357721228,
+      "learning_rate": 5.138951243879608e-06,
+      "loss": 0.5956,
+      "step": 9565
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.134172803566005,
+      "learning_rate": 5.134007565959986e-06,
+      "loss": 0.5759,
+      "step": 9570
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.083896517270064,
+      "learning_rate": 5.1290637569371504e-06,
+      "loss": 0.5877,
+      "step": 9575
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 19.32150938461579,
+      "learning_rate": 5.124119821647759e-06,
+      "loss": 0.5677,
+      "step": 9580
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.6213419594999525,
+      "learning_rate": 5.119175764928599e-06,
+      "loss": 0.5851,
+      "step": 9585
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.008101960171448,
+      "learning_rate": 5.114231591616573e-06,
+      "loss": 0.5703,
+      "step": 9590
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.606662403436859,
+      "learning_rate": 5.1092873065487e-06,
+      "loss": 0.5754,
+      "step": 9595
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.279716551224174,
+      "learning_rate": 5.104342914562107e-06,
+      "loss": 0.5706,
+      "step": 9600
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.533706059181452,
+      "learning_rate": 5.0993984204940265e-06,
+      "loss": 0.5842,
+      "step": 9605
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 27.84735675215916,
+      "learning_rate": 5.0944538291817904e-06,
+      "loss": 0.569,
+      "step": 9610
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.401334890929006,
+      "learning_rate": 5.089509145462825e-06,
+      "loss": 0.5832,
+      "step": 9615
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.823274768680948,
+      "learning_rate": 5.084564374174649e-06,
+      "loss": 0.5911,
+      "step": 9620
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.325720202355667,
+      "learning_rate": 5.079619520154865e-06,
+      "loss": 0.5927,
+      "step": 9625
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.803797542503668,
+      "learning_rate": 5.074674588241157e-06,
+      "loss": 0.5723,
+      "step": 9630
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.970088435158851,
+      "learning_rate": 5.069729583271285e-06,
+      "loss": 0.5959,
+      "step": 9635
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 24.213551346003864,
+      "learning_rate": 5.0647845100830805e-06,
+      "loss": 0.5713,
+      "step": 9640
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.224768652073715,
+      "learning_rate": 5.059839373514441e-06,
+      "loss": 0.5854,
+      "step": 9645
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.8733589879068955,
+      "learning_rate": 5.05489417840333e-06,
+      "loss": 0.5768,
+      "step": 9650
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.377198052908513,
+      "learning_rate": 5.049948929587764e-06,
+      "loss": 0.5836,
+      "step": 9655
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.73180875581141,
+      "learning_rate": 5.045003631905813e-06,
+      "loss": 0.5842,
+      "step": 9660
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 25.008213493513512,
+      "learning_rate": 5.040058290195594e-06,
+      "loss": 0.5947,
+      "step": 9665
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.41992480953314,
+      "learning_rate": 5.0351129092952685e-06,
+      "loss": 0.5729,
+      "step": 9670
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.591693362525524,
+      "learning_rate": 5.030167494043039e-06,
+      "loss": 0.5941,
+      "step": 9675
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.601293521527685,
+      "learning_rate": 5.025222049277136e-06,
+      "loss": 0.5862,
+      "step": 9680
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.893404495312224,
+      "learning_rate": 5.020276579835821e-06,
+      "loss": 0.5965,
+      "step": 9685
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.84532963343138,
+      "learning_rate": 5.0153310905573815e-06,
+      "loss": 0.5938,
+      "step": 9690
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.06878354567659,
+      "learning_rate": 5.0103855862801235e-06,
+      "loss": 0.5921,
+      "step": 9695
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 31.63918706961253,
+      "learning_rate": 5.005440071842365e-06,
+      "loss": 0.5951,
+      "step": 9700
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.578560024309926,
+      "learning_rate": 5.000494552082437e-06,
+      "loss": 0.5893,
+      "step": 9705
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 24.101825694531986,
+      "learning_rate": 4.995549031838675e-06,
+      "loss": 0.583,
+      "step": 9710
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 29.86430102484688,
+      "learning_rate": 4.990603515949416e-06,
+      "loss": 0.5828,
+      "step": 9715
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.822902890432811,
+      "learning_rate": 4.985658009252992e-06,
+      "loss": 0.5617,
+      "step": 9720
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 26.90223147024694,
+      "learning_rate": 4.980712516587724e-06,
+      "loss": 0.5908,
+      "step": 9725
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 34.950449323343015,
+      "learning_rate": 4.975767042791921e-06,
+      "loss": 0.5766,
+      "step": 9730
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.257839463549047,
+      "learning_rate": 4.970821592703874e-06,
+      "loss": 0.5859,
+      "step": 9735
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.386917175444268,
+      "learning_rate": 4.965876171161848e-06,
+      "loss": 0.5797,
+      "step": 9740
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 22.194715608074517,
+      "learning_rate": 4.960930783004085e-06,
+      "loss": 0.5698,
+      "step": 9745
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 21.670717750103282,
+      "learning_rate": 4.955985433068791e-06,
+      "loss": 0.5858,
+      "step": 9750
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 24.122686284627108,
+      "learning_rate": 4.951040126194135e-06,
+      "loss": 0.5921,
+      "step": 9755
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.764344436833607,
+      "learning_rate": 4.946094867218243e-06,
+      "loss": 0.5677,
+      "step": 9760
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 19.31689565138828,
+      "learning_rate": 4.941149660979201e-06,
+      "loss": 0.5904,
+      "step": 9765
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.942338302320184,
+      "learning_rate": 4.936204512315029e-06,
+      "loss": 0.5971,
+      "step": 9770
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.9944235406172,
+      "learning_rate": 4.931259426063704e-06,
+      "loss": 0.5843,
+      "step": 9775
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 21.039797821647916,
+      "learning_rate": 4.926314407063136e-06,
+      "loss": 0.577,
+      "step": 9780
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.865950004267344,
+      "learning_rate": 4.9213694601511714e-06,
+      "loss": 0.5809,
+      "step": 9785
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.363302596636814,
+      "learning_rate": 4.9164245901655845e-06,
+      "loss": 0.5825,
+      "step": 9790
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.5785200823388505,
+      "learning_rate": 4.911479801944076e-06,
+      "loss": 0.5878,
+      "step": 9795
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 29.945610829274326,
+      "learning_rate": 4.906535100324264e-06,
+      "loss": 0.5877,
+      "step": 9800
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.813855108965385,
+      "learning_rate": 4.901590490143686e-06,
+      "loss": 0.5727,
+      "step": 9805
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.498703494352979,
+      "learning_rate": 4.896645976239785e-06,
+      "loss": 0.5813,
+      "step": 9810
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.94551286609573,
+      "learning_rate": 4.8917015634499125e-06,
+      "loss": 0.5937,
+      "step": 9815
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.006180458225585,
+      "learning_rate": 4.886757256611323e-06,
+      "loss": 0.5783,
+      "step": 9820
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.992668492922249,
+      "learning_rate": 4.881813060561162e-06,
+      "loss": 0.5793,
+      "step": 9825
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.09319164500253,
+      "learning_rate": 4.876868980136472e-06,
+      "loss": 0.5774,
+      "step": 9830
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 19.408392042835462,
+      "learning_rate": 4.87192502017418e-06,
+      "loss": 0.5756,
+      "step": 9835
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 17.818354113938415,
+      "learning_rate": 4.866981185511095e-06,
+      "loss": 0.5755,
+      "step": 9840
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.504576142845876,
+      "learning_rate": 4.862037480983906e-06,
+      "loss": 0.5762,
+      "step": 9845
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 21.296112438487995,
+      "learning_rate": 4.857093911429169e-06,
+      "loss": 0.5928,
+      "step": 9850
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.99023787677679,
+      "learning_rate": 4.852150481683313e-06,
+      "loss": 0.5809,
+      "step": 9855
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.901437789610188,
+      "learning_rate": 4.847207196582628e-06,
+      "loss": 0.5883,
+      "step": 9860
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 14.855858691741162,
+      "learning_rate": 4.842264060963265e-06,
+      "loss": 0.5962,
+      "step": 9865
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 27.968729321502533,
+      "learning_rate": 4.837321079661225e-06,
+      "loss": 0.5897,
+      "step": 9870
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.75320400924953,
+      "learning_rate": 4.83237825751236e-06,
+      "loss": 0.5836,
+      "step": 9875
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.5378797267475,
+      "learning_rate": 4.827435599352367e-06,
+      "loss": 0.5679,
+      "step": 9880
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 26.342093199593943,
+      "learning_rate": 4.822493110016785e-06,
+      "loss": 0.5743,
+      "step": 9885
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.642984156198046,
+      "learning_rate": 4.817550794340977e-06,
+      "loss": 0.5702,
+      "step": 9890
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.668599007883666,
+      "learning_rate": 4.812608657160149e-06,
+      "loss": 0.5739,
+      "step": 9895
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.331624056006664,
+      "learning_rate": 4.807666703309327e-06,
+      "loss": 0.5671,
+      "step": 9900
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 25.002481425875065,
+      "learning_rate": 4.802724937623355e-06,
+      "loss": 0.5769,
+      "step": 9905
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 18.291991609730832,
+      "learning_rate": 4.7977833649369e-06,
+      "loss": 0.5797,
+      "step": 9910
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.706725759701401,
+      "learning_rate": 4.7928419900844316e-06,
+      "loss": 0.5755,
+      "step": 9915
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.17580923262539,
+      "learning_rate": 4.787900817900232e-06,
+      "loss": 0.5749,
+      "step": 9920
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.019499640302362,
+      "learning_rate": 4.782959853218386e-06,
+      "loss": 0.5819,
+      "step": 9925
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.759696126688855,
+      "learning_rate": 4.778019100872767e-06,
+      "loss": 0.5792,
+      "step": 9930
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 26.831148200227613,
+      "learning_rate": 4.773078565697048e-06,
+      "loss": 0.5565,
+      "step": 9935
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.270356675527533,
+      "learning_rate": 4.76813825252469e-06,
+      "loss": 0.5843,
+      "step": 9940
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.231540027130091,
+      "learning_rate": 4.763198166188933e-06,
+      "loss": 0.5774,
+      "step": 9945
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.9990434864943545,
+      "learning_rate": 4.758258311522798e-06,
+      "loss": 0.5961,
+      "step": 9950
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.583122420995502,
+      "learning_rate": 4.7533186933590766e-06,
+      "loss": 0.5779,
+      "step": 9955
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.375315990982129,
+      "learning_rate": 4.748379316530331e-06,
+      "loss": 0.5557,
+      "step": 9960
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 33.351291331106125,
+      "learning_rate": 4.743440185868888e-06,
+      "loss": 0.5775,
+      "step": 9965
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.461224217936112,
+      "learning_rate": 4.738501306206831e-06,
+      "loss": 0.5718,
+      "step": 9970
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 20.359831655936695,
+      "learning_rate": 4.733562682375999e-06,
+      "loss": 0.5673,
+      "step": 9975
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.654800710201974,
+      "learning_rate": 4.728624319207979e-06,
+      "loss": 0.5715,
+      "step": 9980
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.9141816673019,
+      "learning_rate": 4.723686221534109e-06,
+      "loss": 0.5612,
+      "step": 9985
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.915149604282366,
+      "learning_rate": 4.7187483941854615e-06,
+      "loss": 0.5743,
+      "step": 9990
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.012368837366665,
+      "learning_rate": 4.713810841992845e-06,
+      "loss": 0.5684,
+      "step": 9995
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 23.658206543376345,
+      "learning_rate": 4.708873569786803e-06,
+      "loss": 0.588,
+      "step": 10000
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.610097088990205,
+      "learning_rate": 4.7039365823976e-06,
+      "loss": 0.557,
+      "step": 10005
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.946014682934994,
+      "learning_rate": 4.6989998846552234e-06,
+      "loss": 0.5691,
+      "step": 10010
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.491836666841432,
+      "learning_rate": 4.694063481389377e-06,
+      "loss": 0.5755,
+      "step": 10015
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.281593897023107,
+      "learning_rate": 4.68912737742948e-06,
+      "loss": 0.5824,
+      "step": 10020
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.680821452661381,
+      "learning_rate": 4.684191577604653e-06,
+      "loss": 0.5528,
+      "step": 10025
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.924164351060437,
+      "learning_rate": 4.679256086743725e-06,
+      "loss": 0.5757,
+      "step": 10030
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.566207423001996,
+      "learning_rate": 4.674320909675218e-06,
+      "loss": 0.5593,
+      "step": 10035
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.428569933754627,
+      "learning_rate": 4.66938605122735e-06,
+      "loss": 0.5751,
+      "step": 10040
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.645878410906143,
+      "learning_rate": 4.664451516228027e-06,
+      "loss": 0.5799,
+      "step": 10045
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.705526960881357,
+      "learning_rate": 4.659517309504834e-06,
+      "loss": 0.5736,
+      "step": 10050
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.44097547370709,
+      "learning_rate": 4.6545834358850415e-06,
+      "loss": 0.5816,
+      "step": 10055
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.76221489972865,
+      "learning_rate": 4.649649900195591e-06,
+      "loss": 0.5735,
+      "step": 10060
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.732878022187395,
+      "learning_rate": 4.644716707263091e-06,
+      "loss": 0.579,
+      "step": 10065
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.972228567376378,
+      "learning_rate": 4.6397838619138205e-06,
+      "loss": 0.574,
+      "step": 10070
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.090807111400876,
+      "learning_rate": 4.634851368973713e-06,
+      "loss": 0.5678,
+      "step": 10075
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.469211038126204,
+      "learning_rate": 4.6299192332683605e-06,
+      "loss": 0.5764,
+      "step": 10080
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.812271489373423,
+      "learning_rate": 4.6249874596230056e-06,
+      "loss": 0.5643,
+      "step": 10085
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.540827875393065,
+      "learning_rate": 4.620056052862532e-06,
+      "loss": 0.5597,
+      "step": 10090
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.747842561670529,
+      "learning_rate": 4.615125017811471e-06,
+      "loss": 0.5727,
+      "step": 10095
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.5164902547307815,
+      "learning_rate": 4.6101943592939855e-06,
+      "loss": 0.5644,
+      "step": 10100
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 27.124478750771498,
+      "learning_rate": 4.605264082133872e-06,
+      "loss": 0.5743,
+      "step": 10105
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.513661237874295,
+      "learning_rate": 4.600334191154554e-06,
+      "loss": 0.555,
+      "step": 10110
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.864940327869741,
+      "learning_rate": 4.595404691179077e-06,
+      "loss": 0.5547,
+      "step": 10115
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.06766107265483,
+      "learning_rate": 4.5904755870301035e-06,
+      "loss": 0.5793,
+      "step": 10120
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.085298717926536,
+      "learning_rate": 4.585546883529911e-06,
+      "loss": 0.578,
+      "step": 10125
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.041429483784615,
+      "learning_rate": 4.5806185855003786e-06,
+      "loss": 0.5677,
+      "step": 10130
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.799211328435109,
+      "learning_rate": 4.575690697762996e-06,
+      "loss": 0.5665,
+      "step": 10135
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.448452244387205,
+      "learning_rate": 4.5707632251388484e-06,
+      "loss": 0.5753,
+      "step": 10140
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 26.61989099099489,
+      "learning_rate": 4.5658361724486165e-06,
+      "loss": 0.5736,
+      "step": 10145
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.459385864952437,
+      "learning_rate": 4.5609095445125665e-06,
+      "loss": 0.5675,
+      "step": 10150
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.14015548413005,
+      "learning_rate": 4.555983346150551e-06,
+      "loss": 0.552,
+      "step": 10155
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.124694141352393,
+      "learning_rate": 4.551057582182005e-06,
+      "loss": 0.5673,
+      "step": 10160
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 22.63468479100616,
+      "learning_rate": 4.546132257425939e-06,
+      "loss": 0.5709,
+      "step": 10165
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.668158526720532,
+      "learning_rate": 4.541207376700924e-06,
+      "loss": 0.5714,
+      "step": 10170
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.798478426561271,
+      "learning_rate": 4.5362829448251076e-06,
+      "loss": 0.5721,
+      "step": 10175
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.943369172881516,
+      "learning_rate": 4.5313589666161935e-06,
+      "loss": 0.5763,
+      "step": 10180
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.960859334311488,
+      "learning_rate": 4.5264354468914425e-06,
+      "loss": 0.5775,
+      "step": 10185
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 15.36159808951015,
+      "learning_rate": 4.521512390467668e-06,
+      "loss": 0.5545,
+      "step": 10190
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.850169100527044,
+      "learning_rate": 4.516589802161228e-06,
+      "loss": 0.5657,
+      "step": 10195
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.8472842610984985,
+      "learning_rate": 4.511667686788022e-06,
+      "loss": 0.5613,
+      "step": 10200
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.381231330134624,
+      "learning_rate": 4.50674604916349e-06,
+      "loss": 0.5597,
+      "step": 10205
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.380650887041463,
+      "learning_rate": 4.501824894102604e-06,
+      "loss": 0.5746,
+      "step": 10210
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.9616226566722625,
+      "learning_rate": 4.49690422641986e-06,
+      "loss": 0.568,
+      "step": 10215
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.128136408250823,
+      "learning_rate": 4.49198405092928e-06,
+      "loss": 0.5635,
+      "step": 10220
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.7778297659247375,
+      "learning_rate": 4.487064372444406e-06,
+      "loss": 0.5661,
+      "step": 10225
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.148750564471536,
+      "learning_rate": 4.4821451957782915e-06,
+      "loss": 0.559,
+      "step": 10230
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.749591740727813,
+      "learning_rate": 4.4772265257435e-06,
+      "loss": 0.5698,
+      "step": 10235
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.949616805563828,
+      "learning_rate": 4.472308367152098e-06,
+      "loss": 0.5518,
+      "step": 10240
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.755053911637036,
+      "learning_rate": 4.467390724815654e-06,
+      "loss": 0.572,
+      "step": 10245
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.60832823711757,
+      "learning_rate": 4.462473603545232e-06,
+      "loss": 0.5794,
+      "step": 10250
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.883077773789456,
+      "learning_rate": 4.457557008151379e-06,
+      "loss": 0.5713,
+      "step": 10255
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.519039378885232,
+      "learning_rate": 4.452640943444137e-06,
+      "loss": 0.5762,
+      "step": 10260
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.12176426140937,
+      "learning_rate": 4.447725414233024e-06,
+      "loss": 0.5575,
+      "step": 10265
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.597445584862303,
+      "learning_rate": 4.442810425327033e-06,
+      "loss": 0.5682,
+      "step": 10270
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.899622508028797,
+      "learning_rate": 4.437895981534632e-06,
+      "loss": 0.5663,
+      "step": 10275
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.23241821477259,
+      "learning_rate": 4.432982087663755e-06,
+      "loss": 0.5706,
+      "step": 10280
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.064482315135152,
+      "learning_rate": 4.428068748521794e-06,
+      "loss": 0.5626,
+      "step": 10285
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.224225338476222,
+      "learning_rate": 4.423155968915605e-06,
+      "loss": 0.5682,
+      "step": 10290
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.104061848723966,
+      "learning_rate": 4.418243753651488e-06,
+      "loss": 0.5513,
+      "step": 10295
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.6983397688120485,
+      "learning_rate": 4.413332107535199e-06,
+      "loss": 0.575,
+      "step": 10300
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.031076235814748,
+      "learning_rate": 4.408421035371932e-06,
+      "loss": 0.5583,
+      "step": 10305
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 23.865179039994,
+      "learning_rate": 4.4035105419663234e-06,
+      "loss": 0.5685,
+      "step": 10310
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.3402880658064,
+      "learning_rate": 4.39860063212244e-06,
+      "loss": 0.5672,
+      "step": 10315
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.376246146568219,
+      "learning_rate": 4.393691310643779e-06,
+      "loss": 0.5549,
+      "step": 10320
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.5076088706963535,
+      "learning_rate": 4.388782582333263e-06,
+      "loss": 0.5456,
+      "step": 10325
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.076302725867919,
+      "learning_rate": 4.3838744519932345e-06,
+      "loss": 0.5735,
+      "step": 10330
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.155201996360862,
+      "learning_rate": 4.378966924425447e-06,
+      "loss": 0.5672,
+      "step": 10335
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.51270787542137,
+      "learning_rate": 4.3740600044310664e-06,
+      "loss": 0.569,
+      "step": 10340
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.943870912242375,
+      "learning_rate": 4.3691536968106675e-06,
+      "loss": 0.5794,
+      "step": 10345
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 16.036753964316397,
+      "learning_rate": 4.364248006364222e-06,
+      "loss": 0.5575,
+      "step": 10350
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.337387313826945,
+      "learning_rate": 4.359342937891099e-06,
+      "loss": 0.56,
+      "step": 10355
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.632921011142148,
+      "learning_rate": 4.354438496190061e-06,
+      "loss": 0.5733,
+      "step": 10360
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.001797244074437,
+      "learning_rate": 4.349534686059255e-06,
+      "loss": 0.5841,
+      "step": 10365
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.677734308629427,
+      "learning_rate": 4.344631512296211e-06,
+      "loss": 0.5562,
+      "step": 10370
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.974231493449727,
+      "learning_rate": 4.3397289796978335e-06,
+      "loss": 0.5582,
+      "step": 10375
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.82580270027655,
+      "learning_rate": 4.334827093060406e-06,
+      "loss": 0.5525,
+      "step": 10380
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.300464842483924,
+      "learning_rate": 4.329925857179573e-06,
+      "loss": 0.5677,
+      "step": 10385
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.408626674668492,
+      "learning_rate": 4.325025276850347e-06,
+      "loss": 0.556,
+      "step": 10390
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.281700862386464,
+      "learning_rate": 4.3201253568671e-06,
+      "loss": 0.5579,
+      "step": 10395
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.4899340428012025,
+      "learning_rate": 4.3152261020235516e-06,
+      "loss": 0.5609,
+      "step": 10400
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.976312177486976,
+      "learning_rate": 4.31032751711278e-06,
+      "loss": 0.5598,
+      "step": 10405
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 17.343807709088885,
+      "learning_rate": 4.305429606927202e-06,
+      "loss": 0.5547,
+      "step": 10410
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.69312788328246,
+      "learning_rate": 4.300532376258571e-06,
+      "loss": 0.5574,
+      "step": 10415
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.87987772594223,
+      "learning_rate": 4.295635829897983e-06,
+      "loss": 0.5494,
+      "step": 10420
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.887701568026129,
+      "learning_rate": 4.2907399726358626e-06,
+      "loss": 0.5559,
+      "step": 10425
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 24.364887801382608,
+      "learning_rate": 4.285844809261955e-06,
+      "loss": 0.556,
+      "step": 10430
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.924058864204879,
+      "learning_rate": 4.280950344565335e-06,
+      "loss": 0.5576,
+      "step": 10435
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.91510691811666,
+      "learning_rate": 4.276056583334386e-06,
+      "loss": 0.5637,
+      "step": 10440
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.736864177986743,
+      "learning_rate": 4.271163530356808e-06,
+      "loss": 0.5521,
+      "step": 10445
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 28.15747552724219,
+      "learning_rate": 4.266271190419609e-06,
+      "loss": 0.562,
+      "step": 10450
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 17.159897130514594,
+      "learning_rate": 4.261379568309093e-06,
+      "loss": 0.574,
+      "step": 10455
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.932452219225763,
+      "learning_rate": 4.256488668810868e-06,
+      "loss": 0.5581,
+      "step": 10460
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.313755617902103,
+      "learning_rate": 4.251598496709832e-06,
+      "loss": 0.5526,
+      "step": 10465
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.111505571206518,
+      "learning_rate": 4.2467090567901735e-06,
+      "loss": 0.5582,
+      "step": 10470
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 17.77722314473156,
+      "learning_rate": 4.241820353835363e-06,
+      "loss": 0.556,
+      "step": 10475
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 14.39601648137402,
+      "learning_rate": 4.236932392628149e-06,
+      "loss": 0.5528,
+      "step": 10480
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 24.56336728851155,
+      "learning_rate": 4.2320451779505575e-06,
+      "loss": 0.5469,
+      "step": 10485
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.487147299657114,
+      "learning_rate": 4.227158714583884e-06,
+      "loss": 0.5588,
+      "step": 10490
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.364733268739467,
+      "learning_rate": 4.222273007308684e-06,
+      "loss": 0.5639,
+      "step": 10495
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.441684490504245,
+      "learning_rate": 4.217388060904778e-06,
+      "loss": 0.5534,
+      "step": 10500
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.84464990205949,
+      "learning_rate": 4.21250388015124e-06,
+      "loss": 0.5473,
+      "step": 10505
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.347143338237716,
+      "learning_rate": 4.207620469826397e-06,
+      "loss": 0.5563,
+      "step": 10510
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 20.032525340936964,
+      "learning_rate": 4.2027378347078225e-06,
+      "loss": 0.541,
+      "step": 10515
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 22.353036370097144,
+      "learning_rate": 4.197855979572326e-06,
+      "loss": 0.5461,
+      "step": 10520
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.145827342584123,
+      "learning_rate": 4.192974909195962e-06,
+      "loss": 0.5538,
+      "step": 10525
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.284130529239555,
+      "learning_rate": 4.188094628354013e-06,
+      "loss": 0.5523,
+      "step": 10530
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 21.944753217572533,
+      "learning_rate": 4.1832151418209865e-06,
+      "loss": 0.5497,
+      "step": 10535
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.726946766485042,
+      "learning_rate": 4.1783364543706165e-06,
+      "loss": 0.5385,
+      "step": 10540
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.947396885772298,
+      "learning_rate": 4.173458570775856e-06,
+      "loss": 0.5655,
+      "step": 10545
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.966857276634125,
+      "learning_rate": 4.1685814958088696e-06,
+      "loss": 0.5756,
+      "step": 10550
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.402810816083218,
+      "learning_rate": 4.1637052342410315e-06,
+      "loss": 0.5623,
+      "step": 10555
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.900704617475461,
+      "learning_rate": 4.1588297908429195e-06,
+      "loss": 0.54,
+      "step": 10560
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.269730671746444,
+      "learning_rate": 4.153955170384312e-06,
+      "loss": 0.5483,
+      "step": 10565
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 18.570352607344855,
+      "learning_rate": 4.149081377634182e-06,
+      "loss": 0.5409,
+      "step": 10570
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.479458512559505,
+      "learning_rate": 4.14420841736069e-06,
+      "loss": 0.5487,
+      "step": 10575
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 20.410675325402618,
+      "learning_rate": 4.1393362943311866e-06,
+      "loss": 0.5704,
+      "step": 10580
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.033492396039378,
+      "learning_rate": 4.1344650133122e-06,
+      "loss": 0.5527,
+      "step": 10585
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.608181918412429,
+      "learning_rate": 4.129594579069436e-06,
+      "loss": 0.5503,
+      "step": 10590
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 16.721543496224996,
+      "learning_rate": 4.1247249963677725e-06,
+      "loss": 0.5527,
+      "step": 10595
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.756527197932947,
+      "learning_rate": 4.119856269971254e-06,
+      "loss": 0.5569,
+      "step": 10600
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.928787184558088,
+      "learning_rate": 4.114988404643086e-06,
+      "loss": 0.551,
+      "step": 10605
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 27.003024941869466,
+      "learning_rate": 4.110121405145634e-06,
+      "loss": 0.5454,
+      "step": 10610
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 24.19555970041987,
+      "learning_rate": 4.105255276240413e-06,
+      "loss": 0.5502,
+      "step": 10615
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.339057342338457,
+      "learning_rate": 4.100390022688087e-06,
+      "loss": 0.5623,
+      "step": 10620
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.422037220563851,
+      "learning_rate": 4.095525649248467e-06,
+      "loss": 0.544,
+      "step": 10625
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 14.684061712602356,
+      "learning_rate": 4.0906621606805e-06,
+      "loss": 0.5611,
+      "step": 10630
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 19.907732420246848,
+      "learning_rate": 4.085799561742269e-06,
+      "loss": 0.5619,
+      "step": 10635
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.149132567592444,
+      "learning_rate": 4.080937857190984e-06,
+      "loss": 0.5567,
+      "step": 10640
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 17.153750553470985,
+      "learning_rate": 4.076077051782983e-06,
+      "loss": 0.5652,
+      "step": 10645
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.938875135207251,
+      "learning_rate": 4.0712171502737245e-06,
+      "loss": 0.5475,
+      "step": 10650
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 28.712317930091356,
+      "learning_rate": 4.0663581574177764e-06,
+      "loss": 0.5492,
+      "step": 10655
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.907991497479526,
+      "learning_rate": 4.061500077968829e-06,
+      "loss": 0.5519,
+      "step": 10660
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.26578908835014,
+      "learning_rate": 4.056642916679666e-06,
+      "loss": 0.5678,
+      "step": 10665
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.9969212767908955,
+      "learning_rate": 4.051786678302182e-06,
+      "loss": 0.5542,
+      "step": 10670
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.811354898565146,
+      "learning_rate": 4.046931367587367e-06,
+      "loss": 0.5547,
+      "step": 10675
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.351676942851948,
+      "learning_rate": 4.042076989285301e-06,
+      "loss": 0.5487,
+      "step": 10680
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.690168235430562,
+      "learning_rate": 4.037223548145155e-06,
+      "loss": 0.5376,
+      "step": 10685
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.10316249163135,
+      "learning_rate": 4.0323710489151816e-06,
+      "loss": 0.5542,
+      "step": 10690
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.238263388190468,
+      "learning_rate": 4.027519496342707e-06,
+      "loss": 0.5477,
+      "step": 10695
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.111697756767847,
+      "learning_rate": 4.0226688951741415e-06,
+      "loss": 0.541,
+      "step": 10700
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.312504651719623,
+      "learning_rate": 4.017819250154957e-06,
+      "loss": 0.5625,
+      "step": 10705
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.855123982730133,
+      "learning_rate": 4.01297056602969e-06,
+      "loss": 0.5456,
+      "step": 10710
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.61601724706946,
+      "learning_rate": 4.008122847541942e-06,
+      "loss": 0.5484,
+      "step": 10715
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.541933682803487,
+      "learning_rate": 4.003276099434365e-06,
+      "loss": 0.5408,
+      "step": 10720
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.02298738803633,
+      "learning_rate": 3.998430326448664e-06,
+      "loss": 0.5623,
+      "step": 10725
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.4550356670551,
+      "learning_rate": 3.993585533325591e-06,
+      "loss": 0.5351,
+      "step": 10730
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.076748410436516,
+      "learning_rate": 3.988741724804935e-06,
+      "loss": 0.5471,
+      "step": 10735
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.060203765038019,
+      "learning_rate": 3.983898905625525e-06,
+      "loss": 0.5606,
+      "step": 10740
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.697863097937426,
+      "learning_rate": 3.979057080525223e-06,
+      "loss": 0.5733,
+      "step": 10745
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.600904312186005,
+      "learning_rate": 3.974216254240917e-06,
+      "loss": 0.5571,
+      "step": 10750
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.41737306070303,
+      "learning_rate": 3.969376431508516e-06,
+      "loss": 0.5434,
+      "step": 10755
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.975078682574957,
+      "learning_rate": 3.964537617062951e-06,
+      "loss": 0.5439,
+      "step": 10760
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 17.113910948663463,
+      "learning_rate": 3.959699815638163e-06,
+      "loss": 0.55,
+      "step": 10765
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.726618538250287,
+      "learning_rate": 3.954863031967108e-06,
+      "loss": 0.5516,
+      "step": 10770
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.149850327053818,
+      "learning_rate": 3.950027270781736e-06,
+      "loss": 0.5534,
+      "step": 10775
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.751245969017911,
+      "learning_rate": 3.945192536813006e-06,
+      "loss": 0.5286,
+      "step": 10780
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.60655821899745,
+      "learning_rate": 3.940358834790867e-06,
+      "loss": 0.5523,
+      "step": 10785
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.613630439695627,
+      "learning_rate": 3.935526169444261e-06,
+      "loss": 0.5413,
+      "step": 10790
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.455914843357684,
+      "learning_rate": 3.930694545501117e-06,
+      "loss": 0.5377,
+      "step": 10795
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.107483270728077,
+      "learning_rate": 3.925863967688339e-06,
+      "loss": 0.5428,
+      "step": 10800
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 24.49841838362915,
+      "learning_rate": 3.921034440731813e-06,
+      "loss": 0.5484,
+      "step": 10805
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.02938698091131,
+      "learning_rate": 3.916205969356399e-06,
+      "loss": 0.5291,
+      "step": 10810
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.780566938233218,
+      "learning_rate": 3.911378558285915e-06,
+      "loss": 0.5498,
+      "step": 10815
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.707041202570831,
+      "learning_rate": 3.906552212243151e-06,
+      "loss": 0.558,
+      "step": 10820
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.845285542740227,
+      "learning_rate": 3.90172693594985e-06,
+      "loss": 0.54,
+      "step": 10825
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.30898838434219,
+      "learning_rate": 3.89690273412671e-06,
+      "loss": 0.5421,
+      "step": 10830
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.456838345718717,
+      "learning_rate": 3.892079611493379e-06,
+      "loss": 0.5268,
+      "step": 10835
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.483645969550005,
+      "learning_rate": 3.8872575727684485e-06,
+      "loss": 0.5467,
+      "step": 10840
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.01162771308466,
+      "learning_rate": 3.882436622669447e-06,
+      "loss": 0.5395,
+      "step": 10845
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.646301598055061,
+      "learning_rate": 3.877616765912843e-06,
+      "loss": 0.5301,
+      "step": 10850
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.206351159952076,
+      "learning_rate": 3.872798007214028e-06,
+      "loss": 0.5457,
+      "step": 10855
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.460428079828267,
+      "learning_rate": 3.867980351287326e-06,
+      "loss": 0.5459,
+      "step": 10860
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.058509320670087,
+      "learning_rate": 3.863163802845979e-06,
+      "loss": 0.5513,
+      "step": 10865
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.923932413895969,
+      "learning_rate": 3.858348366602147e-06,
+      "loss": 0.5312,
+      "step": 10870
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.311599242077515,
+      "learning_rate": 3.853534047266902e-06,
+      "loss": 0.5456,
+      "step": 10875
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.53201788074447,
+      "learning_rate": 3.848720849550221e-06,
+      "loss": 0.5415,
+      "step": 10880
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.801911731802575,
+      "learning_rate": 3.843908778160986e-06,
+      "loss": 0.5391,
+      "step": 10885
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.338680249038083,
+      "learning_rate": 3.839097837806977e-06,
+      "loss": 0.5482,
+      "step": 10890
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.535259979212849,
+      "learning_rate": 3.834288033194864e-06,
+      "loss": 0.546,
+      "step": 10895
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.935828201938994,
+      "learning_rate": 3.829479369030211e-06,
+      "loss": 0.5377,
+      "step": 10900
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.688589701889827,
+      "learning_rate": 3.824671850017462e-06,
+      "loss": 0.5372,
+      "step": 10905
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 31.066244041899697,
+      "learning_rate": 3.819865480859943e-06,
+      "loss": 0.5361,
+      "step": 10910
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.014970505246703,
+      "learning_rate": 3.815060266259856e-06,
+      "loss": 0.5493,
+      "step": 10915
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.526054798739284,
+      "learning_rate": 3.8102562109182713e-06,
+      "loss": 0.543,
+      "step": 10920
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.449964955547665,
+      "learning_rate": 3.805453319535126e-06,
+      "loss": 0.5313,
+      "step": 10925
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 17.86887071360535,
+      "learning_rate": 3.8006515968092176e-06,
+      "loss": 0.5498,
+      "step": 10930
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.578980091545276,
+      "learning_rate": 3.7958510474382027e-06,
+      "loss": 0.5485,
+      "step": 10935
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.05700161143195,
+      "learning_rate": 3.7910516761185864e-06,
+      "loss": 0.5416,
+      "step": 10940
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.476537423347814,
+      "learning_rate": 3.7862534875457226e-06,
+      "loss": 0.539,
+      "step": 10945
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.67141676921221,
+      "learning_rate": 3.781456486413809e-06,
+      "loss": 0.5176,
+      "step": 10950
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.665045618301248,
+      "learning_rate": 3.7766606774158828e-06,
+      "loss": 0.5338,
+      "step": 10955
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.702942638630219,
+      "learning_rate": 3.7718660652438115e-06,
+      "loss": 0.5382,
+      "step": 10960
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.780251465298803,
+      "learning_rate": 3.7670726545882945e-06,
+      "loss": 0.5342,
+      "step": 10965
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.284822222671517,
+      "learning_rate": 3.7622804501388554e-06,
+      "loss": 0.5461,
+      "step": 10970
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.327317512329649,
+      "learning_rate": 3.7574894565838364e-06,
+      "loss": 0.5494,
+      "step": 10975
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.938292463573944,
+      "learning_rate": 3.752699678610395e-06,
+      "loss": 0.5337,
+      "step": 10980
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.505501291362319,
+      "learning_rate": 3.747911120904501e-06,
+      "loss": 0.5291,
+      "step": 10985
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.4303775627203015,
+      "learning_rate": 3.7431237881509287e-06,
+      "loss": 0.5246,
+      "step": 10990
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.991768485557685,
+      "learning_rate": 3.7383376850332546e-06,
+      "loss": 0.5243,
+      "step": 10995
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.495221718797736,
+      "learning_rate": 3.733552816233854e-06,
+      "loss": 0.5337,
+      "step": 11000
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.162826956882016,
+      "learning_rate": 3.7287691864338926e-06,
+      "loss": 0.5384,
+      "step": 11005
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.199569881998118,
+      "learning_rate": 3.723986800313324e-06,
+      "loss": 0.5195,
+      "step": 11010
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.756550802893384,
+      "learning_rate": 3.7192056625508877e-06,
+      "loss": 0.539,
+      "step": 11015
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.743213887283496,
+      "learning_rate": 3.7144257778240955e-06,
+      "loss": 0.5365,
+      "step": 11020
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.84475921976795,
+      "learning_rate": 3.70964715080924e-06,
+      "loss": 0.5423,
+      "step": 11025
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.338435642943008,
+      "learning_rate": 3.704869786181382e-06,
+      "loss": 0.5352,
+      "step": 11030
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.387321434056955,
+      "learning_rate": 3.700093688614344e-06,
+      "loss": 0.5361,
+      "step": 11035
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.6140088905394885,
+      "learning_rate": 3.695318862780712e-06,
+      "loss": 0.5336,
+      "step": 11040
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.222058289128649,
+      "learning_rate": 3.6905453133518266e-06,
+      "loss": 0.5342,
+      "step": 11045
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.083708298443856,
+      "learning_rate": 3.6857730449977807e-06,
+      "loss": 0.532,
+      "step": 11050
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.372511390623997,
+      "learning_rate": 3.6810020623874143e-06,
+      "loss": 0.5339,
+      "step": 11055
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.828966228991336,
+      "learning_rate": 3.676232370188305e-06,
+      "loss": 0.5449,
+      "step": 11060
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.061506476661475,
+      "learning_rate": 3.6714639730667733e-06,
+      "loss": 0.5439,
+      "step": 11065
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.986951778483274,
+      "learning_rate": 3.6666968756878706e-06,
+      "loss": 0.5391,
+      "step": 11070
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.697073681510214,
+      "learning_rate": 3.6619310827153777e-06,
+      "loss": 0.5255,
+      "step": 11075
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.48917727035367,
+      "learning_rate": 3.6571665988117964e-06,
+      "loss": 0.5353,
+      "step": 11080
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.95409608824707,
+      "learning_rate": 3.6524034286383512e-06,
+      "loss": 0.5388,
+      "step": 11085
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.739997106110412,
+      "learning_rate": 3.647641576854979e-06,
+      "loss": 0.5397,
+      "step": 11090
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.702913793269424,
+      "learning_rate": 3.6428810481203314e-06,
+      "loss": 0.5392,
+      "step": 11095
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.152523882009815,
+      "learning_rate": 3.6381218470917566e-06,
+      "loss": 0.5367,
+      "step": 11100
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 14.924401718875243,
+      "learning_rate": 3.6333639784253116e-06,
+      "loss": 0.5328,
+      "step": 11105
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.220084367718904,
+      "learning_rate": 3.6286074467757488e-06,
+      "loss": 0.5378,
+      "step": 11110
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.6355010730653685,
+      "learning_rate": 3.623852256796511e-06,
+      "loss": 0.5417,
+      "step": 11115
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.697080845793394,
+      "learning_rate": 3.6190984131397277e-06,
+      "loss": 0.5458,
+      "step": 11120
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.4948154448340105,
+      "learning_rate": 3.6143459204562128e-06,
+      "loss": 0.529,
+      "step": 11125
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.98852882390966,
+      "learning_rate": 3.609594783395458e-06,
+      "loss": 0.5299,
+      "step": 11130
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.254750013758175,
+      "learning_rate": 3.604845006605632e-06,
+      "loss": 0.5268,
+      "step": 11135
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.746954504175433,
+      "learning_rate": 3.600096594733564e-06,
+      "loss": 0.5483,
+      "step": 11140
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.365857124321258,
+      "learning_rate": 3.5953495524247573e-06,
+      "loss": 0.5281,
+      "step": 11145
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.329693543819747,
+      "learning_rate": 3.5906038843233693e-06,
+      "loss": 0.5374,
+      "step": 11150
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.801535909625722,
+      "learning_rate": 3.585859595072216e-06,
+      "loss": 0.5276,
+      "step": 11155
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.397119396349337,
+      "learning_rate": 3.5811166893127646e-06,
+      "loss": 0.5316,
+      "step": 11160
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 21.842813538079348,
+      "learning_rate": 3.576375171685126e-06,
+      "loss": 0.5408,
+      "step": 11165
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.231416398872692,
+      "learning_rate": 3.5716350468280553e-06,
+      "loss": 0.5301,
+      "step": 11170
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.519093114896531,
+      "learning_rate": 3.566896319378947e-06,
+      "loss": 0.5304,
+      "step": 11175
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.940488754631659,
+      "learning_rate": 3.562158993973821e-06,
+      "loss": 0.5465,
+      "step": 11180
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.299602282158513,
+      "learning_rate": 3.5574230752473336e-06,
+      "loss": 0.5242,
+      "step": 11185
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.554576788811912,
+      "learning_rate": 3.5526885678327617e-06,
+      "loss": 0.5384,
+      "step": 11190
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.904279904285887,
+      "learning_rate": 3.5479554763620016e-06,
+      "loss": 0.5238,
+      "step": 11195
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.20256437712205,
+      "learning_rate": 3.5432238054655633e-06,
+      "loss": 0.5232,
+      "step": 11200
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.122344904169903,
+      "learning_rate": 3.53849355977257e-06,
+      "loss": 0.5497,
+      "step": 11205
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.152097367658841,
+      "learning_rate": 3.533764743910747e-06,
+      "loss": 0.5314,
+      "step": 11210
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.84948937139005,
+      "learning_rate": 3.529037362506424e-06,
+      "loss": 0.537,
+      "step": 11215
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.577212182155189,
+      "learning_rate": 3.5243114201845242e-06,
+      "loss": 0.5112,
+      "step": 11220
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.170134880436668,
+      "learning_rate": 3.519586921568564e-06,
+      "loss": 0.5338,
+      "step": 11225
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.220419192071601,
+      "learning_rate": 3.5148638712806486e-06,
+      "loss": 0.5082,
+      "step": 11230
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.7340417910071455,
+      "learning_rate": 3.5101422739414657e-06,
+      "loss": 0.5183,
+      "step": 11235
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.07802322303307,
+      "learning_rate": 3.5054221341702815e-06,
+      "loss": 0.5209,
+      "step": 11240
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.700269535527903,
+      "learning_rate": 3.500703456584935e-06,
+      "loss": 0.5133,
+      "step": 11245
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 19.52570025614485,
+      "learning_rate": 3.495986245801839e-06,
+      "loss": 0.5262,
+      "step": 11250
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.298198955013202,
+      "learning_rate": 3.4912705064359643e-06,
+      "loss": 0.5205,
+      "step": 11255
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.119842025529667,
+      "learning_rate": 3.486556243100847e-06,
+      "loss": 0.5261,
+      "step": 11260
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.108706346417808,
+      "learning_rate": 3.481843460408579e-06,
+      "loss": 0.5328,
+      "step": 11265
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.859265643783456,
+      "learning_rate": 3.4771321629698008e-06,
+      "loss": 0.5452,
+      "step": 11270
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.659912383492827,
+      "learning_rate": 3.472422355393703e-06,
+      "loss": 0.5287,
+      "step": 11275
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.979094507974204,
+      "learning_rate": 3.4677140422880172e-06,
+      "loss": 0.5408,
+      "step": 11280
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.489033753103027,
+      "learning_rate": 3.4630072282590135e-06,
+      "loss": 0.5448,
+      "step": 11285
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.576397068629151,
+      "learning_rate": 3.4583019179114948e-06,
+      "loss": 0.5268,
+      "step": 11290
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.561886265766148,
+      "learning_rate": 3.453598115848795e-06,
+      "loss": 0.5308,
+      "step": 11295
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.515198041434377,
+      "learning_rate": 3.448895826672767e-06,
+      "loss": 0.5279,
+      "step": 11300
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.760420913784969,
+      "learning_rate": 3.444195054983788e-06,
+      "loss": 0.5204,
+      "step": 11305
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 20.619225924917643,
+      "learning_rate": 3.439495805380752e-06,
+      "loss": 0.5316,
+      "step": 11310
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.939695464936968,
+      "learning_rate": 3.4347980824610593e-06,
+      "loss": 0.5282,
+      "step": 11315
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.492607689047357,
+      "learning_rate": 3.4301018908206198e-06,
+      "loss": 0.5282,
+      "step": 11320
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.5726352155456,
+      "learning_rate": 3.4254072350538437e-06,
+      "loss": 0.5154,
+      "step": 11325
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.521845531151163,
+      "learning_rate": 3.420714119753641e-06,
+      "loss": 0.5391,
+      "step": 11330
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.394116452417119,
+      "learning_rate": 3.4160225495114134e-06,
+      "loss": 0.541,
+      "step": 11335
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.445030342660791,
+      "learning_rate": 3.4113325289170475e-06,
+      "loss": 0.5275,
+      "step": 11340
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.940504419644416,
+      "learning_rate": 3.4066440625589186e-06,
+      "loss": 0.5179,
+      "step": 11345
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.6345428352537965,
+      "learning_rate": 3.4019571550238816e-06,
+      "loss": 0.5142,
+      "step": 11350
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.461369428184476,
+      "learning_rate": 3.3972718108972612e-06,
+      "loss": 0.5224,
+      "step": 11355
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 29.354729919652407,
+      "learning_rate": 3.3925880347628577e-06,
+      "loss": 0.5264,
+      "step": 11360
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.403518167501197,
+      "learning_rate": 3.3879058312029354e-06,
+      "loss": 0.5282,
+      "step": 11365
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.988468122093689,
+      "learning_rate": 3.3832252047982206e-06,
+      "loss": 0.5311,
+      "step": 11370
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.384055884376355,
+      "learning_rate": 3.378546160127899e-06,
+      "loss": 0.5235,
+      "step": 11375
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.98755000776099,
+      "learning_rate": 3.3738687017696004e-06,
+      "loss": 0.5422,
+      "step": 11380
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.865728926427982,
+      "learning_rate": 3.3691928342994117e-06,
+      "loss": 0.5299,
+      "step": 11385
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.669535823097867,
+      "learning_rate": 3.364518562291861e-06,
+      "loss": 0.5216,
+      "step": 11390
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 20.042979966523028,
+      "learning_rate": 3.359845890319914e-06,
+      "loss": 0.5181,
+      "step": 11395
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.242460329586381,
+      "learning_rate": 3.3551748229549695e-06,
+      "loss": 0.5314,
+      "step": 11400
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.409967046664902,
+      "learning_rate": 3.3505053647668616e-06,
+      "loss": 0.5403,
+      "step": 11405
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 15.280372712285226,
+      "learning_rate": 3.3458375203238456e-06,
+      "loss": 0.5252,
+      "step": 11410
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 19.565929616107933,
+      "learning_rate": 3.3411712941926027e-06,
+      "loss": 0.5301,
+      "step": 11415
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.77112520924579,
+      "learning_rate": 3.3365066909382233e-06,
+      "loss": 0.5281,
+      "step": 11420
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.77064494485486,
+      "learning_rate": 3.331843715124216e-06,
+      "loss": 0.5188,
+      "step": 11425
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.3463404607000555,
+      "learning_rate": 3.3271823713124973e-06,
+      "loss": 0.5279,
+      "step": 11430
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.202088692534696,
+      "learning_rate": 3.3225226640633835e-06,
+      "loss": 0.5118,
+      "step": 11435
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.539992983116485,
+      "learning_rate": 3.317864597935595e-06,
+      "loss": 0.5225,
+      "step": 11440
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.612631868756232,
+      "learning_rate": 3.3132081774862403e-06,
+      "loss": 0.527,
+      "step": 11445
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.363407207718291,
+      "learning_rate": 3.308553407270822e-06,
+      "loss": 0.5433,
+      "step": 11450
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.753353395071274,
+      "learning_rate": 3.30390029184323e-06,
+      "loss": 0.5056,
+      "step": 11455
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.972212324066545,
+      "learning_rate": 3.299248835755728e-06,
+      "loss": 0.5234,
+      "step": 11460
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.2989976391151785,
+      "learning_rate": 3.2945990435589636e-06,
+      "loss": 0.5205,
+      "step": 11465
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.382665579425526,
+      "learning_rate": 3.289950919801954e-06,
+      "loss": 0.5344,
+      "step": 11470
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.614165948682854,
+      "learning_rate": 3.2853044690320836e-06,
+      "loss": 0.5335,
+      "step": 11475
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 18.54799855202085,
+      "learning_rate": 3.2806596957951003e-06,
+      "loss": 0.518,
+      "step": 11480
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.260869030494261,
+      "learning_rate": 3.2760166046351127e-06,
+      "loss": 0.5177,
+      "step": 11485
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.396344032296104,
+      "learning_rate": 3.2713752000945792e-06,
+      "loss": 0.5346,
+      "step": 11490
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 23.880420866969953,
+      "learning_rate": 3.266735486714314e-06,
+      "loss": 0.5279,
+      "step": 11495
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 22.70728514331903,
+      "learning_rate": 3.2620974690334723e-06,
+      "loss": 0.5239,
+      "step": 11500
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 16.493266956279886,
+      "learning_rate": 3.257461151589551e-06,
+      "loss": 0.5259,
+      "step": 11505
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 22.029387258063146,
+      "learning_rate": 3.2528265389183857e-06,
+      "loss": 0.5322,
+      "step": 11510
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 28.678050417980476,
+      "learning_rate": 3.2481936355541425e-06,
+      "loss": 0.5215,
+      "step": 11515
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.50147938948209,
+      "learning_rate": 3.2435624460293163e-06,
+      "loss": 0.5234,
+      "step": 11520
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 14.973628043093376,
+      "learning_rate": 3.2389329748747246e-06,
+      "loss": 0.5102,
+      "step": 11525
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.999565283031597,
+      "learning_rate": 3.2343052266195044e-06,
+      "loss": 0.5101,
+      "step": 11530
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.374175195674296,
+      "learning_rate": 3.2296792057911064e-06,
+      "loss": 0.5122,
+      "step": 11535
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.887472400804496,
+      "learning_rate": 3.22505491691529e-06,
+      "loss": 0.5198,
+      "step": 11540
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.888601482195552,
+      "learning_rate": 3.220432364516124e-06,
+      "loss": 0.5243,
+      "step": 11545
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.117889813946124,
+      "learning_rate": 3.215811553115974e-06,
+      "loss": 0.517,
+      "step": 11550
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.489191260544375,
+      "learning_rate": 3.2111924872355055e-06,
+      "loss": 0.5341,
+      "step": 11555
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.108146934501852,
+      "learning_rate": 3.2065751713936757e-06,
+      "loss": 0.5316,
+      "step": 11560
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.939846904829047,
+      "learning_rate": 3.2019596101077276e-06,
+      "loss": 0.5276,
+      "step": 11565
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.798886561943347,
+      "learning_rate": 3.197345807893191e-06,
+      "loss": 0.5185,
+      "step": 11570
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.562157928123866,
+      "learning_rate": 3.192733769263874e-06,
+      "loss": 0.5167,
+      "step": 11575
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.255025145423334,
+      "learning_rate": 3.1881234987318554e-06,
+      "loss": 0.5287,
+      "step": 11580
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.00519837851638,
+      "learning_rate": 3.183515000807488e-06,
+      "loss": 0.5323,
+      "step": 11585
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.33740503879071,
+      "learning_rate": 3.178908279999392e-06,
+      "loss": 0.5257,
+      "step": 11590
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.941774823430713,
+      "learning_rate": 3.174303340814443e-06,
+      "loss": 0.5405,
+      "step": 11595
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.764515099338116,
+      "learning_rate": 3.16970018775778e-06,
+      "loss": 0.5194,
+      "step": 11600
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.769381791545381,
+      "learning_rate": 3.1650988253327906e-06,
+      "loss": 0.5261,
+      "step": 11605
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.430163930611821,
+      "learning_rate": 3.160499258041112e-06,
+      "loss": 0.5204,
+      "step": 11610
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.8678313815022545,
+      "learning_rate": 3.1559014903826245e-06,
+      "loss": 0.5267,
+      "step": 11615
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.961306154968712,
+      "learning_rate": 3.1513055268554518e-06,
+      "loss": 0.5123,
+      "step": 11620
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.386787947968249,
+      "learning_rate": 3.146711371955943e-06,
+      "loss": 0.5293,
+      "step": 11625
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.380022599514302,
+      "learning_rate": 3.142119030178688e-06,
+      "loss": 0.519,
+      "step": 11630
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.601956854275004,
+      "learning_rate": 3.1375285060164963e-06,
+      "loss": 0.5218,
+      "step": 11635
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.956209340415413,
+      "learning_rate": 3.132939803960402e-06,
+      "loss": 0.5152,
+      "step": 11640
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.5091717789471,
+      "learning_rate": 3.128352928499657e-06,
+      "loss": 0.5064,
+      "step": 11645
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 27.95353269895987,
+      "learning_rate": 3.123767884121725e-06,
+      "loss": 0.5348,
+      "step": 11650
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.97543978293298,
+      "learning_rate": 3.1191846753122783e-06,
+      "loss": 0.5094,
+      "step": 11655
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.329302870828528,
+      "learning_rate": 3.1146033065551964e-06,
+      "loss": 0.5126,
+      "step": 11660
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 26.946136861360525,
+      "learning_rate": 3.11002378233255e-06,
+      "loss": 0.5222,
+      "step": 11665
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 26.85683438822377,
+      "learning_rate": 3.1054461071246155e-06,
+      "loss": 0.5176,
+      "step": 11670
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 32.54303271050792,
+      "learning_rate": 3.100870285409856e-06,
+      "loss": 0.5285,
+      "step": 11675
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 22.70343510365221,
+      "learning_rate": 3.0962963216649196e-06,
+      "loss": 0.5281,
+      "step": 11680
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.917301804129767,
+      "learning_rate": 3.0917242203646385e-06,
+      "loss": 0.5043,
+      "step": 11685
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.110336385901124,
+      "learning_rate": 3.087153985982024e-06,
+      "loss": 0.5175,
+      "step": 11690
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.677973496169243,
+      "learning_rate": 3.0825856229882584e-06,
+      "loss": 0.5124,
+      "step": 11695
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.415213598801499,
+      "learning_rate": 3.078019135852698e-06,
+      "loss": 0.5211,
+      "step": 11700
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 16.720758651500194,
+      "learning_rate": 3.073454529042854e-06,
+      "loss": 0.5154,
+      "step": 11705
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.50117824566899,
+      "learning_rate": 3.068891807024409e-06,
+      "loss": 0.52,
+      "step": 11710
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.441597888328433,
+      "learning_rate": 3.064330974261196e-06,
+      "loss": 0.5222,
+      "step": 11715
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.022956215814487,
+      "learning_rate": 3.0597720352152004e-06,
+      "loss": 0.5113,
+      "step": 11720
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.45410754613074,
+      "learning_rate": 3.0552149943465554e-06,
+      "loss": 0.5313,
+      "step": 11725
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.656188542011064,
+      "learning_rate": 3.0506598561135362e-06,
+      "loss": 0.5154,
+      "step": 11730
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.627915183745318,
+      "learning_rate": 3.0461066249725584e-06,
+      "loss": 0.5188,
+      "step": 11735
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.616229006704577,
+      "learning_rate": 3.0415553053781725e-06,
+      "loss": 0.5208,
+      "step": 11740
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.0943460014009,
+      "learning_rate": 3.037005901783053e-06,
+      "loss": 0.5194,
+      "step": 11745
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.190068268493544,
+      "learning_rate": 3.032458418638008e-06,
+      "loss": 0.5208,
+      "step": 11750
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.1286362517199136,
+      "learning_rate": 3.0279128603919593e-06,
+      "loss": 0.5125,
+      "step": 11755
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.220220298533826,
+      "learning_rate": 3.0233692314919525e-06,
+      "loss": 0.5235,
+      "step": 11760
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.3905054032802582,
+      "eval_runtime": 75.0128,
+      "eval_samples_per_second": 4.826,
+      "eval_steps_per_second": 0.613,
+      "step": 11764
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.611507507714085,
+      "learning_rate": 3.018827536383142e-06,
+      "loss": 0.4829,
+      "step": 11765
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.752742447036074,
+      "learning_rate": 3.0142877795087876e-06,
+      "loss": 0.3723,
+      "step": 11770
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.804207868987879,
+      "learning_rate": 3.009749965310259e-06,
+      "loss": 0.3751,
+      "step": 11775
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.48577340760751,
+      "learning_rate": 3.0052140982270228e-06,
+      "loss": 0.3786,
+      "step": 11780
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.96000499515076,
+      "learning_rate": 3.0006801826966366e-06,
+      "loss": 0.3715,
+      "step": 11785
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.683037451115184,
+      "learning_rate": 2.9961482231547535e-06,
+      "loss": 0.3728,
+      "step": 11790
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.947056755410368,
+      "learning_rate": 2.9916182240351132e-06,
+      "loss": 0.3658,
+      "step": 11795
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 9.175381671925878,
+      "learning_rate": 2.987090189769535e-06,
+      "loss": 0.3672,
+      "step": 11800
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 7.202043240807192,
+      "learning_rate": 2.9825641247879167e-06,
+      "loss": 0.3792,
+      "step": 11805
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.919930890072514,
+      "learning_rate": 2.9780400335182312e-06,
+      "loss": 0.357,
+      "step": 11810
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.458484845692553,
+      "learning_rate": 2.973517920386517e-06,
+      "loss": 0.3642,
+      "step": 11815
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 12.605985113201717,
+      "learning_rate": 2.968997789816882e-06,
+      "loss": 0.366,
+      "step": 11820
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.17730548605672,
+      "learning_rate": 2.9644796462314897e-06,
+      "loss": 0.3685,
+      "step": 11825
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.988272278395259,
+      "learning_rate": 2.959963494050562e-06,
+      "loss": 0.3653,
+      "step": 11830
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.730790106634236,
+      "learning_rate": 2.955449337692372e-06,
+      "loss": 0.3605,
+      "step": 11835
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.492211689650247,
+      "learning_rate": 2.9509371815732415e-06,
+      "loss": 0.3645,
+      "step": 11840
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 7.540976427148941,
+      "learning_rate": 2.946427030107534e-06,
+      "loss": 0.3642,
+      "step": 11845
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.579453176000582,
+      "learning_rate": 2.9419188877076534e-06,
+      "loss": 0.3693,
+      "step": 11850
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.304416580656278,
+      "learning_rate": 2.9374127587840373e-06,
+      "loss": 0.3522,
+      "step": 11855
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.3593333215489,
+      "learning_rate": 2.932908647745152e-06,
+      "loss": 0.367,
+      "step": 11860
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.497643274627988,
+      "learning_rate": 2.9284065589974915e-06,
+      "loss": 0.3633,
+      "step": 11865
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 13.575720144681346,
+      "learning_rate": 2.9239064969455686e-06,
+      "loss": 0.3656,
+      "step": 11870
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.418062852100978,
+      "learning_rate": 2.9194084659919176e-06,
+      "loss": 0.3624,
+      "step": 11875
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.713970180416974,
+      "learning_rate": 2.914912470537081e-06,
+      "loss": 0.3605,
+      "step": 11880
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 6.3675790522641655,
+      "learning_rate": 2.9104185149796166e-06,
+      "loss": 0.3577,
+      "step": 11885
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.380851351218553,
+      "learning_rate": 2.9059266037160804e-06,
+      "loss": 0.3642,
+      "step": 11890
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.881021953784742,
+      "learning_rate": 2.9014367411410272e-06,
+      "loss": 0.3689,
+      "step": 11895
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.224788190005951,
+      "learning_rate": 2.896948931647018e-06,
+      "loss": 0.3706,
+      "step": 11900
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 7.267290416239944,
+      "learning_rate": 2.8924631796245896e-06,
+      "loss": 0.3626,
+      "step": 11905
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.722158276142785,
+      "learning_rate": 2.8879794894622794e-06,
+      "loss": 0.3549,
+      "step": 11910
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.665683638849933,
+      "learning_rate": 2.883497865546599e-06,
+      "loss": 0.3582,
+      "step": 11915
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 11.717660943671277,
+      "learning_rate": 2.8790183122620455e-06,
+      "loss": 0.3557,
+      "step": 11920
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.744844839312123,
+      "learning_rate": 2.8745408339910857e-06,
+      "loss": 0.3629,
+      "step": 11925
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.254584277873432,
+      "learning_rate": 2.8700654351141546e-06,
+      "loss": 0.3626,
+      "step": 11930
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.076849011568416,
+      "learning_rate": 2.865592120009659e-06,
+      "loss": 0.3665,
+      "step": 11935
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 9.89105922641013,
+      "learning_rate": 2.8611208930539635e-06,
+      "loss": 0.3748,
+      "step": 11940
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.278339030031562,
+      "learning_rate": 2.8566517586213895e-06,
+      "loss": 0.3752,
+      "step": 11945
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.055266207685965,
+      "learning_rate": 2.852184721084208e-06,
+      "loss": 0.3645,
+      "step": 11950
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 7.148044998103296,
+      "learning_rate": 2.847719784812648e-06,
+      "loss": 0.3722,
+      "step": 11955
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.54289037896895,
+      "learning_rate": 2.8432569541748728e-06,
+      "loss": 0.3609,
+      "step": 11960
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.931371433064498,
+      "learning_rate": 2.8387962335369935e-06,
+      "loss": 0.3658,
+      "step": 11965
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.930370778361347,
+      "learning_rate": 2.8343376272630524e-06,
+      "loss": 0.3487,
+      "step": 11970
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.26844641650417,
+      "learning_rate": 2.8298811397150217e-06,
+      "loss": 0.3588,
+      "step": 11975
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.382054822171709,
+      "learning_rate": 2.825426775252806e-06,
+      "loss": 0.3578,
+      "step": 11980
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.611521064546941,
+      "learning_rate": 2.8209745382342312e-06,
+      "loss": 0.3558,
+      "step": 11985
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.068563602349961,
+      "learning_rate": 2.8165244330150383e-06,
+      "loss": 0.364,
+      "step": 11990
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.357350781394905,
+      "learning_rate": 2.812076463948884e-06,
+      "loss": 0.3672,
+      "step": 11995
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.591605468812917,
+      "learning_rate": 2.8076306353873413e-06,
+      "loss": 0.3651,
+      "step": 12000
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.106233634138435,
+      "learning_rate": 2.8031869516798794e-06,
+      "loss": 0.3523,
+      "step": 12005
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.718458446762907,
+      "learning_rate": 2.798745417173877e-06,
+      "loss": 0.3632,
+      "step": 12010
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.2380488137176675,
+      "learning_rate": 2.7943060362146068e-06,
+      "loss": 0.357,
+      "step": 12015
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.238093997370028,
+      "learning_rate": 2.7898688131452344e-06,
+      "loss": 0.3616,
+      "step": 12020
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.566161035481949,
+      "learning_rate": 2.785433752306812e-06,
+      "loss": 0.3568,
+      "step": 12025
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.614923482223124,
+      "learning_rate": 2.7810008580382843e-06,
+      "loss": 0.358,
+      "step": 12030
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.801101666626508,
+      "learning_rate": 2.776570134676469e-06,
+      "loss": 0.3627,
+      "step": 12035
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.401932712886612,
+      "learning_rate": 2.77214158655606e-06,
+      "loss": 0.3577,
+      "step": 12040
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.629093893600516,
+      "learning_rate": 2.76771521800963e-06,
+      "loss": 0.3602,
+      "step": 12045
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.84615246777452,
+      "learning_rate": 2.76329103336761e-06,
+      "loss": 0.3629,
+      "step": 12050
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 6.493679745481419,
+      "learning_rate": 2.7588690369583025e-06,
+      "loss": 0.3584,
+      "step": 12055
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.650688020597786,
+      "learning_rate": 2.7544492331078667e-06,
+      "loss": 0.3573,
+      "step": 12060
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.875759976969528,
+      "learning_rate": 2.750031626140313e-06,
+      "loss": 0.3633,
+      "step": 12065
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.228597460057165,
+      "learning_rate": 2.745616220377504e-06,
+      "loss": 0.3627,
+      "step": 12070
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.359671405062235,
+      "learning_rate": 2.7412030201391553e-06,
+      "loss": 0.3614,
+      "step": 12075
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 6.22622924385044,
+      "learning_rate": 2.7367920297428174e-06,
+      "loss": 0.3591,
+      "step": 12080
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.388816381649836,
+      "learning_rate": 2.7323832535038787e-06,
+      "loss": 0.3547,
+      "step": 12085
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.70592439616937,
+      "learning_rate": 2.727976695735568e-06,
+      "loss": 0.3613,
+      "step": 12090
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.429611413044588,
+      "learning_rate": 2.7235723607489357e-06,
+      "loss": 0.3639,
+      "step": 12095
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.446747254918879,
+      "learning_rate": 2.719170252852868e-06,
+      "loss": 0.3565,
+      "step": 12100
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.694603382556856,
+      "learning_rate": 2.7147703763540567e-06,
+      "loss": 0.3575,
+      "step": 12105
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 10.344482002591908,
+      "learning_rate": 2.7103727355570277e-06,
+      "loss": 0.3599,
+      "step": 12110
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.230043253110429,
+      "learning_rate": 2.7059773347641048e-06,
+      "loss": 0.3783,
+      "step": 12115
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.911178835416952,
+      "learning_rate": 2.701584178275433e-06,
+      "loss": 0.3507,
+      "step": 12120
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.060856281038126,
+      "learning_rate": 2.6971932703889534e-06,
+      "loss": 0.3577,
+      "step": 12125
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.095825633746809,
+      "learning_rate": 2.6928046154004083e-06,
+      "loss": 0.348,
+      "step": 12130
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.33022918116732,
+      "learning_rate": 2.6884182176033397e-06,
+      "loss": 0.3597,
+      "step": 12135
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.14854869257719,
+      "learning_rate": 2.684034081289078e-06,
+      "loss": 0.3654,
+      "step": 12140
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.175699068398426,
+      "learning_rate": 2.6796522107467417e-06,
+      "loss": 0.3471,
+      "step": 12145
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 11.87073571076359,
+      "learning_rate": 2.6752726102632307e-06,
+      "loss": 0.3646,
+      "step": 12150
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.265772381937061,
+      "learning_rate": 2.670895284123231e-06,
+      "loss": 0.3583,
+      "step": 12155
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.51482295186804,
+      "learning_rate": 2.666520236609196e-06,
+      "loss": 0.3474,
+      "step": 12160
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.52752360844317,
+      "learning_rate": 2.662147472001352e-06,
+      "loss": 0.3687,
+      "step": 12165
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.657318428452209,
+      "learning_rate": 2.6577769945776942e-06,
+      "loss": 0.3708,
+      "step": 12170
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 14.585897856994992,
+      "learning_rate": 2.653408808613977e-06,
+      "loss": 0.3748,
+      "step": 12175
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 12.455744504231912,
+      "learning_rate": 2.6490429183837195e-06,
+      "loss": 0.3574,
+      "step": 12180
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.3858811970077705,
+      "learning_rate": 2.6446793281581815e-06,
+      "loss": 0.356,
+      "step": 12185
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 13.118397592546135,
+      "learning_rate": 2.640318042206387e-06,
+      "loss": 0.3654,
+      "step": 12190
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.898867934283263,
+      "learning_rate": 2.635959064795097e-06,
+      "loss": 0.3513,
+      "step": 12195
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.454220363751431,
+      "learning_rate": 2.6316024001888195e-06,
+      "loss": 0.3684,
+      "step": 12200
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.251292384038412,
+      "learning_rate": 2.6272480526497952e-06,
+      "loss": 0.3517,
+      "step": 12205
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.71362624669713,
+      "learning_rate": 2.622896026437998e-06,
+      "loss": 0.3635,
+      "step": 12210
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.455864586011419,
+      "learning_rate": 2.6185463258111355e-06,
+      "loss": 0.3568,
+      "step": 12215
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.345032911435628,
+      "learning_rate": 2.614198955024637e-06,
+      "loss": 0.3621,
+      "step": 12220
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.085325440928504,
+      "learning_rate": 2.6098539183316508e-06,
+      "loss": 0.3572,
+      "step": 12225
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.755306263476994,
+      "learning_rate": 2.6055112199830423e-06,
+      "loss": 0.3632,
+      "step": 12230
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.37675672449101,
+      "learning_rate": 2.601170864227395e-06,
+      "loss": 0.362,
+      "step": 12235
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 10.004487280961618,
+      "learning_rate": 2.5968328553109912e-06,
+      "loss": 0.3607,
+      "step": 12240
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 11.383232716403265,
+      "learning_rate": 2.5924971974778257e-06,
+      "loss": 0.3695,
+      "step": 12245
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.952935057045326,
+      "learning_rate": 2.5881638949695886e-06,
+      "loss": 0.3528,
+      "step": 12250
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.056211499559283,
+      "learning_rate": 2.5838329520256645e-06,
+      "loss": 0.3605,
+      "step": 12255
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.277901307181128,
+      "learning_rate": 2.579504372883134e-06,
+      "loss": 0.3646,
+      "step": 12260
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.69486470616986,
+      "learning_rate": 2.575178161776763e-06,
+      "loss": 0.3647,
+      "step": 12265
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.170144636951997,
+      "learning_rate": 2.5708543229389995e-06,
+      "loss": 0.3556,
+      "step": 12270
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.872922231980344,
+      "learning_rate": 2.5665328605999696e-06,
+      "loss": 0.3726,
+      "step": 12275
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 7.288978068586903,
+      "learning_rate": 2.5622137789874803e-06,
+      "loss": 0.3619,
+      "step": 12280
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.131810848831009,
+      "learning_rate": 2.557897082327002e-06,
+      "loss": 0.3658,
+      "step": 12285
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.919657846426334,
+      "learning_rate": 2.5535827748416797e-06,
+      "loss": 0.3698,
+      "step": 12290
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.168193650629757,
+      "learning_rate": 2.5492708607523144e-06,
+      "loss": 0.3661,
+      "step": 12295
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.285402235676456,
+      "learning_rate": 2.544961344277368e-06,
+      "loss": 0.3745,
+      "step": 12300
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.514162836792714,
+      "learning_rate": 2.540654229632955e-06,
+      "loss": 0.3721,
+      "step": 12305
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.014041099796996,
+      "learning_rate": 2.536349521032846e-06,
+      "loss": 0.3486,
+      "step": 12310
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.404746919512409,
+      "learning_rate": 2.5320472226884506e-06,
+      "loss": 0.3553,
+      "step": 12315
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.926602435491768,
+      "learning_rate": 2.527747338808822e-06,
+      "loss": 0.3705,
+      "step": 12320
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.43243287500068,
+      "learning_rate": 2.5234498736006563e-06,
+      "loss": 0.3758,
+      "step": 12325
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.5893380700500135,
+      "learning_rate": 2.5191548312682758e-06,
+      "loss": 0.3685,
+      "step": 12330
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.739196866534467,
+      "learning_rate": 2.5148622160136406e-06,
+      "loss": 0.3644,
+      "step": 12335
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.223888941673534,
+      "learning_rate": 2.5105720320363287e-06,
+      "loss": 0.3566,
+      "step": 12340
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.760901049999837,
+      "learning_rate": 2.5062842835335442e-06,
+      "loss": 0.3517,
+      "step": 12345
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 8.717510780154683,
+      "learning_rate": 2.5019989747001043e-06,
+      "loss": 0.3576,
+      "step": 12350
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 9.274190474502552,
+      "learning_rate": 2.4977161097284468e-06,
+      "loss": 0.3555,
+      "step": 12355
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.661456821603926,
+      "learning_rate": 2.4934356928086124e-06,
+      "loss": 0.3526,
+      "step": 12360
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 13.14206177144619,
+      "learning_rate": 2.489157728128245e-06,
+      "loss": 0.3536,
+      "step": 12365
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.8359452582871425,
+      "learning_rate": 2.4848822198725974e-06,
+      "loss": 0.3545,
+      "step": 12370
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.562661402570011,
+      "learning_rate": 2.4806091722245114e-06,
+      "loss": 0.3557,
+      "step": 12375
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.216730692799468,
+      "learning_rate": 2.4763385893644283e-06,
+      "loss": 0.3527,
+      "step": 12380
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.175993021978491,
+      "learning_rate": 2.4720704754703715e-06,
+      "loss": 0.3705,
+      "step": 12385
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.558130455617105,
+      "learning_rate": 2.467804834717954e-06,
+      "loss": 0.3611,
+      "step": 12390
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 12.155852876216986,
+      "learning_rate": 2.4635416712803635e-06,
+      "loss": 0.3584,
+      "step": 12395
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.056660813979677,
+      "learning_rate": 2.4592809893283725e-06,
+      "loss": 0.3523,
+      "step": 12400
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 12.155316694768354,
+      "learning_rate": 2.455022793030319e-06,
+      "loss": 0.351,
+      "step": 12405
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.137254627367353,
+      "learning_rate": 2.4507670865521093e-06,
+      "loss": 0.3555,
+      "step": 12410
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.335378484644814,
+      "learning_rate": 2.446513874057219e-06,
+      "loss": 0.3475,
+      "step": 12415
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.202345783395633,
+      "learning_rate": 2.442263159706678e-06,
+      "loss": 0.3627,
+      "step": 12420
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.040729939592325,
+      "learning_rate": 2.4380149476590805e-06,
+      "loss": 0.362,
+      "step": 12425
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 14.663934834177535,
+      "learning_rate": 2.4337692420705578e-06,
+      "loss": 0.352,
+      "step": 12430
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.6804959838619205,
+      "learning_rate": 2.4295260470948058e-06,
+      "loss": 0.3551,
+      "step": 12435
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 8.268339570358487,
+      "learning_rate": 2.425285366883053e-06,
+      "loss": 0.3601,
+      "step": 12440
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.822741932012296,
+      "learning_rate": 2.42104720558407e-06,
+      "loss": 0.3484,
+      "step": 12445
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.813642685372308,
+      "learning_rate": 2.416811567344169e-06,
+      "loss": 0.3679,
+      "step": 12450
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 10.837042565036516,
+      "learning_rate": 2.4125784563071843e-06,
+      "loss": 0.3476,
+      "step": 12455
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.782218813101773,
+      "learning_rate": 2.4083478766144863e-06,
+      "loss": 0.3606,
+      "step": 12460
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 11.059837450405126,
+      "learning_rate": 2.4041198324049634e-06,
+      "loss": 0.3672,
+      "step": 12465
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.250155523400553,
+      "learning_rate": 2.3998943278150265e-06,
+      "loss": 0.3653,
+      "step": 12470
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.595823286695708,
+      "learning_rate": 2.3956713669785974e-06,
+      "loss": 0.3572,
+      "step": 12475
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.43108384835417,
+      "learning_rate": 2.391450954027117e-06,
+      "loss": 0.3578,
+      "step": 12480
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.888909316789247,
+      "learning_rate": 2.387233093089527e-06,
+      "loss": 0.3716,
+      "step": 12485
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.484941394571873,
+      "learning_rate": 2.383017788292273e-06,
+      "loss": 0.3568,
+      "step": 12490
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.746259770066021,
+      "learning_rate": 2.3788050437593042e-06,
+      "loss": 0.3651,
+      "step": 12495
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.9129131917140505,
+      "learning_rate": 2.37459486361206e-06,
+      "loss": 0.3515,
+      "step": 12500
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 6.071685633633804,
+      "learning_rate": 2.370387251969477e-06,
+      "loss": 0.3692,
+      "step": 12505
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.804804003375496,
+      "learning_rate": 2.366182212947969e-06,
+      "loss": 0.3534,
+      "step": 12510
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.700029141854401,
+      "learning_rate": 2.3619797506614447e-06,
+      "loss": 0.3575,
+      "step": 12515
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.871791337356873,
+      "learning_rate": 2.3577798692212817e-06,
+      "loss": 0.3479,
+      "step": 12520
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.445874880209733,
+      "learning_rate": 2.353582572736342e-06,
+      "loss": 0.3593,
+      "step": 12525
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 6.388800020380725,
+      "learning_rate": 2.349387865312951e-06,
+      "loss": 0.3607,
+      "step": 12530
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 7.061201502997077,
+      "learning_rate": 2.3451957510549034e-06,
+      "loss": 0.3547,
+      "step": 12535
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.40790926491953,
+      "learning_rate": 2.34100623406346e-06,
+      "loss": 0.3569,
+      "step": 12540
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 6.112629901368969,
+      "learning_rate": 2.336819318437338e-06,
+      "loss": 0.3571,
+      "step": 12545
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.5893900154448986,
+      "learning_rate": 2.3326350082727093e-06,
+      "loss": 0.3549,
+      "step": 12550
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.773355334844262,
+      "learning_rate": 2.3284533076631954e-06,
+      "loss": 0.3628,
+      "step": 12555
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.176848576237044,
+      "learning_rate": 2.3242742206998703e-06,
+      "loss": 0.3525,
+      "step": 12560
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.064208261078508,
+      "learning_rate": 2.3200977514712434e-06,
+      "loss": 0.3646,
+      "step": 12565
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 8.572635391436043,
+      "learning_rate": 2.3159239040632725e-06,
+      "loss": 0.3496,
+      "step": 12570
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.203301780474356,
+      "learning_rate": 2.3117526825593417e-06,
+      "loss": 0.3509,
+      "step": 12575
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.019304670820222,
+      "learning_rate": 2.307584091040268e-06,
+      "loss": 0.3532,
+      "step": 12580
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.743250315644167,
+      "learning_rate": 2.303418133584301e-06,
+      "loss": 0.3549,
+      "step": 12585
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.598745561581672,
+      "learning_rate": 2.299254814267107e-06,
+      "loss": 0.3494,
+      "step": 12590
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 10.03454149898803,
+      "learning_rate": 2.295094137161774e-06,
+      "loss": 0.3588,
+      "step": 12595
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.63513008315468,
+      "learning_rate": 2.2909361063388024e-06,
+      "loss": 0.3514,
+      "step": 12600
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 12.429596505739113,
+      "learning_rate": 2.28678072586611e-06,
+      "loss": 0.3504,
+      "step": 12605
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.747186379508115,
+      "learning_rate": 2.282627999809014e-06,
+      "loss": 0.3517,
+      "step": 12610
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.766026429105853,
+      "learning_rate": 2.2784779322302408e-06,
+      "loss": 0.3515,
+      "step": 12615
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.19796998489907,
+      "learning_rate": 2.274330527189913e-06,
+      "loss": 0.3563,
+      "step": 12620
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.687175841710557,
+      "learning_rate": 2.2701857887455482e-06,
+      "loss": 0.3553,
+      "step": 12625
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.492044808621296,
+      "learning_rate": 2.266043720952054e-06,
+      "loss": 0.346,
+      "step": 12630
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.698079597137065,
+      "learning_rate": 2.2619043278617307e-06,
+      "loss": 0.355,
+      "step": 12635
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.647526337896694,
+      "learning_rate": 2.2577676135242566e-06,
+      "loss": 0.356,
+      "step": 12640
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.7027116634400175,
+      "learning_rate": 2.253633581986689e-06,
+      "loss": 0.3624,
+      "step": 12645
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.394766540445826,
+      "learning_rate": 2.249502237293466e-06,
+      "loss": 0.3621,
+      "step": 12650
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.1860900535938,
+      "learning_rate": 2.2453735834863897e-06,
+      "loss": 0.3479,
+      "step": 12655
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.678739729089573,
+      "learning_rate": 2.2412476246046377e-06,
+      "loss": 0.3508,
+      "step": 12660
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 7.542513089344192,
+      "learning_rate": 2.2371243646847444e-06,
+      "loss": 0.3579,
+      "step": 12665
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.907712371792205,
+      "learning_rate": 2.233003807760607e-06,
+      "loss": 0.3553,
+      "step": 12670
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 7.918275784958399,
+      "learning_rate": 2.228885957863477e-06,
+      "loss": 0.3488,
+      "step": 12675
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.0345251496741605,
+      "learning_rate": 2.224770819021956e-06,
+      "loss": 0.3516,
+      "step": 12680
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.360579403001685,
+      "learning_rate": 2.220658395261999e-06,
+      "loss": 0.3505,
+      "step": 12685
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.071208825906468,
+      "learning_rate": 2.216548690606898e-06,
+      "loss": 0.3503,
+      "step": 12690
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 7.3340027233657255,
+      "learning_rate": 2.2124417090772903e-06,
+      "loss": 0.3498,
+      "step": 12695
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.058112027557419,
+      "learning_rate": 2.2083374546911444e-06,
+      "loss": 0.364,
+      "step": 12700
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.4348235788957675,
+      "learning_rate": 2.2042359314637683e-06,
+      "loss": 0.3618,
+      "step": 12705
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 9.226367642118475,
+      "learning_rate": 2.200137143407785e-06,
+      "loss": 0.3644,
+      "step": 12710
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.612977427099218,
+      "learning_rate": 2.196041094533155e-06,
+      "loss": 0.3588,
+      "step": 12715
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.7396878974458385,
+      "learning_rate": 2.1919477888471522e-06,
+      "loss": 0.3558,
+      "step": 12720
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.955526064834778,
+      "learning_rate": 2.1878572303543655e-06,
+      "loss": 0.3601,
+      "step": 12725
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.447660951252212,
+      "learning_rate": 2.183769423056702e-06,
+      "loss": 0.3562,
+      "step": 12730
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 7.579752616290158,
+      "learning_rate": 2.179684370953371e-06,
+      "loss": 0.3577,
+      "step": 12735
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.6468210027675445,
+      "learning_rate": 2.175602078040892e-06,
+      "loss": 0.352,
+      "step": 12740
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.578563303804438,
+      "learning_rate": 2.1715225483130815e-06,
+      "loss": 0.3456,
+      "step": 12745
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.508536921531665,
+      "learning_rate": 2.167445785761052e-06,
+      "loss": 0.358,
+      "step": 12750
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.501753862781778,
+      "learning_rate": 2.1633717943732098e-06,
+      "loss": 0.3413,
+      "step": 12755
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.196928032996506,
+      "learning_rate": 2.159300578135254e-06,
+      "loss": 0.3581,
+      "step": 12760
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.074066358859239,
+      "learning_rate": 2.1552321410301626e-06,
+      "loss": 0.3529,
+      "step": 12765
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.809753435779445,
+      "learning_rate": 2.1511664870381956e-06,
+      "loss": 0.3591,
+      "step": 12770
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.725861909837628,
+      "learning_rate": 2.1471036201368968e-06,
+      "loss": 0.352,
+      "step": 12775
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.314575776197405,
+      "learning_rate": 2.1430435443010733e-06,
+      "loss": 0.3479,
+      "step": 12780
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.603148181328639,
+      "learning_rate": 2.1389862635028136e-06,
+      "loss": 0.352,
+      "step": 12785
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.357563224881423,
+      "learning_rate": 2.134931781711457e-06,
+      "loss": 0.3524,
+      "step": 12790
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.679609149108052,
+      "learning_rate": 2.130880102893618e-06,
+      "loss": 0.3547,
+      "step": 12795
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.653289583363171,
+      "learning_rate": 2.126831231013159e-06,
+      "loss": 0.3433,
+      "step": 12800
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.150465354689677,
+      "learning_rate": 2.122785170031205e-06,
+      "loss": 0.3553,
+      "step": 12805
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.49152039269724,
+      "learning_rate": 2.118741923906125e-06,
+      "loss": 0.3545,
+      "step": 12810
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.968638433195927,
+      "learning_rate": 2.1147014965935327e-06,
+      "loss": 0.3511,
+      "step": 12815
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.476562767724293,
+      "learning_rate": 2.110663892046292e-06,
+      "loss": 0.3475,
+      "step": 12820
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.193980384967873,
+      "learning_rate": 2.1066291142144978e-06,
+      "loss": 0.3561,
+      "step": 12825
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 20.28983778856591,
+      "learning_rate": 2.1025971670454827e-06,
+      "loss": 0.3646,
+      "step": 12830
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.287610844059453,
+      "learning_rate": 2.098568054483807e-06,
+      "loss": 0.3655,
+      "step": 12835
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 21.597390203424133,
+      "learning_rate": 2.094541780471264e-06,
+      "loss": 0.3608,
+      "step": 12840
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 13.785556614757647,
+      "learning_rate": 2.0905183489468623e-06,
+      "loss": 0.3641,
+      "step": 12845
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 7.368036043761198,
+      "learning_rate": 2.0864977638468376e-06,
+      "loss": 0.3596,
+      "step": 12850
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.858919450916035,
+      "learning_rate": 2.0824800291046347e-06,
+      "loss": 0.371,
+      "step": 12855
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.328796201165607,
+      "learning_rate": 2.07846514865091e-06,
+      "loss": 0.3434,
+      "step": 12860
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.384829402670883,
+      "learning_rate": 2.0744531264135327e-06,
+      "loss": 0.3554,
+      "step": 12865
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.8962393661322,
+      "learning_rate": 2.0704439663175714e-06,
+      "loss": 0.3517,
+      "step": 12870
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.3887445041630295,
+      "learning_rate": 2.0664376722852948e-06,
+      "loss": 0.3555,
+      "step": 12875
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 19.531962902066855,
+      "learning_rate": 2.0624342482361664e-06,
+      "loss": 0.3501,
+      "step": 12880
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.102304932018561,
+      "learning_rate": 2.058433698086848e-06,
+      "loss": 0.3382,
+      "step": 12885
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.280312009210922,
+      "learning_rate": 2.0544360257511826e-06,
+      "loss": 0.3578,
+      "step": 12890
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.072495666599368,
+      "learning_rate": 2.050441235140203e-06,
+      "loss": 0.3507,
+      "step": 12895
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.7319798260058095,
+      "learning_rate": 2.046449330162121e-06,
+      "loss": 0.3427,
+      "step": 12900
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.997556168618728,
+      "learning_rate": 2.0424603147223228e-06,
+      "loss": 0.3487,
+      "step": 12905
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.668830722747984,
+      "learning_rate": 2.0384741927233687e-06,
+      "loss": 0.3471,
+      "step": 12910
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.498365158807844,
+      "learning_rate": 2.0344909680649937e-06,
+      "loss": 0.3444,
+      "step": 12915
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.035491184131442,
+      "learning_rate": 2.030510644644091e-06,
+      "loss": 0.3531,
+      "step": 12920
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.3573911283481035,
+      "learning_rate": 2.0265332263547175e-06,
+      "loss": 0.3546,
+      "step": 12925
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 12.465513201544523,
+      "learning_rate": 2.022558717088092e-06,
+      "loss": 0.351,
+      "step": 12930
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.656423916603554,
+      "learning_rate": 2.01858712073258e-06,
+      "loss": 0.3478,
+      "step": 12935
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.118832543339823,
+      "learning_rate": 2.0146184411737057e-06,
+      "loss": 0.3521,
+      "step": 12940
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.423815170561722,
+      "learning_rate": 2.0106526822941336e-06,
+      "loss": 0.3537,
+      "step": 12945
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.497609982111329,
+      "learning_rate": 2.006689847973672e-06,
+      "loss": 0.362,
+      "step": 12950
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 8.724143543231287,
+      "learning_rate": 2.0027299420892687e-06,
+      "loss": 0.3605,
+      "step": 12955
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.545635461939133,
+      "learning_rate": 1.9987729685150054e-06,
+      "loss": 0.3477,
+      "step": 12960
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 8.329174384014586,
+      "learning_rate": 1.994818931122099e-06,
+      "loss": 0.353,
+      "step": 12965
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.631646120301006,
+      "learning_rate": 1.9908678337788866e-06,
+      "loss": 0.3623,
+      "step": 12970
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.587745944032032,
+      "learning_rate": 1.9869196803508383e-06,
+      "loss": 0.3512,
+      "step": 12975
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.614550001666569,
+      "learning_rate": 1.9829744747005355e-06,
+      "loss": 0.3486,
+      "step": 12980
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 6.188568299003901,
+      "learning_rate": 1.979032220687683e-06,
+      "loss": 0.3576,
+      "step": 12985
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.165804827286981,
+      "learning_rate": 1.975092922169089e-06,
+      "loss": 0.3532,
+      "step": 12990
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.097043978976538,
+      "learning_rate": 1.9711565829986795e-06,
+      "loss": 0.3489,
+      "step": 12995
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.939939440373387,
+      "learning_rate": 1.9672232070274803e-06,
+      "loss": 0.3582,
+      "step": 13000
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.601542152733169,
+      "learning_rate": 1.963292798103617e-06,
+      "loss": 0.3549,
+      "step": 13005
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 10.330609981982413,
+      "learning_rate": 1.9593653600723184e-06,
+      "loss": 0.3459,
+      "step": 13010
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.109084999537791,
+      "learning_rate": 1.9554408967758996e-06,
+      "loss": 0.3554,
+      "step": 13015
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.586639449236359,
+      "learning_rate": 1.951519412053772e-06,
+      "loss": 0.3496,
+      "step": 13020
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 8.917383245650884,
+      "learning_rate": 1.947600909742427e-06,
+      "loss": 0.3497,
+      "step": 13025
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.944441252305898,
+      "learning_rate": 1.9436853936754456e-06,
+      "loss": 0.3571,
+      "step": 13030
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 15.326786665243585,
+      "learning_rate": 1.9397728676834772e-06,
+      "loss": 0.3525,
+      "step": 13035
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.722319121428556,
+      "learning_rate": 1.9358633355942547e-06,
+      "loss": 0.3523,
+      "step": 13040
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.987853647705775,
+      "learning_rate": 1.9319568012325785e-06,
+      "loss": 0.346,
+      "step": 13045
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.821900932918403,
+      "learning_rate": 1.928053268420314e-06,
+      "loss": 0.3535,
+      "step": 13050
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.759054880221154,
+      "learning_rate": 1.924152740976397e-06,
+      "loss": 0.3507,
+      "step": 13055
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.184172045514366,
+      "learning_rate": 1.920255222716815e-06,
+      "loss": 0.3547,
+      "step": 13060
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.550200708516813,
+      "learning_rate": 1.916360717454618e-06,
+      "loss": 0.3586,
+      "step": 13065
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.201643974584935,
+      "learning_rate": 1.9124692289999043e-06,
+      "loss": 0.3534,
+      "step": 13070
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.145930153650735,
+      "learning_rate": 1.908580761159822e-06,
+      "loss": 0.3493,
+      "step": 13075
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.265935915615525,
+      "learning_rate": 1.9046953177385623e-06,
+      "loss": 0.3474,
+      "step": 13080
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.750015111410342,
+      "learning_rate": 1.9008129025373629e-06,
+      "loss": 0.3522,
+      "step": 13085
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.615266133527854,
+      "learning_rate": 1.896933519354493e-06,
+      "loss": 0.3465,
+      "step": 13090
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.064289801459999,
+      "learning_rate": 1.893057171985257e-06,
+      "loss": 0.3433,
+      "step": 13095
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.437705557970718,
+      "learning_rate": 1.889183864221993e-06,
+      "loss": 0.3432,
+      "step": 13100
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.890599420576279,
+      "learning_rate": 1.885313599854059e-06,
+      "loss": 0.3489,
+      "step": 13105
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.3124975455109755,
+      "learning_rate": 1.8814463826678442e-06,
+      "loss": 0.3511,
+      "step": 13110
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 9.871965635586927,
+      "learning_rate": 1.877582216446745e-06,
+      "loss": 0.3502,
+      "step": 13115
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.646315847922393,
+      "learning_rate": 1.873721104971184e-06,
+      "loss": 0.3505,
+      "step": 13120
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.027378772777941,
+      "learning_rate": 1.8698630520185874e-06,
+      "loss": 0.3479,
+      "step": 13125
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.038666599099896,
+      "learning_rate": 1.8660080613633963e-06,
+      "loss": 0.347,
+      "step": 13130
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.010673079253129,
+      "learning_rate": 1.8621561367770497e-06,
+      "loss": 0.3529,
+      "step": 13135
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 8.602362535722602,
+      "learning_rate": 1.8583072820279885e-06,
+      "loss": 0.3444,
+      "step": 13140
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.896038325823834,
+      "learning_rate": 1.8544615008816536e-06,
+      "loss": 0.3442,
+      "step": 13145
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.876220617037956,
+      "learning_rate": 1.8506187971004753e-06,
+      "loss": 0.3434,
+      "step": 13150
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.195201460399498,
+      "learning_rate": 1.8467791744438735e-06,
+      "loss": 0.3522,
+      "step": 13155
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.816129760139536,
+      "learning_rate": 1.842942636668254e-06,
+      "loss": 0.3514,
+      "step": 13160
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.810474650882271,
+      "learning_rate": 1.8391091875270083e-06,
+      "loss": 0.3541,
+      "step": 13165
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.278313176164161,
+      "learning_rate": 1.8352788307704994e-06,
+      "loss": 0.3551,
+      "step": 13170
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.719003770806738,
+      "learning_rate": 1.8314515701460728e-06,
+      "loss": 0.3526,
+      "step": 13175
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.4429678197703275,
+      "learning_rate": 1.8276274093980378e-06,
+      "loss": 0.3465,
+      "step": 13180
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 8.14256452509486,
+      "learning_rate": 1.823806352267673e-06,
+      "loss": 0.3431,
+      "step": 13185
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.746051066675634,
+      "learning_rate": 1.8199884024932269e-06,
+      "loss": 0.3499,
+      "step": 13190
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.986311949760354,
+      "learning_rate": 1.8161735638098954e-06,
+      "loss": 0.3448,
+      "step": 13195
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.914570318914096,
+      "learning_rate": 1.8123618399498443e-06,
+      "loss": 0.3507,
+      "step": 13200
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.1579141005865,
+      "learning_rate": 1.8085532346421813e-06,
+      "loss": 0.3493,
+      "step": 13205
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.884851366646707,
+      "learning_rate": 1.8047477516129714e-06,
+      "loss": 0.3462,
+      "step": 13210
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.922576484520524,
+      "learning_rate": 1.800945394585218e-06,
+      "loss": 0.3484,
+      "step": 13215
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.022729298673702,
+      "learning_rate": 1.797146167278873e-06,
+      "loss": 0.3559,
+      "step": 13220
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.15404561937025,
+      "learning_rate": 1.7933500734108217e-06,
+      "loss": 0.3461,
+      "step": 13225
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.677057476274528,
+      "learning_rate": 1.7895571166948839e-06,
+      "loss": 0.3531,
+      "step": 13230
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 16.877583132023236,
+      "learning_rate": 1.7857673008418126e-06,
+      "loss": 0.3499,
+      "step": 13235
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.918368999806285,
+      "learning_rate": 1.7819806295592846e-06,
+      "loss": 0.3479,
+      "step": 13240
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 13.074015583450297,
+      "learning_rate": 1.778197106551906e-06,
+      "loss": 0.3469,
+      "step": 13245
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.204338553008846,
+      "learning_rate": 1.7744167355211967e-06,
+      "loss": 0.3437,
+      "step": 13250
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.941548788313813,
+      "learning_rate": 1.770639520165598e-06,
+      "loss": 0.3347,
+      "step": 13255
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.311730600707324,
+      "learning_rate": 1.7668654641804583e-06,
+      "loss": 0.3517,
+      "step": 13260
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 6.954675892706177,
+      "learning_rate": 1.7630945712580427e-06,
+      "loss": 0.3657,
+      "step": 13265
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.517930593178529,
+      "learning_rate": 1.7593268450875145e-06,
+      "loss": 0.3505,
+      "step": 13270
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 10.857689798245556,
+      "learning_rate": 1.7555622893549429e-06,
+      "loss": 0.3502,
+      "step": 13275
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 6.640722053004046,
+      "learning_rate": 1.751800907743294e-06,
+      "loss": 0.3531,
+      "step": 13280
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.540364031508203,
+      "learning_rate": 1.7480427039324266e-06,
+      "loss": 0.3543,
+      "step": 13285
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 6.191867604546337,
+      "learning_rate": 1.7442876815990972e-06,
+      "loss": 0.3579,
+      "step": 13290
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 10.730074278228338,
+      "learning_rate": 1.7405358444169413e-06,
+      "loss": 0.3585,
+      "step": 13295
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 6.655593162182091,
+      "learning_rate": 1.7367871960564865e-06,
+      "loss": 0.3635,
+      "step": 13300
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.232812680915814,
+      "learning_rate": 1.7330417401851317e-06,
+      "loss": 0.3463,
+      "step": 13305
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.764333836098204,
+      "learning_rate": 1.7292994804671648e-06,
+      "loss": 0.3489,
+      "step": 13310
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.560495803368249,
+      "learning_rate": 1.7255604205637305e-06,
+      "loss": 0.3567,
+      "step": 13315
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 6.599711992803756,
+      "learning_rate": 1.7218245641328585e-06,
+      "loss": 0.3439,
+      "step": 13320
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.5763943704902665,
+      "learning_rate": 1.7180919148294356e-06,
+      "loss": 0.3478,
+      "step": 13325
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.192559204919637,
+      "learning_rate": 1.7143624763052113e-06,
+      "loss": 0.3555,
+      "step": 13330
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.530792324752998,
+      "learning_rate": 1.7106362522088e-06,
+      "loss": 0.3484,
+      "step": 13335
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.030676755391766,
+      "learning_rate": 1.7069132461856636e-06,
+      "loss": 0.3502,
+      "step": 13340
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.3288817707798675,
+      "learning_rate": 1.703193461878122e-06,
+      "loss": 0.3426,
+      "step": 13345
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.86473762505183,
+      "learning_rate": 1.69947690292534e-06,
+      "loss": 0.3425,
+      "step": 13350
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.179694085709815,
+      "learning_rate": 1.6957635729633265e-06,
+      "loss": 0.3411,
+      "step": 13355
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.590706960514583,
+      "learning_rate": 1.6920534756249313e-06,
+      "loss": 0.3386,
+      "step": 13360
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.085792274530593,
+      "learning_rate": 1.6883466145398458e-06,
+      "loss": 0.3517,
+      "step": 13365
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 9.457605435397486,
+      "learning_rate": 1.6846429933345909e-06,
+      "loss": 0.3608,
+      "step": 13370
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.232165753807701,
+      "learning_rate": 1.6809426156325165e-06,
+      "loss": 0.3487,
+      "step": 13375
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.420720267383381,
+      "learning_rate": 1.6772454850538062e-06,
+      "loss": 0.3393,
+      "step": 13380
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 7.003590199085304,
+      "learning_rate": 1.6735516052154581e-06,
+      "loss": 0.3444,
+      "step": 13385
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.89964111274835,
+      "learning_rate": 1.6698609797313015e-06,
+      "loss": 0.3614,
+      "step": 13390
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.51153800199699,
+      "learning_rate": 1.666173612211966e-06,
+      "loss": 0.3566,
+      "step": 13395
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.191882587472384,
+      "learning_rate": 1.6624895062649087e-06,
+      "loss": 0.3313,
+      "step": 13400
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.991928299959772,
+      "learning_rate": 1.658808665494387e-06,
+      "loss": 0.3519,
+      "step": 13405
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.8209245667833525,
+      "learning_rate": 1.6551310935014686e-06,
+      "loss": 0.3414,
+      "step": 13410
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.351080819575392,
+      "learning_rate": 1.6514567938840215e-06,
+      "loss": 0.3518,
+      "step": 13415
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.231585925437265,
+      "learning_rate": 1.6477857702367088e-06,
+      "loss": 0.3517,
+      "step": 13420
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.649497456792598,
+      "learning_rate": 1.644118026150996e-06,
+      "loss": 0.3473,
+      "step": 13425
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.598271100066865,
+      "learning_rate": 1.640453565215135e-06,
+      "loss": 0.3471,
+      "step": 13430
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.560216547673767,
+      "learning_rate": 1.636792391014166e-06,
+      "loss": 0.3402,
+      "step": 13435
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.525873620362791,
+      "learning_rate": 1.6331345071299126e-06,
+      "loss": 0.3444,
+      "step": 13440
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.65064534196253,
+      "learning_rate": 1.6294799171409847e-06,
+      "loss": 0.3358,
+      "step": 13445
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.884507828410077,
+      "learning_rate": 1.6258286246227639e-06,
+      "loss": 0.3444,
+      "step": 13450
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.0941968873682155,
+      "learning_rate": 1.6221806331474105e-06,
+      "loss": 0.3451,
+      "step": 13455
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.836567699217456,
+      "learning_rate": 1.6185359462838517e-06,
+      "loss": 0.342,
+      "step": 13460
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.717508618110533,
+      "learning_rate": 1.614894567597781e-06,
+      "loss": 0.3408,
+      "step": 13465
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.417072529234567,
+      "learning_rate": 1.6112565006516628e-06,
+      "loss": 0.348,
+      "step": 13470
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.232586277623032,
+      "learning_rate": 1.6076217490047092e-06,
+      "loss": 0.3471,
+      "step": 13475
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 12.738691740688402,
+      "learning_rate": 1.6039903162129005e-06,
+      "loss": 0.346,
+      "step": 13480
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.476783163143757,
+      "learning_rate": 1.6003622058289625e-06,
+      "loss": 0.339,
+      "step": 13485
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.1786516010413735,
+      "learning_rate": 1.5967374214023767e-06,
+      "loss": 0.3426,
+      "step": 13490
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.240119993457492,
+      "learning_rate": 1.5931159664793638e-06,
+      "loss": 0.3576,
+      "step": 13495
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.617265515344786,
+      "learning_rate": 1.5894978446028948e-06,
+      "loss": 0.3571,
+      "step": 13500
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.100899996233863,
+      "learning_rate": 1.5858830593126733e-06,
+      "loss": 0.3476,
+      "step": 13505
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.180130314731739,
+      "learning_rate": 1.582271614145142e-06,
+      "loss": 0.3479,
+      "step": 13510
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.490433584866408,
+      "learning_rate": 1.5786635126334748e-06,
+      "loss": 0.3493,
+      "step": 13515
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.815477303917275,
+      "learning_rate": 1.5750587583075732e-06,
+      "loss": 0.3326,
+      "step": 13520
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.583652687710298,
+      "learning_rate": 1.5714573546940692e-06,
+      "loss": 0.3457,
+      "step": 13525
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.454232968768451,
+      "learning_rate": 1.5678593053163093e-06,
+      "loss": 0.3405,
+      "step": 13530
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.43772017874306,
+      "learning_rate": 1.5642646136943657e-06,
+      "loss": 0.3352,
+      "step": 13535
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.122617314101993,
+      "learning_rate": 1.5606732833450189e-06,
+      "loss": 0.3398,
+      "step": 13540
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.665676662010156,
+      "learning_rate": 1.5570853177817675e-06,
+      "loss": 0.3396,
+      "step": 13545
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.511840143186782,
+      "learning_rate": 1.5535007205148134e-06,
+      "loss": 0.3554,
+      "step": 13550
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.798262124458812,
+      "learning_rate": 1.549919495051065e-06,
+      "loss": 0.3454,
+      "step": 13555
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 9.802067833224381,
+      "learning_rate": 1.546341644894131e-06,
+      "loss": 0.3409,
+      "step": 13560
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.21854918317779,
+      "learning_rate": 1.5427671735443179e-06,
+      "loss": 0.3396,
+      "step": 13565
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 9.027154433511654,
+      "learning_rate": 1.5391960844986303e-06,
+      "loss": 0.3347,
+      "step": 13570
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.9085323414991215,
+      "learning_rate": 1.5356283812507583e-06,
+      "loss": 0.3556,
+      "step": 13575
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.97029483532059,
+      "learning_rate": 1.5320640672910847e-06,
+      "loss": 0.3553,
+      "step": 13580
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.560186552875563,
+      "learning_rate": 1.5285031461066707e-06,
+      "loss": 0.3388,
+      "step": 13585
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.753874546064374,
+      "learning_rate": 1.524945621181267e-06,
+      "loss": 0.3467,
+      "step": 13590
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.343057377128513,
+      "learning_rate": 1.521391495995289e-06,
+      "loss": 0.3394,
+      "step": 13595
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.353250063424364,
+      "learning_rate": 1.517840774025839e-06,
+      "loss": 0.3498,
+      "step": 13600
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 7.217358273512686,
+      "learning_rate": 1.5142934587466818e-06,
+      "loss": 0.3369,
+      "step": 13605
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.135377518107784,
+      "learning_rate": 1.5107495536282501e-06,
+      "loss": 0.3443,
+      "step": 13610
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.732260437297867,
+      "learning_rate": 1.507209062137645e-06,
+      "loss": 0.3451,
+      "step": 13615
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.850580660514156,
+      "learning_rate": 1.5036719877386219e-06,
+      "loss": 0.3456,
+      "step": 13620
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.058081962272013,
+      "learning_rate": 1.5001383338915992e-06,
+      "loss": 0.3438,
+      "step": 13625
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.18034489260238,
+      "learning_rate": 1.496608104053644e-06,
+      "loss": 0.342,
+      "step": 13630
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.738934622372447,
+      "learning_rate": 1.4930813016784757e-06,
+      "loss": 0.3525,
+      "step": 13635
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 7.446377218869251,
+      "learning_rate": 1.4895579302164582e-06,
+      "loss": 0.344,
+      "step": 13640
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.134281629105313,
+      "learning_rate": 1.486037993114604e-06,
+      "loss": 0.3424,
+      "step": 13645
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.422097460058927,
+      "learning_rate": 1.482521493816561e-06,
+      "loss": 0.3414,
+      "step": 13650
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 8.43992691411807,
+      "learning_rate": 1.4790084357626144e-06,
+      "loss": 0.3343,
+      "step": 13655
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.523273549487832,
+      "learning_rate": 1.475498822389686e-06,
+      "loss": 0.3338,
+      "step": 13660
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.435185890787804,
+      "learning_rate": 1.4719926571313225e-06,
+      "loss": 0.3386,
+      "step": 13665
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 8.036234796730652,
+      "learning_rate": 1.4684899434177042e-06,
+      "loss": 0.3418,
+      "step": 13670
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.985559920161983,
+      "learning_rate": 1.4649906846756246e-06,
+      "loss": 0.3483,
+      "step": 13675
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.198127442793323,
+      "learning_rate": 1.4614948843285075e-06,
+      "loss": 0.3383,
+      "step": 13680
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.405724182538725,
+      "learning_rate": 1.4580025457963853e-06,
+      "loss": 0.3441,
+      "step": 13685
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.447408149158544,
+      "learning_rate": 1.4545136724959103e-06,
+      "loss": 0.3393,
+      "step": 13690
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.682546730853727,
+      "learning_rate": 1.4510282678403398e-06,
+      "loss": 0.3432,
+      "step": 13695
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.311866051573056,
+      "learning_rate": 1.4475463352395375e-06,
+      "loss": 0.3409,
+      "step": 13700
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.960718718240998,
+      "learning_rate": 1.4440678780999755e-06,
+      "loss": 0.3392,
+      "step": 13705
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.020210423031231,
+      "learning_rate": 1.4405928998247198e-06,
+      "loss": 0.3457,
+      "step": 13710
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 7.037044644399787,
+      "learning_rate": 1.4371214038134369e-06,
+      "loss": 0.3529,
+      "step": 13715
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 6.904175449827646,
+      "learning_rate": 1.4336533934623815e-06,
+      "loss": 0.3381,
+      "step": 13720
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.645039334534665,
+      "learning_rate": 1.4301888721644059e-06,
+      "loss": 0.3474,
+      "step": 13725
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.999952519704578,
+      "learning_rate": 1.426727843308942e-06,
+      "loss": 0.3431,
+      "step": 13730
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.703475806749105,
+      "learning_rate": 1.42327031028201e-06,
+      "loss": 0.3368,
+      "step": 13735
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 7.277357316094131,
+      "learning_rate": 1.419816276466206e-06,
+      "loss": 0.338,
+      "step": 13740
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.857459866364857,
+      "learning_rate": 1.4163657452407037e-06,
+      "loss": 0.3473,
+      "step": 13745
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.60983307528463,
+      "learning_rate": 1.4129187199812539e-06,
+      "loss": 0.341,
+      "step": 13750
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.131631743151706,
+      "learning_rate": 1.4094752040601722e-06,
+      "loss": 0.3404,
+      "step": 13755
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.013967935247216,
+      "learning_rate": 1.4060352008463434e-06,
+      "loss": 0.3361,
+      "step": 13760
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.727283192259992,
+      "learning_rate": 1.4025987137052138e-06,
+      "loss": 0.3404,
+      "step": 13765
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.777124384197018,
+      "learning_rate": 1.399165745998795e-06,
+      "loss": 0.3486,
+      "step": 13770
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.506027592709238,
+      "learning_rate": 1.3957363010856485e-06,
+      "loss": 0.3348,
+      "step": 13775
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.325554245222836,
+      "learning_rate": 1.3923103823208956e-06,
+      "loss": 0.3505,
+      "step": 13780
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.141132954617194,
+      "learning_rate": 1.3888879930562033e-06,
+      "loss": 0.3359,
+      "step": 13785
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.567212775700831,
+      "learning_rate": 1.3854691366397866e-06,
+      "loss": 0.3457,
+      "step": 13790
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.297810518315015,
+      "learning_rate": 1.3820538164164093e-06,
+      "loss": 0.3438,
+      "step": 13795
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 9.281204002436738,
+      "learning_rate": 1.378642035727365e-06,
+      "loss": 0.3369,
+      "step": 13800
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.36525808599186,
+      "learning_rate": 1.375233797910495e-06,
+      "loss": 0.3365,
+      "step": 13805
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.4490501640437765,
+      "learning_rate": 1.3718291063001682e-06,
+      "loss": 0.3398,
+      "step": 13810
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.715687665200049,
+      "learning_rate": 1.3684279642272885e-06,
+      "loss": 0.3455,
+      "step": 13815
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 6.08889261493712,
+      "learning_rate": 1.3650303750192817e-06,
+      "loss": 0.344,
+      "step": 13820
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.533607736470023,
+      "learning_rate": 1.361636342000105e-06,
+      "loss": 0.3381,
+      "step": 13825
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 9.119921821708115,
+      "learning_rate": 1.35824586849023e-06,
+      "loss": 0.3522,
+      "step": 13830
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 7.88098686226356,
+      "learning_rate": 1.3548589578066483e-06,
+      "loss": 0.3347,
+      "step": 13835
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.405121362717537,
+      "learning_rate": 1.351475613262867e-06,
+      "loss": 0.343,
+      "step": 13840
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.427892201945487,
+      "learning_rate": 1.3480958381689007e-06,
+      "loss": 0.3356,
+      "step": 13845
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.109650437688848,
+      "learning_rate": 1.3447196358312785e-06,
+      "loss": 0.336,
+      "step": 13850
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.899595498191738,
+      "learning_rate": 1.3413470095530267e-06,
+      "loss": 0.3448,
+      "step": 13855
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 13.105610717701218,
+      "learning_rate": 1.3379779626336792e-06,
+      "loss": 0.3493,
+      "step": 13860
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.477048054743193,
+      "learning_rate": 1.3346124983692633e-06,
+      "loss": 0.3459,
+      "step": 13865
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.3026896755353565,
+      "learning_rate": 1.3312506200523056e-06,
+      "loss": 0.3442,
+      "step": 13870
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.830324581300535,
+      "learning_rate": 1.3278923309718216e-06,
+      "loss": 0.3424,
+      "step": 13875
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.589760157400933,
+      "learning_rate": 1.3245376344133154e-06,
+      "loss": 0.3367,
+      "step": 13880
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.4284197300354275,
+      "learning_rate": 1.321186533658778e-06,
+      "loss": 0.345,
+      "step": 13885
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.245485052632024,
+      "learning_rate": 1.3178390319866796e-06,
+      "loss": 0.343,
+      "step": 13890
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.468765278758668,
+      "learning_rate": 1.314495132671974e-06,
+      "loss": 0.3358,
+      "step": 13895
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.730884937980976,
+      "learning_rate": 1.3111548389860856e-06,
+      "loss": 0.3532,
+      "step": 13900
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.02533076184626,
+      "learning_rate": 1.307818154196917e-06,
+      "loss": 0.3387,
+      "step": 13905
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.535237415331175,
+      "learning_rate": 1.3044850815688336e-06,
+      "loss": 0.3372,
+      "step": 13910
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.078079559136438,
+      "learning_rate": 1.3011556243626744e-06,
+      "loss": 0.3395,
+      "step": 13915
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.862215623470651,
+      "learning_rate": 1.2978297858357319e-06,
+      "loss": 0.3286,
+      "step": 13920
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.075655997097684,
+      "learning_rate": 1.294507569241768e-06,
+      "loss": 0.3476,
+      "step": 13925
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.601342646915005,
+      "learning_rate": 1.291188977830995e-06,
+      "loss": 0.3412,
+      "step": 13930
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.377658288362509,
+      "learning_rate": 1.2878740148500784e-06,
+      "loss": 0.3469,
+      "step": 13935
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.988578510111449,
+      "learning_rate": 1.2845626835421405e-06,
+      "loss": 0.3386,
+      "step": 13940
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.089807389276085,
+      "learning_rate": 1.2812549871467417e-06,
+      "loss": 0.3514,
+      "step": 13945
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.255751972220425,
+      "learning_rate": 1.2779509288998937e-06,
+      "loss": 0.3441,
+      "step": 13950
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 7.210664216579693,
+      "learning_rate": 1.2746505120340447e-06,
+      "loss": 0.3439,
+      "step": 13955
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.13247241723326,
+      "learning_rate": 1.271353739778081e-06,
+      "loss": 0.3329,
+      "step": 13960
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.318434472721774,
+      "learning_rate": 1.2680606153573233e-06,
+      "loss": 0.3416,
+      "step": 13965
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 7.922606543380108,
+      "learning_rate": 1.264771141993526e-06,
+      "loss": 0.3374,
+      "step": 13970
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 7.009083560944203,
+      "learning_rate": 1.261485322904869e-06,
+      "loss": 0.3347,
+      "step": 13975
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.43461490473705,
+      "learning_rate": 1.2582031613059553e-06,
+      "loss": 0.3399,
+      "step": 13980
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 7.245591319250889,
+      "learning_rate": 1.2549246604078164e-06,
+      "loss": 0.3403,
+      "step": 13985
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 8.551751274701207,
+      "learning_rate": 1.2516498234178937e-06,
+      "loss": 0.3415,
+      "step": 13990
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.776017027952166,
+      "learning_rate": 1.2483786535400538e-06,
+      "loss": 0.3407,
+      "step": 13995
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.242890436893526,
+      "learning_rate": 1.2451111539745646e-06,
+      "loss": 0.3426,
+      "step": 14000
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.3470389354152426,
+      "learning_rate": 1.2418473279181132e-06,
+      "loss": 0.3338,
+      "step": 14005
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.534546174432391,
+      "learning_rate": 1.238587178563786e-06,
+      "loss": 0.3398,
+      "step": 14010
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.389675063098023,
+      "learning_rate": 1.2353307091010775e-06,
+      "loss": 0.3376,
+      "step": 14015
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.168606843433336,
+      "learning_rate": 1.2320779227158786e-06,
+      "loss": 0.3459,
+      "step": 14020
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.196101054526336,
+      "learning_rate": 1.2288288225904766e-06,
+      "loss": 0.3403,
+      "step": 14025
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.274286352545161,
+      "learning_rate": 1.225583411903556e-06,
+      "loss": 0.3325,
+      "step": 14030
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.637182685870529,
+      "learning_rate": 1.2223416938301885e-06,
+      "loss": 0.3291,
+      "step": 14035
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.649448721624295,
+      "learning_rate": 1.2191036715418347e-06,
+      "loss": 0.3358,
+      "step": 14040
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.74179880034857,
+      "learning_rate": 1.2158693482063377e-06,
+      "loss": 0.3353,
+      "step": 14045
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.631733274197324,
+      "learning_rate": 1.2126387269879254e-06,
+      "loss": 0.3378,
+      "step": 14050
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.935487631839041,
+      "learning_rate": 1.2094118110471998e-06,
+      "loss": 0.3474,
+      "step": 14055
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.9412089332004983,
+      "learning_rate": 1.2061886035411412e-06,
+      "loss": 0.3331,
+      "step": 14060
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.5029694305110555,
+      "learning_rate": 1.202969107623101e-06,
+      "loss": 0.3372,
+      "step": 14065
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.074013228108049,
+      "learning_rate": 1.1997533264427958e-06,
+      "loss": 0.3297,
+      "step": 14070
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.3155285236036915,
+      "learning_rate": 1.1965412631463164e-06,
+      "loss": 0.3376,
+      "step": 14075
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.655360501282674,
+      "learning_rate": 1.193332920876104e-06,
+      "loss": 0.3341,
+      "step": 14080
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.8202273238215465,
+      "learning_rate": 1.190128302770972e-06,
+      "loss": 0.3317,
+      "step": 14085
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.661250004762066,
+      "learning_rate": 1.1869274119660818e-06,
+      "loss": 0.3341,
+      "step": 14090
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.750711228093705,
+      "learning_rate": 1.1837302515929526e-06,
+      "loss": 0.3319,
+      "step": 14095
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.636421776212692,
+      "learning_rate": 1.180536824779452e-06,
+      "loss": 0.3328,
+      "step": 14100
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.944300055635633,
+      "learning_rate": 1.177347134649796e-06,
+      "loss": 0.35,
+      "step": 14105
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.783967696727534,
+      "learning_rate": 1.1741611843245448e-06,
+      "loss": 0.3384,
+      "step": 14110
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.422475390958299,
+      "learning_rate": 1.1709789769205993e-06,
+      "loss": 0.3407,
+      "step": 14115
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.936647200414145,
+      "learning_rate": 1.1678005155511984e-06,
+      "loss": 0.3398,
+      "step": 14120
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.250589249776812,
+      "learning_rate": 1.164625803325915e-06,
+      "loss": 0.3458,
+      "step": 14125
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.970889235761936,
+      "learning_rate": 1.1614548433506596e-06,
+      "loss": 0.3436,
+      "step": 14130
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.836692761132451,
+      "learning_rate": 1.1582876387276636e-06,
+      "loss": 0.336,
+      "step": 14135
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 9.944435936043519,
+      "learning_rate": 1.1551241925554923e-06,
+      "loss": 0.3367,
+      "step": 14140
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.236576529492092,
+      "learning_rate": 1.1519645079290277e-06,
+      "loss": 0.3384,
+      "step": 14145
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.470189006998729,
+      "learning_rate": 1.1488085879394773e-06,
+      "loss": 0.3485,
+      "step": 14150
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.5188328073800585,
+      "learning_rate": 1.145656435674361e-06,
+      "loss": 0.3329,
+      "step": 14155
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.352260040907254,
+      "learning_rate": 1.1425080542175143e-06,
+      "loss": 0.3341,
+      "step": 14160
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.822775962292289,
+      "learning_rate": 1.1393634466490843e-06,
+      "loss": 0.3405,
+      "step": 14165
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.361876934910723,
+      "learning_rate": 1.1362226160455237e-06,
+      "loss": 0.3344,
+      "step": 14170
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 8.88786874074134,
+      "learning_rate": 1.1330855654795948e-06,
+      "loss": 0.3389,
+      "step": 14175
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 14.262895017022414,
+      "learning_rate": 1.1299522980203554e-06,
+      "loss": 0.3286,
+      "step": 14180
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.149422771083544,
+      "learning_rate": 1.1268228167331686e-06,
+      "loss": 0.3461,
+      "step": 14185
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.574946646085239,
+      "learning_rate": 1.123697124679688e-06,
+      "loss": 0.349,
+      "step": 14190
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 16.266679457171783,
+      "learning_rate": 1.120575224917866e-06,
+      "loss": 0.3411,
+      "step": 14195
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.663923228409695,
+      "learning_rate": 1.1174571205019358e-06,
+      "loss": 0.3334,
+      "step": 14200
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.356748299028564,
+      "learning_rate": 1.114342814482428e-06,
+      "loss": 0.3316,
+      "step": 14205
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.616836478636501,
+      "learning_rate": 1.11123230990615e-06,
+      "loss": 0.3516,
+      "step": 14210
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.507223327833415,
+      "learning_rate": 1.1081256098161913e-06,
+      "loss": 0.3408,
+      "step": 14215
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.034187350742189,
+      "learning_rate": 1.1050227172519234e-06,
+      "loss": 0.3415,
+      "step": 14220
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 15.689685827911726,
+      "learning_rate": 1.1019236352489865e-06,
+      "loss": 0.3412,
+      "step": 14225
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.8734984726878245,
+      "learning_rate": 1.098828366839299e-06,
+      "loss": 0.3357,
+      "step": 14230
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.382933277091549,
+      "learning_rate": 1.0957369150510445e-06,
+      "loss": 0.3475,
+      "step": 14235
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.1833242818970815,
+      "learning_rate": 1.0926492829086728e-06,
+      "loss": 0.3384,
+      "step": 14240
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.661831755769597,
+      "learning_rate": 1.089565473432897e-06,
+      "loss": 0.3314,
+      "step": 14245
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.278366302398714,
+      "learning_rate": 1.086485489640694e-06,
+      "loss": 0.3407,
+      "step": 14250
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.140543659429433,
+      "learning_rate": 1.0834093345452934e-06,
+      "loss": 0.3349,
+      "step": 14255
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.1394738008804595,
+      "learning_rate": 1.0803370111561789e-06,
+      "loss": 0.3335,
+      "step": 14260
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.4178291423734315,
+      "learning_rate": 1.0772685224790907e-06,
+      "loss": 0.3406,
+      "step": 14265
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.519217722729064,
+      "learning_rate": 1.0742038715160108e-06,
+      "loss": 0.3403,
+      "step": 14270
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.799302187562816,
+      "learning_rate": 1.0711430612651747e-06,
+      "loss": 0.3458,
+      "step": 14275
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.851722917122433,
+      "learning_rate": 1.0680860947210492e-06,
+      "loss": 0.3329,
+      "step": 14280
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 7.937356964578476,
+      "learning_rate": 1.065032974874352e-06,
+      "loss": 0.332,
+      "step": 14285
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.096391120098633,
+      "learning_rate": 1.0619837047120296e-06,
+      "loss": 0.3375,
+      "step": 14290
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.967495395371477,
+      "learning_rate": 1.0589382872172682e-06,
+      "loss": 0.3434,
+      "step": 14295
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.419841773328182,
+      "learning_rate": 1.0558967253694802e-06,
+      "loss": 0.336,
+      "step": 14300
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.0973018817246984,
+      "learning_rate": 1.0528590221443064e-06,
+      "loss": 0.3383,
+      "step": 14305
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.074301199329289,
+      "learning_rate": 1.0498251805136162e-06,
+      "loss": 0.3361,
+      "step": 14310
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.976987817402853,
+      "learning_rate": 1.0467952034454976e-06,
+      "loss": 0.334,
+      "step": 14315
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.122772139785621,
+      "learning_rate": 1.0437690939042594e-06,
+      "loss": 0.3376,
+      "step": 14320
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.564482802942002,
+      "learning_rate": 1.0407468548504234e-06,
+      "loss": 0.3302,
+      "step": 14325
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.250238575968118,
+      "learning_rate": 1.0377284892407318e-06,
+      "loss": 0.3333,
+      "step": 14330
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.218686261207107,
+      "learning_rate": 1.0347140000281297e-06,
+      "loss": 0.3353,
+      "step": 14335
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.063182793310301,
+      "learning_rate": 1.0317033901617763e-06,
+      "loss": 0.3419,
+      "step": 14340
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.656996118444409,
+      "learning_rate": 1.0286966625870304e-06,
+      "loss": 0.3317,
+      "step": 14345
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.4213939092521874,
+      "learning_rate": 1.0256938202454536e-06,
+      "loss": 0.3429,
+      "step": 14350
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 8.364522415132814,
+      "learning_rate": 1.022694866074812e-06,
+      "loss": 0.334,
+      "step": 14355
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.928770765768932,
+      "learning_rate": 1.0196998030090577e-06,
+      "loss": 0.33,
+      "step": 14360
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.018321446051179,
+      "learning_rate": 1.0167086339783455e-06,
+      "loss": 0.3488,
+      "step": 14365
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.015085898456728,
+      "learning_rate": 1.0137213619090142e-06,
+      "loss": 0.336,
+      "step": 14370
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 10.398286045146108,
+      "learning_rate": 1.0107379897235959e-06,
+      "loss": 0.356,
+      "step": 14375
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.891907399568669,
+      "learning_rate": 1.0077585203408003e-06,
+      "loss": 0.3352,
+      "step": 14380
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.407190738784242,
+      "learning_rate": 1.0047829566755262e-06,
+      "loss": 0.3306,
+      "step": 14385
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.082744545042239,
+      "learning_rate": 1.001811301638846e-06,
+      "loss": 0.3399,
+      "step": 14390
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.318854243071567,
+      "learning_rate": 9.988435581380102e-07,
+      "loss": 0.3395,
+      "step": 14395
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.723240011964577,
+      "learning_rate": 9.95879729076442e-07,
+      "loss": 0.3308,
+      "step": 14400
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.3405520836823985,
+      "learning_rate": 9.929198173537346e-07,
+      "loss": 0.3298,
+      "step": 14405
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.770996377634228,
+      "learning_rate": 9.899638258656518e-07,
+      "loss": 0.3364,
+      "step": 14410
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.911279512730939,
+      "learning_rate": 9.870117575041172e-07,
+      "loss": 0.3254,
+      "step": 14415
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.785308442014827,
+      "learning_rate": 9.840636151572215e-07,
+      "loss": 0.3395,
+      "step": 14420
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.7477711398773295,
+      "learning_rate": 9.811194017092086e-07,
+      "loss": 0.3422,
+      "step": 14425
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.243548036699693,
+      "learning_rate": 9.781791200404855e-07,
+      "loss": 0.3408,
+      "step": 14430
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.289453405665378,
+      "learning_rate": 9.752427730276076e-07,
+      "loss": 0.3391,
+      "step": 14435
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.724613198135481,
+      "learning_rate": 9.723103635432823e-07,
+      "loss": 0.3361,
+      "step": 14440
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 8.191455303424572,
+      "learning_rate": 9.693818944563644e-07,
+      "loss": 0.341,
+      "step": 14445
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 11.469298515326782,
+      "learning_rate": 9.664573686318535e-07,
+      "loss": 0.3313,
+      "step": 14450
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.138448172334449,
+      "learning_rate": 9.635367889308945e-07,
+      "loss": 0.3467,
+      "step": 14455
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.280952234714455,
+      "learning_rate": 9.606201582107666e-07,
+      "loss": 0.3464,
+      "step": 14460
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.780709911013356,
+      "learning_rate": 9.577074793248908e-07,
+      "loss": 0.3312,
+      "step": 14465
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.421025545234076,
+      "learning_rate": 9.547987551228172e-07,
+      "loss": 0.3322,
+      "step": 14470
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.595092239009826,
+      "learning_rate": 9.518939884502315e-07,
+      "loss": 0.3373,
+      "step": 14475
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 6.702879081922312,
+      "learning_rate": 9.489931821489439e-07,
+      "loss": 0.3409,
+      "step": 14480
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.4970994508148445,
+      "learning_rate": 9.460963390568922e-07,
+      "loss": 0.3347,
+      "step": 14485
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.376402649743197,
+      "learning_rate": 9.432034620081349e-07,
+      "loss": 0.3371,
+      "step": 14490
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.8537603318498395,
+      "learning_rate": 9.403145538328512e-07,
+      "loss": 0.3202,
+      "step": 14495
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.55627346536457,
+      "learning_rate": 9.3742961735734e-07,
+      "loss": 0.3382,
+      "step": 14500
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.337405894436932,
+      "learning_rate": 9.3454865540401e-07,
+      "loss": 0.3389,
+      "step": 14505
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.558365503858435,
+      "learning_rate": 9.31671670791387e-07,
+      "loss": 0.3355,
+      "step": 14510
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.468843945765404,
+      "learning_rate": 9.287986663340998e-07,
+      "loss": 0.3321,
+      "step": 14515
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 8.908580862152249,
+      "learning_rate": 9.259296448428895e-07,
+      "loss": 0.3367,
+      "step": 14520
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.877623694657907,
+      "learning_rate": 9.230646091245932e-07,
+      "loss": 0.3399,
+      "step": 14525
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.82838509264817,
+      "learning_rate": 9.202035619821553e-07,
+      "loss": 0.3391,
+      "step": 14530
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.028694803195866,
+      "learning_rate": 9.173465062146148e-07,
+      "loss": 0.3369,
+      "step": 14535
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.237706071931036,
+      "learning_rate": 9.14493444617105e-07,
+      "loss": 0.334,
+      "step": 14540
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.825383076615841,
+      "learning_rate": 9.11644379980855e-07,
+      "loss": 0.3352,
+      "step": 14545
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.316499817256464,
+      "learning_rate": 9.087993150931801e-07,
+      "loss": 0.3326,
+      "step": 14550
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.911143231324706,
+      "learning_rate": 9.05958252737486e-07,
+      "loss": 0.3298,
+      "step": 14555
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.6358832021455445,
+      "learning_rate": 9.03121195693259e-07,
+      "loss": 0.332,
+      "step": 14560
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.947430264832714,
+      "learning_rate": 9.002881467360692e-07,
+      "loss": 0.3344,
+      "step": 14565
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 7.248449927175377,
+      "learning_rate": 8.974591086375634e-07,
+      "loss": 0.335,
+      "step": 14570
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.458948458123161,
+      "learning_rate": 8.946340841654677e-07,
+      "loss": 0.331,
+      "step": 14575
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.198133276647935,
+      "learning_rate": 8.918130760835797e-07,
+      "loss": 0.329,
+      "step": 14580
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.712618679236314,
+      "learning_rate": 8.88996087151765e-07,
+      "loss": 0.3355,
+      "step": 14585
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.002139102868113,
+      "learning_rate": 8.861831201259635e-07,
+      "loss": 0.3386,
+      "step": 14590
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.133980115384712,
+      "learning_rate": 8.833741777581739e-07,
+      "loss": 0.3274,
+      "step": 14595
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 8.734894342249994,
+      "learning_rate": 8.80569262796464e-07,
+      "loss": 0.333,
+      "step": 14600
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.053567103499762,
+      "learning_rate": 8.777683779849527e-07,
+      "loss": 0.3423,
+      "step": 14605
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.718447412100042,
+      "learning_rate": 8.749715260638247e-07,
+      "loss": 0.3273,
+      "step": 14610
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.513812256605959,
+      "learning_rate": 8.721787097693141e-07,
+      "loss": 0.3336,
+      "step": 14615
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.621238267112457,
+      "learning_rate": 8.693899318337095e-07,
+      "loss": 0.3302,
+      "step": 14620
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.445122668669881,
+      "learning_rate": 8.666051949853472e-07,
+      "loss": 0.3257,
+      "step": 14625
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.709037015599132,
+      "learning_rate": 8.638245019486091e-07,
+      "loss": 0.3338,
+      "step": 14630
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.360940980024568,
+      "learning_rate": 8.610478554439244e-07,
+      "loss": 0.3343,
+      "step": 14635
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.6143899100962,
+      "learning_rate": 8.582752581877607e-07,
+      "loss": 0.3326,
+      "step": 14640
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.7097884600896,
+      "learning_rate": 8.555067128926236e-07,
+      "loss": 0.3355,
+      "step": 14645
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.298828685434539,
+      "learning_rate": 8.52742222267055e-07,
+      "loss": 0.3291,
+      "step": 14650
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.260336253602938,
+      "learning_rate": 8.499817890156331e-07,
+      "loss": 0.3317,
+      "step": 14655
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.064517398086245,
+      "learning_rate": 8.47225415838962e-07,
+      "loss": 0.3272,
+      "step": 14660
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.483713586833253,
+      "learning_rate": 8.44473105433678e-07,
+      "loss": 0.3346,
+      "step": 14665
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.264668270536011,
+      "learning_rate": 8.417248604924394e-07,
+      "loss": 0.3399,
+      "step": 14670
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.172403867722749,
+      "learning_rate": 8.389806837039272e-07,
+      "loss": 0.3337,
+      "step": 14675
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.8912479936973945,
+      "learning_rate": 8.362405777528471e-07,
+      "loss": 0.3433,
+      "step": 14680
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.494072929459606,
+      "learning_rate": 8.335045453199142e-07,
+      "loss": 0.3418,
+      "step": 14685
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.405357055704501,
+      "learning_rate": 8.307725890818658e-07,
+      "loss": 0.3388,
+      "step": 14690
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.187527389254743,
+      "learning_rate": 8.280447117114465e-07,
+      "loss": 0.3323,
+      "step": 14695
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.665466163749549,
+      "learning_rate": 8.25320915877415e-07,
+      "loss": 0.3339,
+      "step": 14700
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.553292854497034,
+      "learning_rate": 8.226012042445308e-07,
+      "loss": 0.3361,
+      "step": 14705
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 6.328542620083238,
+      "learning_rate": 8.198855794735644e-07,
+      "loss": 0.3329,
+      "step": 14710
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.279384402377715,
+      "learning_rate": 8.171740442212833e-07,
+      "loss": 0.3492,
+      "step": 14715
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.645996678991626,
+      "learning_rate": 8.144666011404556e-07,
+      "loss": 0.3399,
+      "step": 14720
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.832297637021295,
+      "learning_rate": 8.117632528798458e-07,
+      "loss": 0.3485,
+      "step": 14725
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.637719837256791,
+      "learning_rate": 8.090640020842117e-07,
+      "loss": 0.3327,
+      "step": 14730
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.937432153179189,
+      "learning_rate": 8.063688513943046e-07,
+      "loss": 0.3298,
+      "step": 14735
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.247228681295939,
+      "learning_rate": 8.036778034468617e-07,
+      "loss": 0.3348,
+      "step": 14740
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.601109205992861,
+      "learning_rate": 8.009908608746097e-07,
+      "loss": 0.3265,
+      "step": 14745
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.610534797221197,
+      "learning_rate": 7.983080263062542e-07,
+      "loss": 0.3333,
+      "step": 14750
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 6.139040766800028,
+      "learning_rate": 7.956293023664879e-07,
+      "loss": 0.3309,
+      "step": 14755
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.412769842768002,
+      "learning_rate": 7.929546916759772e-07,
+      "loss": 0.3306,
+      "step": 14760
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.884387501900604,
+      "learning_rate": 7.902841968513652e-07,
+      "loss": 0.3341,
+      "step": 14765
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 6.846825276989522,
+      "learning_rate": 7.876178205052698e-07,
+      "loss": 0.3415,
+      "step": 14770
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.665868376155784,
+      "learning_rate": 7.849555652462775e-07,
+      "loss": 0.3269,
+      "step": 14775
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.423483816559562,
+      "learning_rate": 7.822974336789468e-07,
+      "loss": 0.3184,
+      "step": 14780
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.625052189667723,
+      "learning_rate": 7.796434284037973e-07,
+      "loss": 0.3375,
+      "step": 14785
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.130904432799836,
+      "learning_rate": 7.769935520173155e-07,
+      "loss": 0.3402,
+      "step": 14790
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.298938169660414,
+      "learning_rate": 7.743478071119459e-07,
+      "loss": 0.3269,
+      "step": 14795
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.9781642622490025,
+      "learning_rate": 7.717061962760947e-07,
+      "loss": 0.3334,
+      "step": 14800
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.914812405158503,
+      "learning_rate": 7.690687220941162e-07,
+      "loss": 0.3292,
+      "step": 14805
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.047822614780733,
+      "learning_rate": 7.664353871463264e-07,
+      "loss": 0.3182,
+      "step": 14810
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 7.91907198905064,
+      "learning_rate": 7.638061940089875e-07,
+      "loss": 0.3363,
+      "step": 14815
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.3497987908934865,
+      "learning_rate": 7.611811452543072e-07,
+      "loss": 0.331,
+      "step": 14820
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.267298014785521,
+      "learning_rate": 7.585602434504453e-07,
+      "loss": 0.3287,
+      "step": 14825
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.802029820505902,
+      "learning_rate": 7.559434911614977e-07,
+      "loss": 0.3252,
+      "step": 14830
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.32890062588132,
+      "learning_rate": 7.533308909475068e-07,
+      "loss": 0.3271,
+      "step": 14835
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.268375713285703,
+      "learning_rate": 7.507224453644474e-07,
+      "loss": 0.3405,
+      "step": 14840
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.399286170087374,
+      "learning_rate": 7.481181569642332e-07,
+      "loss": 0.3243,
+      "step": 14845
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.304404015063495,
+      "learning_rate": 7.455180282947083e-07,
+      "loss": 0.3337,
+      "step": 14850
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.029073052501918,
+      "learning_rate": 7.429220618996507e-07,
+      "loss": 0.3262,
+      "step": 14855
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.378391554889924,
+      "learning_rate": 7.40330260318764e-07,
+      "loss": 0.3384,
+      "step": 14860
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.621059063167957,
+      "learning_rate": 7.377426260876757e-07,
+      "loss": 0.3335,
+      "step": 14865
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.791322359845037,
+      "learning_rate": 7.351591617379411e-07,
+      "loss": 0.3324,
+      "step": 14870
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.678482618842939,
+      "learning_rate": 7.325798697970305e-07,
+      "loss": 0.3365,
+      "step": 14875
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.583563067785552,
+      "learning_rate": 7.300047527883375e-07,
+      "loss": 0.3338,
+      "step": 14880
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.391075538129845,
+      "learning_rate": 7.274338132311653e-07,
+      "loss": 0.3414,
+      "step": 14885
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.478843806589986,
+      "learning_rate": 7.248670536407354e-07,
+      "loss": 0.3129,
+      "step": 14890
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.5267513455016415,
+      "learning_rate": 7.223044765281767e-07,
+      "loss": 0.3326,
+      "step": 14895
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.660762723698685,
+      "learning_rate": 7.197460844005294e-07,
+      "loss": 0.3379,
+      "step": 14900
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 8.65434649620914,
+      "learning_rate": 7.171918797607369e-07,
+      "loss": 0.3294,
+      "step": 14905
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 7.1632545699544945,
+      "learning_rate": 7.146418651076443e-07,
+      "loss": 0.3294,
+      "step": 14910
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 10.121338816522481,
+      "learning_rate": 7.12096042936003e-07,
+      "loss": 0.3395,
+      "step": 14915
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.581526830547222,
+      "learning_rate": 7.095544157364575e-07,
+      "loss": 0.3359,
+      "step": 14920
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.044604807753416,
+      "learning_rate": 7.070169859955506e-07,
+      "loss": 0.335,
+      "step": 14925
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.000469258632066,
+      "learning_rate": 7.04483756195718e-07,
+      "loss": 0.3355,
+      "step": 14930
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.017536310415441,
+      "learning_rate": 7.019547288152872e-07,
+      "loss": 0.3344,
+      "step": 14935
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.900616115213088,
+      "learning_rate": 6.994299063284738e-07,
+      "loss": 0.3161,
+      "step": 14940
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.694735372740905,
+      "learning_rate": 6.969092912053798e-07,
+      "loss": 0.3301,
+      "step": 14945
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.03599084217576,
+      "learning_rate": 6.943928859119914e-07,
+      "loss": 0.3233,
+      "step": 14950
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.050873466380064,
+      "learning_rate": 6.91880692910174e-07,
+      "loss": 0.3257,
+      "step": 14955
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.478973666964457,
+      "learning_rate": 6.893727146576773e-07,
+      "loss": 0.3301,
+      "step": 14960
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.442444582301273,
+      "learning_rate": 6.868689536081197e-07,
+      "loss": 0.3403,
+      "step": 14965
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.26838730409746,
+      "learning_rate": 6.843694122110017e-07,
+      "loss": 0.3344,
+      "step": 14970
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.4244297663026595,
+      "learning_rate": 6.81874092911689e-07,
+      "loss": 0.327,
+      "step": 14975
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.237606573282461,
+      "learning_rate": 6.793829981514228e-07,
+      "loss": 0.3373,
+      "step": 14980
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.296940542734641,
+      "learning_rate": 6.768961303673055e-07,
+      "loss": 0.318,
+      "step": 14985
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.660218223419286,
+      "learning_rate": 6.744134919923096e-07,
+      "loss": 0.3253,
+      "step": 14990
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 7.067592424361357,
+      "learning_rate": 6.719350854552659e-07,
+      "loss": 0.3245,
+      "step": 14995
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.552954424557108,
+      "learning_rate": 6.694609131808666e-07,
+      "loss": 0.3287,
+      "step": 15000
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.652184851420003,
+      "learning_rate": 6.669909775896605e-07,
+      "loss": 0.322,
+      "step": 15005
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.855649009523,
+      "learning_rate": 6.645252810980519e-07,
+      "loss": 0.3314,
+      "step": 15010
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.9922125059213265,
+      "learning_rate": 6.620638261182998e-07,
+      "loss": 0.3279,
+      "step": 15015
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.399204658913523,
+      "learning_rate": 6.596066150585107e-07,
+      "loss": 0.3338,
+      "step": 15020
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.694809045991095,
+      "learning_rate": 6.571536503226411e-07,
+      "loss": 0.3288,
+      "step": 15025
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.276032527676246,
+      "learning_rate": 6.547049343104916e-07,
+      "loss": 0.342,
+      "step": 15030
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.683686403735051,
+      "learning_rate": 6.522604694177093e-07,
+      "loss": 0.3233,
+      "step": 15035
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.314664074593344,
+      "learning_rate": 6.498202580357788e-07,
+      "loss": 0.3339,
+      "step": 15040
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.271181418939168,
+      "learning_rate": 6.473843025520243e-07,
+      "loss": 0.3413,
+      "step": 15045
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.026745251495655,
+      "learning_rate": 6.449526053496069e-07,
+      "loss": 0.3275,
+      "step": 15050
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.973143978132485,
+      "learning_rate": 6.425251688075212e-07,
+      "loss": 0.3409,
+      "step": 15055
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.734733913002473,
+      "learning_rate": 6.401019953005949e-07,
+      "loss": 0.3299,
+      "step": 15060
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.264578374363782,
+      "learning_rate": 6.376830871994827e-07,
+      "loss": 0.3257,
+      "step": 15065
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 6.804642974562434,
+      "learning_rate": 6.352684468706699e-07,
+      "loss": 0.3347,
+      "step": 15070
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.75835234659942,
+      "learning_rate": 6.328580766764613e-07,
+      "loss": 0.3276,
+      "step": 15075
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.707073653371905,
+      "learning_rate": 6.304519789749907e-07,
+      "loss": 0.3185,
+      "step": 15080
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.670681759892701,
+      "learning_rate": 6.28050156120204e-07,
+      "loss": 0.3323,
+      "step": 15085
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.666206453580839,
+      "learning_rate": 6.256526104618732e-07,
+      "loss": 0.338,
+      "step": 15090
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.61513244052899,
+      "learning_rate": 6.232593443455797e-07,
+      "loss": 0.3252,
+      "step": 15095
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.2786063959854825,
+      "learning_rate": 6.208703601127198e-07,
+      "loss": 0.3361,
+      "step": 15100
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.458489925474581,
+      "learning_rate": 6.184856601005035e-07,
+      "loss": 0.3347,
+      "step": 15105
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.942362062095153,
+      "learning_rate": 6.161052466419449e-07,
+      "loss": 0.3317,
+      "step": 15110
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.201628485037839,
+      "learning_rate": 6.137291220658687e-07,
+      "loss": 0.3323,
+      "step": 15115
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 6.107208916924716,
+      "learning_rate": 6.113572886968994e-07,
+      "loss": 0.3258,
+      "step": 15120
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.377864924772755,
+      "learning_rate": 6.089897488554685e-07,
+      "loss": 0.3304,
+      "step": 15125
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.212508677166943,
+      "learning_rate": 6.066265048578007e-07,
+      "loss": 0.3345,
+      "step": 15130
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.611635469346788,
+      "learning_rate": 6.042675590159241e-07,
+      "loss": 0.3303,
+      "step": 15135
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.553090232084375,
+      "learning_rate": 6.019129136376578e-07,
+      "loss": 0.3343,
+      "step": 15140
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.034051982014956,
+      "learning_rate": 5.99562571026614e-07,
+      "loss": 0.3354,
+      "step": 15145
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.272555573157402,
+      "learning_rate": 5.972165334821983e-07,
+      "loss": 0.3366,
+      "step": 15150
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 6.82950133715513,
+      "learning_rate": 5.94874803299601e-07,
+      "loss": 0.3234,
+      "step": 15155
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 6.959497100574585,
+      "learning_rate": 5.925373827698011e-07,
+      "loss": 0.3163,
+      "step": 15160
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.261785061241256,
+      "learning_rate": 5.902042741795594e-07,
+      "loss": 0.3275,
+      "step": 15165
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.01198603476133,
+      "learning_rate": 5.878754798114189e-07,
+      "loss": 0.3256,
+      "step": 15170
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.558766408563095,
+      "learning_rate": 5.855510019437011e-07,
+      "loss": 0.327,
+      "step": 15175
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.556461907057693,
+      "learning_rate": 5.83230842850508e-07,
+      "loss": 0.3342,
+      "step": 15180
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 8.951444850543265,
+      "learning_rate": 5.809150048017115e-07,
+      "loss": 0.3355,
+      "step": 15185
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.21513520852271,
+      "learning_rate": 5.786034900629584e-07,
+      "loss": 0.3265,
+      "step": 15190
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.822438591492039,
+      "learning_rate": 5.762963008956674e-07,
+      "loss": 0.3196,
+      "step": 15195
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.088036776916842,
+      "learning_rate": 5.739934395570224e-07,
+      "loss": 0.329,
+      "step": 15200
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.32928136338505,
+      "learning_rate": 5.716949082999773e-07,
+      "loss": 0.3319,
+      "step": 15205
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.406320074819915,
+      "learning_rate": 5.694007093732434e-07,
+      "loss": 0.3246,
+      "step": 15210
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.283054897769128,
+      "learning_rate": 5.671108450213009e-07,
+      "loss": 0.3255,
+      "step": 15215
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.293700654067463,
+      "learning_rate": 5.648253174843826e-07,
+      "loss": 0.3305,
+      "step": 15220
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 6.174784098397865,
+      "learning_rate": 5.625441289984851e-07,
+      "loss": 0.309,
+      "step": 15225
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.554691783757481,
+      "learning_rate": 5.602672817953547e-07,
+      "loss": 0.3309,
+      "step": 15230
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 7.632898096662614,
+      "learning_rate": 5.579947781024919e-07,
+      "loss": 0.326,
+      "step": 15235
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 6.394787708139916,
+      "learning_rate": 5.5572662014315e-07,
+      "loss": 0.3218,
+      "step": 15240
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 8.185992925087373,
+      "learning_rate": 5.534628101363287e-07,
+      "loss": 0.3383,
+      "step": 15245
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.80027617699584,
+      "learning_rate": 5.51203350296774e-07,
+      "loss": 0.3289,
+      "step": 15250
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.719033975776043,
+      "learning_rate": 5.489482428349751e-07,
+      "loss": 0.3349,
+      "step": 15255
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.394999674061311,
+      "learning_rate": 5.46697489957167e-07,
+      "loss": 0.3406,
+      "step": 15260
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.365908251012121,
+      "learning_rate": 5.444510938653191e-07,
+      "loss": 0.3295,
+      "step": 15265
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.531697620960981,
+      "learning_rate": 5.422090567571448e-07,
+      "loss": 0.3352,
+      "step": 15270
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.818715998208998,
+      "learning_rate": 5.399713808260871e-07,
+      "loss": 0.332,
+      "step": 15275
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.722096102165326,
+      "learning_rate": 5.377380682613243e-07,
+      "loss": 0.3303,
+      "step": 15280
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.677021215940097,
+      "learning_rate": 5.355091212477693e-07,
+      "loss": 0.3251,
+      "step": 15285
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.33303885654041,
+      "learning_rate": 5.33284541966057e-07,
+      "loss": 0.3324,
+      "step": 15290
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.959350957878013,
+      "learning_rate": 5.310643325925563e-07,
+      "loss": 0.3139,
+      "step": 15295
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 7.711470113011544,
+      "learning_rate": 5.288484952993556e-07,
+      "loss": 0.3306,
+      "step": 15300
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.881691636537076,
+      "learning_rate": 5.266370322542713e-07,
+      "loss": 0.3299,
+      "step": 15305
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.341716160517968,
+      "learning_rate": 5.244299456208341e-07,
+      "loss": 0.3282,
+      "step": 15310
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.6246624806682375,
+      "learning_rate": 5.222272375582993e-07,
+      "loss": 0.3269,
+      "step": 15315
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.538722876182721,
+      "learning_rate": 5.200289102216338e-07,
+      "loss": 0.3291,
+      "step": 15320
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.1239510295431305,
+      "learning_rate": 5.178349657615217e-07,
+      "loss": 0.3253,
+      "step": 15325
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.21239231566736,
+      "learning_rate": 5.156454063243566e-07,
+      "loss": 0.3269,
+      "step": 15330
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 6.697072352286349,
+      "learning_rate": 5.134602340522437e-07,
+      "loss": 0.328,
+      "step": 15335
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 6.146424287472719,
+      "learning_rate": 5.112794510829977e-07,
+      "loss": 0.3259,
+      "step": 15340
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.001634187370734,
+      "learning_rate": 5.091030595501351e-07,
+      "loss": 0.3314,
+      "step": 15345
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.675020804818011,
+      "learning_rate": 5.069310615828804e-07,
+      "loss": 0.3298,
+      "step": 15350
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.884773853738382,
+      "learning_rate": 5.047634593061562e-07,
+      "loss": 0.341,
+      "step": 15355
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.686820504059619,
+      "learning_rate": 5.026002548405878e-07,
+      "loss": 0.3222,
+      "step": 15360
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 6.570517269507593,
+      "learning_rate": 5.004414503024962e-07,
+      "loss": 0.3336,
+      "step": 15365
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.382963602034246,
+      "learning_rate": 4.982870478038976e-07,
+      "loss": 0.3315,
+      "step": 15370
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.8333082592597085,
+      "learning_rate": 4.96137049452502e-07,
+      "loss": 0.3108,
+      "step": 15375
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.703537645888523,
+      "learning_rate": 4.939914573517097e-07,
+      "loss": 0.33,
+      "step": 15380
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.055508354992322,
+      "learning_rate": 4.918502736006136e-07,
+      "loss": 0.3272,
+      "step": 15385
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.140679591165281,
+      "learning_rate": 4.897135002939896e-07,
+      "loss": 0.316,
+      "step": 15390
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.6297789890950165,
+      "learning_rate": 4.875811395223023e-07,
+      "loss": 0.317,
+      "step": 15395
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.887885997760296,
+      "learning_rate": 4.85453193371696e-07,
+      "loss": 0.33,
+      "step": 15400
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.376128690566044,
+      "learning_rate": 4.83329663924001e-07,
+      "loss": 0.3245,
+      "step": 15405
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.8938060055908865,
+      "learning_rate": 4.812105532567191e-07,
+      "loss": 0.3427,
+      "step": 15410
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.11132863259498,
+      "learning_rate": 4.790958634430365e-07,
+      "loss": 0.3317,
+      "step": 15415
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.724591336463593,
+      "learning_rate": 4.7698559655181e-07,
+      "loss": 0.3357,
+      "step": 15420
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.951037262704773,
+      "learning_rate": 4.748797546475703e-07,
+      "loss": 0.3268,
+      "step": 15425
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.8802443891193565,
+      "learning_rate": 4.727783397905211e-07,
+      "loss": 0.3331,
+      "step": 15430
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.373028538851771,
+      "learning_rate": 4.706813540365313e-07,
+      "loss": 0.3267,
+      "step": 15435
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 6.122718060013959,
+      "learning_rate": 4.6858879943713965e-07,
+      "loss": 0.3286,
+      "step": 15440
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.987546920660024,
+      "learning_rate": 4.665006780395492e-07,
+      "loss": 0.3243,
+      "step": 15445
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.851157448444356,
+      "learning_rate": 4.6441699188662424e-07,
+      "loss": 0.3309,
+      "step": 15450
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.222443952757895,
+      "learning_rate": 4.623377430168913e-07,
+      "loss": 0.325,
+      "step": 15455
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.9519127965546392,
+      "learning_rate": 4.6026293346453644e-07,
+      "loss": 0.3299,
+      "step": 15460
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.281245218736686,
+      "learning_rate": 4.581925652594016e-07,
+      "loss": 0.3293,
+      "step": 15465
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.425900579865196,
+      "learning_rate": 4.561266404269826e-07,
+      "loss": 0.3161,
+      "step": 15470
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.723935686904716,
+      "learning_rate": 4.5406516098843166e-07,
+      "loss": 0.326,
+      "step": 15475
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.551245390291361,
+      "learning_rate": 4.5200812896054714e-07,
+      "loss": 0.3285,
+      "step": 15480
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.385884864216109,
+      "learning_rate": 4.49955546355782e-07,
+      "loss": 0.3308,
+      "step": 15485
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.182454178384056,
+      "learning_rate": 4.479074151822299e-07,
+      "loss": 0.337,
+      "step": 15490
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.427294600601199,
+      "learning_rate": 4.458637374436353e-07,
+      "loss": 0.3282,
+      "step": 15495
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.293405692501016,
+      "learning_rate": 4.4382451513938163e-07,
+      "loss": 0.3333,
+      "step": 15500
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.942125262624205,
+      "learning_rate": 4.4178975026449634e-07,
+      "loss": 0.3322,
+      "step": 15505
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.20532593695376,
+      "learning_rate": 4.397594448096448e-07,
+      "loss": 0.3221,
+      "step": 15510
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.226395874757872,
+      "learning_rate": 4.377336007611277e-07,
+      "loss": 0.3227,
+      "step": 15515
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.300060759903124,
+      "learning_rate": 4.357122201008851e-07,
+      "loss": 0.3342,
+      "step": 15520
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.533745505806048,
+      "learning_rate": 4.3369530480648737e-07,
+      "loss": 0.332,
+      "step": 15525
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.335895542748602,
+      "learning_rate": 4.316828568511372e-07,
+      "loss": 0.3218,
+      "step": 15530
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.288348245563574,
+      "learning_rate": 4.296748782036658e-07,
+      "loss": 0.3345,
+      "step": 15535
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.385575084440829,
+      "learning_rate": 4.276713708285346e-07,
+      "loss": 0.3355,
+      "step": 15540
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.16265164571498,
+      "learning_rate": 4.256723366858267e-07,
+      "loss": 0.3326,
+      "step": 15545
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.65564953851326,
+      "learning_rate": 4.236777777312534e-07,
+      "loss": 0.3206,
+      "step": 15550
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.706592827325074,
+      "learning_rate": 4.2168769591614476e-07,
+      "loss": 0.3363,
+      "step": 15555
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.220505773528335,
+      "learning_rate": 4.197020931874507e-07,
+      "loss": 0.3274,
+      "step": 15560
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.213911131499332,
+      "learning_rate": 4.1772097148774173e-07,
+      "loss": 0.3255,
+      "step": 15565
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.282918448441257,
+      "learning_rate": 4.1574433275519963e-07,
+      "loss": 0.3278,
+      "step": 15570
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.404275303777677,
+      "learning_rate": 4.1377217892362653e-07,
+      "loss": 0.3296,
+      "step": 15575
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.876173248789661,
+      "learning_rate": 4.118045119224312e-07,
+      "loss": 0.3306,
+      "step": 15580
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.255389054786569,
+      "learning_rate": 4.0984133367663717e-07,
+      "loss": 0.3213,
+      "step": 15585
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.024576780909882,
+      "learning_rate": 4.07882646106873e-07,
+      "loss": 0.3324,
+      "step": 15590
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.443562631028897,
+      "learning_rate": 4.0592845112937764e-07,
+      "loss": 0.3287,
+      "step": 15595
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.128444756291535,
+      "learning_rate": 4.0397875065599225e-07,
+      "loss": 0.3266,
+      "step": 15600
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.473835014555687,
+      "learning_rate": 4.0203354659415995e-07,
+      "loss": 0.3269,
+      "step": 15605
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.302270282731905,
+      "learning_rate": 4.0009284084692734e-07,
+      "loss": 0.3195,
+      "step": 15610
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.244008416918633,
+      "learning_rate": 3.981566353129385e-07,
+      "loss": 0.3298,
+      "step": 15615
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.511728239678076,
+      "learning_rate": 3.9622493188643695e-07,
+      "loss": 0.3235,
+      "step": 15620
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.7468068473137235,
+      "learning_rate": 3.9429773245725836e-07,
+      "loss": 0.33,
+      "step": 15625
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.532017615851619,
+      "learning_rate": 3.9237503891083604e-07,
+      "loss": 0.3328,
+      "step": 15630
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.9984381654002403,
+      "learning_rate": 3.904568531281905e-07,
+      "loss": 0.3259,
+      "step": 15635
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.369737981053874,
+      "learning_rate": 3.8854317698593713e-07,
+      "loss": 0.3219,
+      "step": 15640
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.891949730049542,
+      "learning_rate": 3.866340123562756e-07,
+      "loss": 0.3275,
+      "step": 15645
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.777212770846676,
+      "learning_rate": 3.8472936110699354e-07,
+      "loss": 0.3311,
+      "step": 15650
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.511407488239449,
+      "learning_rate": 3.8282922510146257e-07,
+      "loss": 0.3275,
+      "step": 15655
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.363462161440987,
+      "learning_rate": 3.8093360619863575e-07,
+      "loss": 0.3244,
+      "step": 15660
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.1267355991414805,
+      "learning_rate": 3.7904250625305006e-07,
+      "loss": 0.3266,
+      "step": 15665
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.521972951609363,
+      "learning_rate": 3.771559271148184e-07,
+      "loss": 0.333,
+      "step": 15670
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.264437267832434,
+      "learning_rate": 3.7527387062963274e-07,
+      "loss": 0.3327,
+      "step": 15675
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.003377017199365,
+      "learning_rate": 3.7339633863875956e-07,
+      "loss": 0.3262,
+      "step": 15680
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.080670004347522,
+      "learning_rate": 3.715233329790391e-07,
+      "loss": 0.3284,
+      "step": 15685
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.331208989720204,
+      "learning_rate": 3.6965485548288217e-07,
+      "loss": 0.3187,
+      "step": 15690
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.463432793538451,
+      "learning_rate": 3.677909079782721e-07,
+      "loss": 0.3214,
+      "step": 15695
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.765134561944942,
+      "learning_rate": 3.6593149228875915e-07,
+      "loss": 0.3294,
+      "step": 15700
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.288035996501678,
+      "learning_rate": 3.640766102334581e-07,
+      "loss": 0.3286,
+      "step": 15705
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.418751421483361,
+      "learning_rate": 3.622262636270518e-07,
+      "loss": 0.3216,
+      "step": 15710
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.867108747967822,
+      "learning_rate": 3.603804542797829e-07,
+      "loss": 0.3272,
+      "step": 15715
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.276748456421316,
+      "learning_rate": 3.5853918399745836e-07,
+      "loss": 0.3212,
+      "step": 15720
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.2089463813160854,
+      "learning_rate": 3.567024545814413e-07,
+      "loss": 0.3262,
+      "step": 15725
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.083727820870298,
+      "learning_rate": 3.548702678286542e-07,
+      "loss": 0.3294,
+      "step": 15730
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.290004487380496,
+      "learning_rate": 3.5304262553157277e-07,
+      "loss": 0.3369,
+      "step": 15735
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 6.074764534438417,
+      "learning_rate": 3.5121952947823166e-07,
+      "loss": 0.3298,
+      "step": 15740
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.5570105710259305,
+      "learning_rate": 3.494009814522137e-07,
+      "loss": 0.3286,
+      "step": 15745
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 6.788132016382686,
+      "learning_rate": 3.475869832326523e-07,
+      "loss": 0.3233,
+      "step": 15750
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.219825053242776,
+      "learning_rate": 3.4577753659423287e-07,
+      "loss": 0.3308,
+      "step": 15755
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.414709813806513,
+      "learning_rate": 3.4397264330718437e-07,
+      "loss": 0.3344,
+      "step": 15760
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.2904459457604265,
+      "learning_rate": 3.421723051372844e-07,
+      "loss": 0.3254,
+      "step": 15765
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.503347436179465,
+      "learning_rate": 3.403765238458495e-07,
+      "loss": 0.3218,
+      "step": 15770
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.221171414137864,
+      "learning_rate": 3.385853011897433e-07,
+      "loss": 0.3319,
+      "step": 15775
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.93086889960561,
+      "learning_rate": 3.367986389213662e-07,
+      "loss": 0.327,
+      "step": 15780
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.106430303910091,
+      "learning_rate": 3.350165387886584e-07,
+      "loss": 0.3215,
+      "step": 15785
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.28094196659452,
+      "learning_rate": 3.3323900253509736e-07,
+      "loss": 0.3275,
+      "step": 15790
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.613510431621033,
+      "learning_rate": 3.314660318996921e-07,
+      "loss": 0.3171,
+      "step": 15795
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.309412604149584,
+      "learning_rate": 3.296976286169906e-07,
+      "loss": 0.3258,
+      "step": 15800
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.364148651082251,
+      "learning_rate": 3.2793379441706854e-07,
+      "loss": 0.333,
+      "step": 15805
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.300943879556085,
+      "learning_rate": 3.261745310255321e-07,
+      "loss": 0.3151,
+      "step": 15810
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.040209421392553,
+      "learning_rate": 3.244198401635157e-07,
+      "loss": 0.3209,
+      "step": 15815
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.794991959613715,
+      "learning_rate": 3.226697235476817e-07,
+      "loss": 0.306,
+      "step": 15820
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.935248036664671,
+      "learning_rate": 3.2092418289021487e-07,
+      "loss": 0.3253,
+      "step": 15825
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.056519209770571,
+      "learning_rate": 3.1918321989882706e-07,
+      "loss": 0.318,
+      "step": 15830
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.269083367129021,
+      "learning_rate": 3.17446836276748e-07,
+      "loss": 0.3314,
+      "step": 15835
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.326862720235322,
+      "learning_rate": 3.157150337227277e-07,
+      "loss": 0.3167,
+      "step": 15840
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.069723672961724,
+      "learning_rate": 3.1398781393103763e-07,
+      "loss": 0.3294,
+      "step": 15845
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.8522878283056725,
+      "learning_rate": 3.1226517859146157e-07,
+      "loss": 0.3302,
+      "step": 15850
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.231814147547882,
+      "learning_rate": 3.105471293893009e-07,
+      "loss": 0.3266,
+      "step": 15855
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.909474166584724,
+      "learning_rate": 3.088336680053688e-07,
+      "loss": 0.3243,
+      "step": 15860
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.235909628911115,
+      "learning_rate": 3.071247961159918e-07,
+      "loss": 0.324,
+      "step": 15865
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.734972101490973,
+      "learning_rate": 3.0542051539300455e-07,
+      "loss": 0.3275,
+      "step": 15870
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.9506257702838843,
+      "learning_rate": 3.037208275037512e-07,
+      "loss": 0.3251,
+      "step": 15875
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.538223913166323,
+      "learning_rate": 3.020257341110827e-07,
+      "loss": 0.3264,
+      "step": 15880
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.358203367211378,
+      "learning_rate": 3.0033523687335363e-07,
+      "loss": 0.3279,
+      "step": 15885
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.231501878147074,
+      "learning_rate": 2.986493374444244e-07,
+      "loss": 0.3239,
+      "step": 15890
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.837316439950837,
+      "learning_rate": 2.969680374736539e-07,
+      "loss": 0.3238,
+      "step": 15895
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.018421603392979,
+      "learning_rate": 2.952913386059053e-07,
+      "loss": 0.318,
+      "step": 15900
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.1435525006903475,
+      "learning_rate": 2.936192424815365e-07,
+      "loss": 0.3175,
+      "step": 15905
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.12970076162308,
+      "learning_rate": 2.919517507364056e-07,
+      "loss": 0.3218,
+      "step": 15910
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.27748140244397,
+      "learning_rate": 2.902888650018648e-07,
+      "loss": 0.3274,
+      "step": 15915
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.885317669031329,
+      "learning_rate": 2.886305869047584e-07,
+      "loss": 0.3287,
+      "step": 15920
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.28260729481659,
+      "learning_rate": 2.86976918067427e-07,
+      "loss": 0.3251,
+      "step": 15925
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.475073429347808,
+      "learning_rate": 2.8532786010769753e-07,
+      "loss": 0.3391,
+      "step": 15930
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.821053704563675,
+      "learning_rate": 2.8368341463888895e-07,
+      "loss": 0.3192,
+      "step": 15935
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.130502760793864,
+      "learning_rate": 2.8204358326980543e-07,
+      "loss": 0.3255,
+      "step": 15940
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.893758489623744,
+      "learning_rate": 2.8040836760473977e-07,
+      "loss": 0.3357,
+      "step": 15945
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.2516531067849845,
+      "learning_rate": 2.7877776924346625e-07,
+      "loss": 0.3132,
+      "step": 15950
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 7.406973652341834,
+      "learning_rate": 2.771517897812437e-07,
+      "loss": 0.3293,
+      "step": 15955
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.322483066999841,
+      "learning_rate": 2.755304308088125e-07,
+      "loss": 0.3166,
+      "step": 15960
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.474306932264857,
+      "learning_rate": 2.7391369391239043e-07,
+      "loss": 0.3288,
+      "step": 15965
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.20725319692026,
+      "learning_rate": 2.723015806736756e-07,
+      "loss": 0.3171,
+      "step": 15970
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.535491629522536,
+      "learning_rate": 2.7069409266984204e-07,
+      "loss": 0.3264,
+      "step": 15975
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.761847179835859,
+      "learning_rate": 2.690912314735383e-07,
+      "loss": 0.3269,
+      "step": 15980
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.109385173462457,
+      "learning_rate": 2.6749299865288626e-07,
+      "loss": 0.3224,
+      "step": 15985
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.879185729585172,
+      "learning_rate": 2.6589939577148115e-07,
+      "loss": 0.3297,
+      "step": 15990
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.056409679988566,
+      "learning_rate": 2.6431042438838707e-07,
+      "loss": 0.322,
+      "step": 15995
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.058301298589732,
+      "learning_rate": 2.6272608605813766e-07,
+      "loss": 0.3372,
+      "step": 16000
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.0829892398529095,
+      "learning_rate": 2.611463823307342e-07,
+      "loss": 0.3256,
+      "step": 16005
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.781469546114506,
+      "learning_rate": 2.595713147516432e-07,
+      "loss": 0.3281,
+      "step": 16010
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.154008309359028,
+      "learning_rate": 2.5800088486179545e-07,
+      "loss": 0.326,
+      "step": 16015
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.4988103210311365,
+      "learning_rate": 2.564350941975852e-07,
+      "loss": 0.3319,
+      "step": 16020
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.230648636823393,
+      "learning_rate": 2.5487394429086764e-07,
+      "loss": 0.328,
+      "step": 16025
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.220354766605433,
+      "learning_rate": 2.5331743666895725e-07,
+      "loss": 0.3251,
+      "step": 16030
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.181949627869047,
+      "learning_rate": 2.517655728546287e-07,
+      "loss": 0.3202,
+      "step": 16035
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.18793309691398,
+      "learning_rate": 2.5021835436611076e-07,
+      "loss": 0.3219,
+      "step": 16040
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.7932237688490047,
+      "learning_rate": 2.486757827170905e-07,
+      "loss": 0.3205,
+      "step": 16045
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.637007418320029,
+      "learning_rate": 2.471378594167062e-07,
+      "loss": 0.3255,
+      "step": 16050
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.274747694027581,
+      "learning_rate": 2.456045859695505e-07,
+      "loss": 0.318,
+      "step": 16055
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.888514536811012,
+      "learning_rate": 2.440759638756651e-07,
+      "loss": 0.3147,
+      "step": 16060
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.070119120643892,
+      "learning_rate": 2.425519946305438e-07,
+      "loss": 0.3255,
+      "step": 16065
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.318133851055312,
+      "learning_rate": 2.4103267972512554e-07,
+      "loss": 0.3202,
+      "step": 16070
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.812137184792789,
+      "learning_rate": 2.39518020645797e-07,
+      "loss": 0.3237,
+      "step": 16075
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.288701026818577,
+      "learning_rate": 2.3800801887439106e-07,
+      "loss": 0.3229,
+      "step": 16080
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.333033708892183,
+      "learning_rate": 2.365026758881822e-07,
+      "loss": 0.3267,
+      "step": 16085
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.423311250484052,
+      "learning_rate": 2.350019931598896e-07,
+      "loss": 0.3196,
+      "step": 16090
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.8143452561896725,
+      "learning_rate": 2.3350597215766878e-07,
+      "loss": 0.3212,
+      "step": 16095
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.9134689439462935,
+      "learning_rate": 2.3201461434512075e-07,
+      "loss": 0.3289,
+      "step": 16100
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.438047747163149,
+      "learning_rate": 2.305279211812783e-07,
+      "loss": 0.3161,
+      "step": 16105
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.079736872305593,
+      "learning_rate": 2.2904589412061528e-07,
+      "loss": 0.3229,
+      "step": 16110
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.72236932999239,
+      "learning_rate": 2.2756853461303797e-07,
+      "loss": 0.33,
+      "step": 16115
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 6.102576174544622,
+      "learning_rate": 2.2609584410388685e-07,
+      "loss": 0.3234,
+      "step": 16120
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 6.018929984105898,
+      "learning_rate": 2.2462782403393557e-07,
+      "loss": 0.3098,
+      "step": 16125
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.505430379967904,
+      "learning_rate": 2.2316447583938694e-07,
+      "loss": 0.3351,
+      "step": 16130
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.084487341521747,
+      "learning_rate": 2.2170580095187466e-07,
+      "loss": 0.3242,
+      "step": 16135
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 6.836398225817673,
+      "learning_rate": 2.2025180079845886e-07,
+      "loss": 0.3267,
+      "step": 16140
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.944041701120263,
+      "learning_rate": 2.1880247680162836e-07,
+      "loss": 0.331,
+      "step": 16145
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.132936779080271,
+      "learning_rate": 2.1735783037929391e-07,
+      "loss": 0.3283,
+      "step": 16150
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 6.9944360578145135,
+      "learning_rate": 2.1591786294479444e-07,
+      "loss": 0.3258,
+      "step": 16155
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.352649165025649,
+      "learning_rate": 2.1448257590688747e-07,
+      "loss": 0.3259,
+      "step": 16160
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.013748515607696,
+      "learning_rate": 2.1305197066975315e-07,
+      "loss": 0.3219,
+      "step": 16165
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.852245030863292,
+      "learning_rate": 2.1162604863299241e-07,
+      "loss": 0.3263,
+      "step": 16170
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.875041183822977,
+      "learning_rate": 2.1020481119162106e-07,
+      "loss": 0.3197,
+      "step": 16175
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.902276796757562,
+      "learning_rate": 2.0878825973607575e-07,
+      "loss": 0.3186,
+      "step": 16180
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.382093906560683,
+      "learning_rate": 2.0737639565220568e-07,
+      "loss": 0.3271,
+      "step": 16185
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 7.1592131547023605,
+      "learning_rate": 2.059692203212771e-07,
+      "loss": 0.3238,
+      "step": 16190
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.699480539733266,
+      "learning_rate": 2.0456673511996705e-07,
+      "loss": 0.3365,
+      "step": 16195
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.58969484654952,
+      "learning_rate": 2.0316894142036303e-07,
+      "loss": 0.3269,
+      "step": 16200
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.709577048665052,
+      "learning_rate": 2.0177584058996667e-07,
+      "loss": 0.3193,
+      "step": 16205
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.366505950527198,
+      "learning_rate": 2.0038743399168504e-07,
+      "loss": 0.3215,
+      "step": 16210
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.957699551759988,
+      "learning_rate": 1.9900372298383374e-07,
+      "loss": 0.323,
+      "step": 16215
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.719601013061237,
+      "learning_rate": 1.976247089201344e-07,
+      "loss": 0.3297,
+      "step": 16220
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 3.935689059330765,
+      "learning_rate": 1.9625039314971394e-07,
+      "loss": 0.3357,
+      "step": 16225
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 6.791720105209688,
+      "learning_rate": 1.9488077701710238e-07,
+      "loss": 0.3336,
+      "step": 16230
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.143417367694795,
+      "learning_rate": 1.9351586186223237e-07,
+      "loss": 0.3238,
+      "step": 16235
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 8.216116273198738,
+      "learning_rate": 1.9215564902043738e-07,
+      "loss": 0.3154,
+      "step": 16240
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.107020221296381,
+      "learning_rate": 1.908001398224496e-07,
+      "loss": 0.3225,
+      "step": 16245
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.588168833266675,
+      "learning_rate": 1.8944933559440105e-07,
+      "loss": 0.3143,
+      "step": 16250
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.79229475914689,
+      "learning_rate": 1.8810323765781956e-07,
+      "loss": 0.3296,
+      "step": 16255
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.36897448726827,
+      "learning_rate": 1.8676184732962953e-07,
+      "loss": 0.3276,
+      "step": 16260
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.567793253702501,
+      "learning_rate": 1.8542516592214788e-07,
+      "loss": 0.3323,
+      "step": 16265
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.50180730085157,
+      "learning_rate": 1.840931947430874e-07,
+      "loss": 0.3236,
+      "step": 16270
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.784739469539132,
+      "learning_rate": 1.8276593509555073e-07,
+      "loss": 0.3254,
+      "step": 16275
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.093679564094168,
+      "learning_rate": 1.8144338827803253e-07,
+      "loss": 0.3212,
+      "step": 16280
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.188116061505996,
+      "learning_rate": 1.8012555558441558e-07,
+      "loss": 0.3268,
+      "step": 16285
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.188884125901071,
+      "learning_rate": 1.7881243830397133e-07,
+      "loss": 0.3254,
+      "step": 16290
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.3028610933926466,
+      "learning_rate": 1.7750403772135716e-07,
+      "loss": 0.3221,
+      "step": 16295
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.748550442739596,
+      "learning_rate": 1.7620035511661748e-07,
+      "loss": 0.3197,
+      "step": 16300
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.34403152927218,
+      "learning_rate": 1.749013917651804e-07,
+      "loss": 0.317,
+      "step": 16305
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.615831309228839,
+      "learning_rate": 1.736071489378549e-07,
+      "loss": 0.3144,
+      "step": 16310
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.699662803614298,
+      "learning_rate": 1.7231762790083594e-07,
+      "loss": 0.3213,
+      "step": 16315
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.770633996070142,
+      "learning_rate": 1.7103282991569548e-07,
+      "loss": 0.3267,
+      "step": 16320
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.955789013505318,
+      "learning_rate": 1.6975275623938637e-07,
+      "loss": 0.3246,
+      "step": 16325
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.521376850056406,
+      "learning_rate": 1.6847740812423962e-07,
+      "loss": 0.3397,
+      "step": 16330
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.013497701337801,
+      "learning_rate": 1.6720678681796165e-07,
+      "loss": 0.3107,
+      "step": 16335
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.053988036087268,
+      "learning_rate": 1.659408935636364e-07,
+      "loss": 0.3183,
+      "step": 16340
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.341405350563486,
+      "learning_rate": 1.6467972959972102e-07,
+      "loss": 0.3217,
+      "step": 16345
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.9646969352659425,
+      "learning_rate": 1.6342329616004683e-07,
+      "loss": 0.3227,
+      "step": 16350
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.1896122643170335,
+      "learning_rate": 1.6217159447381502e-07,
+      "loss": 0.3202,
+      "step": 16355
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.565785467430683,
+      "learning_rate": 1.609246257656011e-07,
+      "loss": 0.3202,
+      "step": 16360
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 6.4492287907440105,
+      "learning_rate": 1.596823912553469e-07,
+      "loss": 0.3148,
+      "step": 16365
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.0934813034293605,
+      "learning_rate": 1.584448921583648e-07,
+      "loss": 0.3248,
+      "step": 16370
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.659745455359364,
+      "learning_rate": 1.5721212968533238e-07,
+      "loss": 0.3165,
+      "step": 16375
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.000420361796629,
+      "learning_rate": 1.5598410504229554e-07,
+      "loss": 0.3213,
+      "step": 16380
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.169494550177258,
+      "learning_rate": 1.5476081943066268e-07,
+      "loss": 0.3222,
+      "step": 16385
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.52313110624559,
+      "learning_rate": 1.5354227404720867e-07,
+      "loss": 0.3212,
+      "step": 16390
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.40933196464466,
+      "learning_rate": 1.523284700840688e-07,
+      "loss": 0.3264,
+      "step": 16395
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.64794423861692,
+      "learning_rate": 1.511194087287393e-07,
+      "loss": 0.3221,
+      "step": 16400
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.240945621396869,
+      "learning_rate": 1.4991509116407842e-07,
+      "loss": 0.3204,
+      "step": 16405
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.319882401349022,
+      "learning_rate": 1.4871551856830259e-07,
+      "loss": 0.332,
+      "step": 16410
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.365765429318108,
+      "learning_rate": 1.475206921149852e-07,
+      "loss": 0.3272,
+      "step": 16415
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.018989346460705,
+      "learning_rate": 1.4633061297305796e-07,
+      "loss": 0.318,
+      "step": 16420
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.323980559831332,
+      "learning_rate": 1.4514528230680726e-07,
+      "loss": 0.3299,
+      "step": 16425
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.19585151471915,
+      "learning_rate": 1.4396470127587382e-07,
+      "loss": 0.3181,
+      "step": 16430
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 6.836398734241736,
+      "learning_rate": 1.4278887103525153e-07,
+      "loss": 0.3227,
+      "step": 16435
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.60566893832175,
+      "learning_rate": 1.4161779273528797e-07,
+      "loss": 0.3225,
+      "step": 16440
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.06213953867403,
+      "learning_rate": 1.4045146752167948e-07,
+      "loss": 0.3299,
+      "step": 16445
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.8865681200680235,
+      "learning_rate": 1.3928989653547498e-07,
+      "loss": 0.3276,
+      "step": 16450
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.753344877185966,
+      "learning_rate": 1.3813308091306876e-07,
+      "loss": 0.3342,
+      "step": 16455
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.38252543437366,
+      "learning_rate": 1.3698102178620664e-07,
+      "loss": 0.3261,
+      "step": 16460
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.443970363526169,
+      "learning_rate": 1.3583372028197704e-07,
+      "loss": 0.326,
+      "step": 16465
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.581110482318744,
+      "learning_rate": 1.3469117752281767e-07,
+      "loss": 0.3259,
+      "step": 16470
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.142538575844111,
+      "learning_rate": 1.335533946265083e-07,
+      "loss": 0.3255,
+      "step": 16475
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.4172771144012435,
+      "learning_rate": 1.3242037270617292e-07,
+      "loss": 0.3406,
+      "step": 16480
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 6.996640252062232,
+      "learning_rate": 1.312921128702771e-07,
+      "loss": 0.3238,
+      "step": 16485
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.147013677924641,
+      "learning_rate": 1.3016861622262788e-07,
+      "loss": 0.3222,
+      "step": 16490
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.969175635638045,
+      "learning_rate": 1.2904988386237272e-07,
+      "loss": 0.3211,
+      "step": 16495
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.304735817041214,
+      "learning_rate": 1.2793591688399665e-07,
+      "loss": 0.3154,
+      "step": 16500
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.417624275374276,
+      "learning_rate": 1.2682671637732512e-07,
+      "loss": 0.3281,
+      "step": 16505
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.3438042106250485,
+      "learning_rate": 1.2572228342751737e-07,
+      "loss": 0.324,
+      "step": 16510
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.222667005638983,
+      "learning_rate": 1.2462261911507124e-07,
+      "loss": 0.3226,
+      "step": 16515
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.666061908982401,
+      "learning_rate": 1.2352772451581784e-07,
+      "loss": 0.318,
+      "step": 16520
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.977480571113075,
+      "learning_rate": 1.2243760070092093e-07,
+      "loss": 0.3197,
+      "step": 16525
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.020839554052547,
+      "learning_rate": 1.21352248736879e-07,
+      "loss": 0.3143,
+      "step": 16530
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.9117798883194865,
+      "learning_rate": 1.2027166968552163e-07,
+      "loss": 0.3185,
+      "step": 16535
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.00269934718407,
+      "learning_rate": 1.19195864604007e-07,
+      "loss": 0.3221,
+      "step": 16540
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.2316602057613455,
+      "learning_rate": 1.1812483454482493e-07,
+      "loss": 0.3251,
+      "step": 16545
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.958986638942587,
+      "learning_rate": 1.1705858055579389e-07,
+      "loss": 0.316,
+      "step": 16550
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.43777934051335,
+      "learning_rate": 1.1599710368005723e-07,
+      "loss": 0.3128,
+      "step": 16555
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.872643950834902,
+      "learning_rate": 1.149404049560876e-07,
+      "loss": 0.3172,
+      "step": 16560
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.0242774912669885,
+      "learning_rate": 1.1388848541768193e-07,
+      "loss": 0.3302,
+      "step": 16565
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.6394760087647064,
+      "learning_rate": 1.1284134609396091e-07,
+      "loss": 0.3134,
+      "step": 16570
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.20248467236056,
+      "learning_rate": 1.117989880093695e-07,
+      "loss": 0.3242,
+      "step": 16575
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.993861326186423,
+      "learning_rate": 1.107614121836742e-07,
+      "loss": 0.3232,
+      "step": 16580
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.140204526928821,
+      "learning_rate": 1.0972861963196469e-07,
+      "loss": 0.3131,
+      "step": 16585
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 6.362294713393221,
+      "learning_rate": 1.0870061136464772e-07,
+      "loss": 0.3183,
+      "step": 16590
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 6.061443962737918,
+      "learning_rate": 1.0767738838745379e-07,
+      "loss": 0.32,
+      "step": 16595
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.021965056921146,
+      "learning_rate": 1.066589517014277e-07,
+      "loss": 0.3165,
+      "step": 16600
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.388055243217996,
+      "learning_rate": 1.0564530230293468e-07,
+      "loss": 0.3165,
+      "step": 16605
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.35685916039647,
+      "learning_rate": 1.0463644118365535e-07,
+      "loss": 0.3169,
+      "step": 16610
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.082810603055631,
+      "learning_rate": 1.0363236933058462e-07,
+      "loss": 0.3223,
+      "step": 16615
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.12496359069897,
+      "learning_rate": 1.0263308772603397e-07,
+      "loss": 0.3259,
+      "step": 16620
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.348108139298327,
+      "learning_rate": 1.0163859734762749e-07,
+      "loss": 0.3187,
+      "step": 16625
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.1193485171849975,
+      "learning_rate": 1.006488991683019e-07,
+      "loss": 0.3202,
+      "step": 16630
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.8436774397202775,
+      "learning_rate": 9.96639941563049e-08,
+      "loss": 0.3316,
+      "step": 16635
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.415829203215659,
+      "learning_rate": 9.868388327519685e-08,
+      "loss": 0.3238,
+      "step": 16640
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.197213779335068,
+      "learning_rate": 9.770856748384516e-08,
+      "loss": 0.3248,
+      "step": 16645
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.196155455714435,
+      "learning_rate": 9.673804773642936e-08,
+      "loss": 0.3249,
+      "step": 16650
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 7.855173996053146,
+      "learning_rate": 9.577232498243383e-08,
+      "loss": 0.3245,
+      "step": 16655
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.120386816838393,
+      "learning_rate": 9.481140016665169e-08,
+      "loss": 0.3151,
+      "step": 16660
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 6.56032471761875,
+      "learning_rate": 9.385527422918095e-08,
+      "loss": 0.3141,
+      "step": 16665
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.838102733097684,
+      "learning_rate": 9.290394810542669e-08,
+      "loss": 0.319,
+      "step": 16670
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.237242674037893,
+      "learning_rate": 9.195742272609609e-08,
+      "loss": 0.3299,
+      "step": 16675
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.069975175557291,
+      "learning_rate": 9.101569901719953e-08,
+      "loss": 0.3269,
+      "step": 16680
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.70395556339215,
+      "learning_rate": 9.007877790005281e-08,
+      "loss": 0.3216,
+      "step": 16685
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.484240354831281,
+      "learning_rate": 8.91466602912694e-08,
+      "loss": 0.3258,
+      "step": 16690
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.619820251785469,
+      "learning_rate": 8.821934710276648e-08,
+      "loss": 0.3264,
+      "step": 16695
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.306695765403448,
+      "learning_rate": 8.72968392417578e-08,
+      "loss": 0.3186,
+      "step": 16700
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.678091335257884,
+      "learning_rate": 8.637913761075922e-08,
+      "loss": 0.3136,
+      "step": 16705
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.140977332989994,
+      "learning_rate": 8.546624310758256e-08,
+      "loss": 0.319,
+      "step": 16710
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.245543097098709,
+      "learning_rate": 8.455815662533617e-08,
+      "loss": 0.3191,
+      "step": 16715
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.2871480871847005,
+      "learning_rate": 8.36548790524272e-08,
+      "loss": 0.3154,
+      "step": 16720
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.821588086606466,
+      "learning_rate": 8.27564112725554e-08,
+      "loss": 0.3282,
+      "step": 16725
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.331167692980463,
+      "learning_rate": 8.186275416471656e-08,
+      "loss": 0.3108,
+      "step": 16730
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.137864192771517,
+      "learning_rate": 8.097390860319909e-08,
+      "loss": 0.3191,
+      "step": 16735
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.561135191481652,
+      "learning_rate": 8.008987545758518e-08,
+      "loss": 0.3168,
+      "step": 16740
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.209127320403764,
+      "learning_rate": 7.921065559274688e-08,
+      "loss": 0.3311,
+      "step": 16745
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.808937295788496,
+      "learning_rate": 7.833624986885058e-08,
+      "loss": 0.3204,
+      "step": 16750
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.061855686895446,
+      "learning_rate": 7.746665914134977e-08,
+      "loss": 0.3224,
+      "step": 16755
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.742392831215071,
+      "learning_rate": 7.66018842609889e-08,
+      "loss": 0.3245,
+      "step": 16760
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.374524764337193,
+      "learning_rate": 7.574192607380071e-08,
+      "loss": 0.3318,
+      "step": 16765
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.019406843271415,
+      "learning_rate": 7.488678542110495e-08,
+      "loss": 0.3212,
+      "step": 16770
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.926694113312759,
+      "learning_rate": 7.403646313950962e-08,
+      "loss": 0.3328,
+      "step": 16775
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.566359108289269,
+      "learning_rate": 7.319096006090654e-08,
+      "loss": 0.3108,
+      "step": 16780
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.408310731840014,
+      "learning_rate": 7.235027701247621e-08,
+      "loss": 0.3212,
+      "step": 16785
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.152290333016591,
+      "learning_rate": 7.151441481667965e-08,
+      "loss": 0.3355,
+      "step": 16790
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.565863091344188,
+      "learning_rate": 7.068337429126437e-08,
+      "loss": 0.3247,
+      "step": 16795
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.721947965138385,
+      "learning_rate": 6.985715624925948e-08,
+      "loss": 0.3219,
+      "step": 16800
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.5162074092895645,
+      "learning_rate": 6.903576149897617e-08,
+      "loss": 0.3204,
+      "step": 16805
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.236228619641246,
+      "learning_rate": 6.821919084400774e-08,
+      "loss": 0.3308,
+      "step": 16810
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.141813607931332,
+      "learning_rate": 6.740744508322683e-08,
+      "loss": 0.3135,
+      "step": 16815
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.665322840238465,
+      "learning_rate": 6.660052501078596e-08,
+      "loss": 0.3252,
+      "step": 16820
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.767201979259087,
+      "learning_rate": 6.579843141611697e-08,
+      "loss": 0.3273,
+      "step": 16825
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.165531389035974,
+      "learning_rate": 6.50011650839305e-08,
+      "loss": 0.3223,
+      "step": 16830
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.192803078576282,
+      "learning_rate": 6.420872679421208e-08,
+      "loss": 0.3151,
+      "step": 16835
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.385516489304573,
+      "learning_rate": 6.342111732222655e-08,
+      "loss": 0.3328,
+      "step": 16840
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.425070913092375,
+      "learning_rate": 6.263833743851367e-08,
+      "loss": 0.3204,
+      "step": 16845
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.471432430744481,
+      "learning_rate": 6.186038790888749e-08,
+      "loss": 0.3248,
+      "step": 16850
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.3856903583224875,
+      "learning_rate": 6.108726949443756e-08,
+      "loss": 0.3201,
+      "step": 16855
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.589501882050007,
+      "learning_rate": 6.031898295152605e-08,
+      "loss": 0.3224,
+      "step": 16860
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.674711932036349,
+      "learning_rate": 5.955552903178896e-08,
+      "loss": 0.3267,
+      "step": 16865
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.10918736978263,
+      "learning_rate": 5.8796908482132706e-08,
+      "loss": 0.3152,
+      "step": 16870
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.9623222898979105,
+      "learning_rate": 5.80431220447375e-08,
+      "loss": 0.3255,
+      "step": 16875
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.83341224963951,
+      "learning_rate": 5.7294170457052326e-08,
+      "loss": 0.3129,
+      "step": 16880
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.207285877907312,
+      "learning_rate": 5.655005445179662e-08,
+      "loss": 0.3302,
+      "step": 16885
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.267072915271972,
+      "learning_rate": 5.581077475695973e-08,
+      "loss": 0.3282,
+      "step": 16890
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.213948223184893,
+      "learning_rate": 5.50763320957981e-08,
+      "loss": 0.328,
+      "step": 16895
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.630031411426699,
+      "learning_rate": 5.4346727186837534e-08,
+      "loss": 0.3341,
+      "step": 16900
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 6.480499974246501,
+      "learning_rate": 5.362196074386983e-08,
+      "loss": 0.3305,
+      "step": 16905
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.359901605698903,
+      "learning_rate": 5.290203347595335e-08,
+      "loss": 0.3232,
+      "step": 16910
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.488707476211648,
+      "learning_rate": 5.218694608741304e-08,
+      "loss": 0.3237,
+      "step": 16915
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.183021393567609,
+      "learning_rate": 5.1476699277837605e-08,
+      "loss": 0.3186,
+      "step": 16920
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.804647675972266,
+      "learning_rate": 5.077129374208012e-08,
+      "loss": 0.3257,
+      "step": 16925
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.494163540567883,
+      "learning_rate": 5.007073017025965e-08,
+      "loss": 0.3314,
+      "step": 16930
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.1840909862780515,
+      "learning_rate": 4.9375009247754626e-08,
+      "loss": 0.3279,
+      "step": 16935
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.572678380765821,
+      "learning_rate": 4.8684131655208353e-08,
+      "loss": 0.3116,
+      "step": 16940
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.196403827581241,
+      "learning_rate": 4.799809806852518e-08,
+      "loss": 0.3184,
+      "step": 16945
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.357098767447156,
+      "learning_rate": 4.7316909158869884e-08,
+      "loss": 0.3297,
+      "step": 16950
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.343916649756927,
+      "learning_rate": 4.6640565592668276e-08,
+      "loss": 0.311,
+      "step": 16955
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.291618378770368,
+      "learning_rate": 4.5969068031604945e-08,
+      "loss": 0.329,
+      "step": 16960
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.25102189320903,
+      "learning_rate": 4.530241713262495e-08,
+      "loss": 0.3209,
+      "step": 16965
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.134754275534955,
+      "learning_rate": 4.4640613547929925e-08,
+      "loss": 0.3101,
+      "step": 16970
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.545599286611421,
+      "learning_rate": 4.398365792498083e-08,
+      "loss": 0.3242,
+      "step": 16975
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.196296248690342,
+      "learning_rate": 4.3331550906494656e-08,
+      "loss": 0.3239,
+      "step": 16980
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.364358485774829,
+      "learning_rate": 4.268429313044553e-08,
+      "loss": 0.3226,
+      "step": 16985
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.446202143056948,
+      "learning_rate": 4.204188523006303e-08,
+      "loss": 0.3247,
+      "step": 16990
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.79378156017565,
+      "learning_rate": 4.140432783383219e-08,
+      "loss": 0.3221,
+      "step": 16995
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.010838753010564,
+      "learning_rate": 4.077162156549297e-08,
+      "loss": 0.3324,
+      "step": 17000
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.9813207535154596,
+      "learning_rate": 4.0143767044038554e-08,
+      "loss": 0.3231,
+      "step": 17005
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.901540079957885,
+      "learning_rate": 3.9520764883715924e-08,
+      "loss": 0.3275,
+      "step": 17010
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.208544291007741,
+      "learning_rate": 3.8902615694025313e-08,
+      "loss": 0.3186,
+      "step": 17015
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.042927808490137,
+      "learning_rate": 3.828932007971797e-08,
+      "loss": 0.3151,
+      "step": 17020
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.514916487809097,
+      "learning_rate": 3.768087864079839e-08,
+      "loss": 0.3234,
+      "step": 17025
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.432387190050278,
+      "learning_rate": 3.707729197252097e-08,
+      "loss": 0.3174,
+      "step": 17030
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 6.9894227760113425,
+      "learning_rate": 3.6478560665390574e-08,
+      "loss": 0.33,
+      "step": 17035
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.295747928089492,
+      "learning_rate": 3.588468530516198e-08,
+      "loss": 0.3266,
+      "step": 17040
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.407466468401592,
+      "learning_rate": 3.529566647284044e-08,
+      "loss": 0.3257,
+      "step": 17045
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.747434671821218,
+      "learning_rate": 3.471150474467777e-08,
+      "loss": 0.3323,
+      "step": 17050
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 6.148672477395743,
+      "learning_rate": 3.413220069217627e-08,
+      "loss": 0.3232,
+      "step": 17055
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.212980371901559,
+      "learning_rate": 3.355775488208368e-08,
+      "loss": 0.3205,
+      "step": 17060
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.595464952249602,
+      "learning_rate": 3.298816787639714e-08,
+      "loss": 0.3308,
+      "step": 17065
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.261453414240882,
+      "learning_rate": 3.242344023235755e-08,
+      "loss": 0.3235,
+      "step": 17070
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.140376013153457,
+      "learning_rate": 3.186357250245409e-08,
+      "loss": 0.3192,
+      "step": 17075
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.201478258105787,
+      "learning_rate": 3.1308565234420275e-08,
+      "loss": 0.3211,
+      "step": 17080
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.146822624316797,
+      "learning_rate": 3.0758418971233995e-08,
+      "loss": 0.3154,
+      "step": 17085
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.174234028597232,
+      "learning_rate": 3.0213134251119716e-08,
+      "loss": 0.3183,
+      "step": 17090
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.298240807790573,
+      "learning_rate": 2.967271160754237e-08,
+      "loss": 0.3268,
+      "step": 17095
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.227795333228735,
+      "learning_rate": 2.9137151569213486e-08,
+      "loss": 0.3214,
+      "step": 17100
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.210920820302368,
+      "learning_rate": 2.8606454660085047e-08,
+      "loss": 0.3201,
+      "step": 17105
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.688694831097193,
+      "learning_rate": 2.8080621399352857e-08,
+      "loss": 0.3274,
+      "step": 17110
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.016657299349907,
+      "learning_rate": 2.7559652301452632e-08,
+      "loss": 0.3165,
+      "step": 17115
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.427423155416518,
+      "learning_rate": 2.704354787606389e-08,
+      "loss": 0.3265,
+      "step": 17120
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.048773944525423,
+      "learning_rate": 2.653230862810441e-08,
+      "loss": 0.316,
+      "step": 17125
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.363179533603372,
+      "learning_rate": 2.602593505773354e-08,
+      "loss": 0.3294,
+      "step": 17130
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.343244247482888,
+      "learning_rate": 2.5524427660351125e-08,
+      "loss": 0.3286,
+      "step": 17135
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.1011993397027995,
+      "learning_rate": 2.5027786926594132e-08,
+      "loss": 0.3294,
+      "step": 17140
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.095024825445343,
+      "learning_rate": 2.453601334234057e-08,
+      "loss": 0.3184,
+      "step": 17145
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.533694904679112,
+      "learning_rate": 2.4049107388705028e-08,
+      "loss": 0.3165,
+      "step": 17150
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.338434029875889,
+      "learning_rate": 2.356706954204091e-08,
+      "loss": 0.3194,
+      "step": 17155
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.59201757614292,
+      "learning_rate": 2.3089900273938758e-08,
+      "loss": 0.3288,
+      "step": 17160
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.1761127099824575,
+      "learning_rate": 2.2617600051226818e-08,
+      "loss": 0.3285,
+      "step": 17165
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.518604552105859,
+      "learning_rate": 2.2150169335968807e-08,
+      "loss": 0.322,
+      "step": 17170
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.641904310325613,
+      "learning_rate": 2.168760858546448e-08,
+      "loss": 0.3237,
+      "step": 17175
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.752248201594694,
+      "learning_rate": 2.1229918252249627e-08,
+      "loss": 0.3242,
+      "step": 17180
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.765539723924974,
+      "learning_rate": 2.0777098784095507e-08,
+      "loss": 0.3281,
+      "step": 17185
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.333491188039683,
+      "learning_rate": 2.0329150624006645e-08,
+      "loss": 0.3343,
+      "step": 17190
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.9243081578917645,
+      "learning_rate": 1.9886074210223592e-08,
+      "loss": 0.3285,
+      "step": 17195
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.142304534229642,
+      "learning_rate": 1.9447869976220167e-08,
+      "loss": 0.3264,
+      "step": 17200
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.05735236631641,
+      "learning_rate": 1.901453835070233e-08,
+      "loss": 0.3115,
+      "step": 17205
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 3.95977328598609,
+      "learning_rate": 1.858607975761095e-08,
+      "loss": 0.3274,
+      "step": 17210
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.0736743171757075,
+      "learning_rate": 1.816249461611852e-08,
+      "loss": 0.3267,
+      "step": 17215
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.349859105884953,
+      "learning_rate": 1.774378334062965e-08,
+      "loss": 0.3272,
+      "step": 17220
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.296848237552809,
+      "learning_rate": 1.732994634078111e-08,
+      "loss": 0.3187,
+      "step": 17225
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.696438474912808,
+      "learning_rate": 1.692098402144071e-08,
+      "loss": 0.331,
+      "step": 17230
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.615718612464431,
+      "learning_rate": 1.6516896782706736e-08,
+      "loss": 0.3345,
+      "step": 17235
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.568930088531357,
+      "learning_rate": 1.6117685019909623e-08,
+      "loss": 0.322,
+      "step": 17240
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.404550793819637,
+      "learning_rate": 1.5723349123608067e-08,
+      "loss": 0.3163,
+      "step": 17245
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.629118831077295,
+      "learning_rate": 1.5333889479592356e-08,
+      "loss": 0.3237,
+      "step": 17250
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.603765706275918,
+      "learning_rate": 1.4949306468880486e-08,
+      "loss": 0.3176,
+      "step": 17255
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.365997654720079,
+      "learning_rate": 1.456960046772149e-08,
+      "loss": 0.3255,
+      "step": 17260
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.130421217693394,
+      "learning_rate": 1.4194771847590994e-08,
+      "loss": 0.3187,
+      "step": 17265
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.305110035082924,
+      "learning_rate": 1.3824820975194553e-08,
+      "loss": 0.3094,
+      "step": 17270
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.422436622511799,
+      "learning_rate": 1.3459748212464318e-08,
+      "loss": 0.3225,
+      "step": 17275
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.431422170606385,
+      "learning_rate": 1.3099553916561813e-08,
+      "loss": 0.3287,
+      "step": 17280
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.241567806730011,
+      "learning_rate": 1.2744238439874046e-08,
+      "loss": 0.3261,
+      "step": 17285
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.461893699555121,
+      "learning_rate": 1.239380213001684e-08,
+      "loss": 0.3182,
+      "step": 17290
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.755366473966939,
+      "learning_rate": 1.2048245329829844e-08,
+      "loss": 0.3261,
+      "step": 17295
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.086374403285044,
+      "learning_rate": 1.1707568377382072e-08,
+      "loss": 0.3237,
+      "step": 17300
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.37192574562722,
+      "learning_rate": 1.137177160596581e-08,
+      "loss": 0.3209,
+      "step": 17305
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.113981142802513,
+      "learning_rate": 1.1040855344101043e-08,
+      "loss": 0.3163,
+      "step": 17310
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.4901520195537765,
+      "learning_rate": 1.0714819915531582e-08,
+      "loss": 0.3317,
+      "step": 17315
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.06515915220569,
+      "learning_rate": 1.0393665639226724e-08,
+      "loss": 0.3286,
+      "step": 17320
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.305528353338569,
+      "learning_rate": 1.007739282938014e-08,
+      "loss": 0.3205,
+      "step": 17325
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.283052479749407,
+      "learning_rate": 9.766001795410984e-09,
+      "loss": 0.3217,
+      "step": 17330
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.007913866226925,
+      "learning_rate": 9.459492841960572e-09,
+      "loss": 0.319,
+      "step": 17335
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.892225816928856,
+      "learning_rate": 9.157866268895144e-09,
+      "loss": 0.3195,
+      "step": 17340
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.2700543147223975,
+      "learning_rate": 8.861122371303654e-09,
+      "loss": 0.3251,
+      "step": 17345
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.746315215212394,
+      "learning_rate": 8.569261439499432e-09,
+      "loss": 0.3291,
+      "step": 17350
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.227497950738878,
+      "learning_rate": 8.282283759017962e-09,
+      "loss": 0.3251,
+      "step": 17355
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.526948346687582,
+      "learning_rate": 8.000189610616883e-09,
+      "loss": 0.3167,
+      "step": 17360
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.397402173863833,
+      "learning_rate": 7.722979270275988e-09,
+      "loss": 0.3139,
+      "step": 17365
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.019686407837929,
+      "learning_rate": 7.450653009198338e-09,
+      "loss": 0.3203,
+      "step": 17370
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.113894086863957,
+      "learning_rate": 7.18321109380804e-09,
+      "loss": 0.3153,
+      "step": 17375
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.9802828861675468,
+      "learning_rate": 6.920653785750797e-09,
+      "loss": 0.3123,
+      "step": 17380
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 7.397341279147858,
+      "learning_rate": 6.662981341892805e-09,
+      "loss": 0.3222,
+      "step": 17385
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.2876178038487325,
+      "learning_rate": 6.410194014322413e-09,
+      "loss": 0.3254,
+      "step": 17390
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 6.362407584057065,
+      "learning_rate": 6.162292050348462e-09,
+      "loss": 0.3215,
+      "step": 17395
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.260197573800328,
+      "learning_rate": 5.919275692500281e-09,
+      "loss": 0.3342,
+      "step": 17400
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.652129288543425,
+      "learning_rate": 5.681145178526581e-09,
+      "loss": 0.3266,
+      "step": 17405
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.0329427262401225,
+      "learning_rate": 5.4479007413976715e-09,
+      "loss": 0.3261,
+      "step": 17410
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.212095217891655,
+      "learning_rate": 5.21954260930213e-09,
+      "loss": 0.3189,
+      "step": 17415
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.388186784150905,
+      "learning_rate": 4.996071005649583e-09,
+      "loss": 0.3258,
+      "step": 17420
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.403577402430027,
+      "learning_rate": 4.777486149067923e-09,
+      "loss": 0.3229,
+      "step": 17425
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.372666014770122,
+      "learning_rate": 4.563788253404422e-09,
+      "loss": 0.3168,
+      "step": 17430
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.920216648474848,
+      "learning_rate": 4.3549775277262895e-09,
+      "loss": 0.3219,
+      "step": 17435
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.75224480939319,
+      "learning_rate": 4.151054176317337e-09,
+      "loss": 0.3224,
+      "step": 17440
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.17217886620779,
+      "learning_rate": 3.9520183986829776e-09,
+      "loss": 0.3291,
+      "step": 17445
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.521520496183765,
+      "learning_rate": 3.757870389544116e-09,
+      "loss": 0.3193,
+      "step": 17450
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.9911833599850017,
+      "learning_rate": 3.5686103388410385e-09,
+      "loss": 0.3261,
+      "step": 17455
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.8047227129063055,
+      "learning_rate": 3.384238431732301e-09,
+      "loss": 0.3251,
+      "step": 17460
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.3984464082370645,
+      "learning_rate": 3.2047548485941714e-09,
+      "loss": 0.3188,
+      "step": 17465
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.579762521361893,
+      "learning_rate": 3.0301597650195247e-09,
+      "loss": 0.3259,
+      "step": 17470
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.880006797112499,
+      "learning_rate": 2.8604533518200585e-09,
+      "loss": 0.3245,
+      "step": 17475
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.29310428512058,
+      "learning_rate": 2.6956357750235195e-09,
+      "loss": 0.3288,
+      "step": 17480
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.429366982118829,
+      "learning_rate": 2.53570719587648e-09,
+      "loss": 0.3261,
+      "step": 17485
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.47672322500388,
+      "learning_rate": 2.3806677708398952e-09,
+      "loss": 0.326,
+      "step": 17490
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.315788135319219,
+      "learning_rate": 2.230517651594655e-09,
+      "loss": 0.3145,
+      "step": 17495
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.145232204854416,
+      "learning_rate": 2.0852569850354778e-09,
+      "loss": 0.3167,
+      "step": 17500
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.61449092787545,
+      "learning_rate": 1.9448859132747965e-09,
+      "loss": 0.3093,
+      "step": 17505
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 6.349955321675779,
+      "learning_rate": 1.809404573642204e-09,
+      "loss": 0.3289,
+      "step": 17510
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.081659098663104,
+      "learning_rate": 1.6788130986816754e-09,
+      "loss": 0.3171,
+      "step": 17515
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.59987534077367,
+      "learning_rate": 1.553111616155456e-09,
+      "loss": 0.3243,
+      "step": 17520
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.241357555063181,
+      "learning_rate": 1.432300249040175e-09,
+      "loss": 0.3163,
+      "step": 17525
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.650346457651398,
+      "learning_rate": 1.316379115529065e-09,
+      "loss": 0.3216,
+      "step": 17530
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.413006568192071,
+      "learning_rate": 1.2053483290308533e-09,
+      "loss": 0.3165,
+      "step": 17535
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.247458222002462,
+      "learning_rate": 1.099207998169205e-09,
+      "loss": 0.3255,
+      "step": 17540
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.332420786778957,
+      "learning_rate": 9.979582267855004e-10,
+      "loss": 0.3211,
+      "step": 17545
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.774136612916247,
+      "learning_rate": 9.015991139338376e-10,
+      "loss": 0.3266,
+      "step": 17550
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.276016406258255,
+      "learning_rate": 8.101307538854741e-10,
+      "loss": 0.3287,
+      "step": 17555
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.9933204057622826,
+      "learning_rate": 7.235532361266062e-10,
+      "loss": 0.3267,
+      "step": 17560
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.015434379642638,
+      "learning_rate": 6.418666453578137e-10,
+      "loss": 0.3279,
+      "step": 17565
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.3100411946404105,
+      "learning_rate": 5.650710614957255e-10,
+      "loss": 0.3261,
+      "step": 17570
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.620246737157527,
+      "learning_rate": 4.931665596713542e-10,
+      "loss": 0.3089,
+      "step": 17575
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.131919436844451,
+      "learning_rate": 4.2615321023065094e-10,
+      "loss": 0.329,
+      "step": 17580
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.213045629785342,
+      "learning_rate": 3.6403107873450583e-10,
+      "loss": 0.3256,
+      "step": 17585
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.629428548039462,
+      "learning_rate": 3.068002259593028e-10,
+      "loss": 0.3195,
+      "step": 17590
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.332419507807568,
+      "learning_rate": 2.5446070789525437e-10,
+      "loss": 0.3204,
+      "step": 17595
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.240834030496161,
+      "learning_rate": 2.0701257574751165e-10,
+      "loss": 0.3203,
+      "step": 17600
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.572494727034468,
+      "learning_rate": 1.6445587593560964e-10,
+      "loss": 0.327,
+      "step": 17605
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.2630889633032805,
+      "learning_rate": 1.267906500940219e-10,
+      "loss": 0.326,
+      "step": 17610
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.181006413141019,
+      "learning_rate": 9.40169350716058e-11,
+      "loss": 0.3255,
+      "step": 17615
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.117134318017797,
+      "learning_rate": 6.613476293160226e-11,
+      "loss": 0.3272,
+      "step": 17620
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.00361578124429,
+      "learning_rate": 4.3144160952746096e-11,
+      "loss": 0.334,
+      "step": 17625
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.3128277913830155,
+      "learning_rate": 2.5045151626490462e-11,
+      "loss": 0.3257,
+      "step": 17630
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.347718473004982,
+      "learning_rate": 1.1837752659782375e-11,
+      "loss": 0.3184,
+      "step": 17635
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.857624199795862,
+      "learning_rate": 3.5219769745076237e-12,
+      "loss": 0.3295,
+      "step": 17640
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.70428450020991,
+      "learning_rate": 9.783270471519502e-14,
+      "loss": 0.3309,
+      "step": 17645
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.273702472448349,
+      "eval_runtime": 75.049,
+      "eval_samples_per_second": 4.824,
+      "eval_steps_per_second": 0.613,
+      "step": 17646
+    },
+    {
+      "epoch": 3.0,
+      "step": 17646,
+      "total_flos": 607478932832256.0,
+      "train_loss": 0.7695021375204772,
+      "train_runtime": 155405.9935,
+      "train_samples_per_second": 1.817,
+      "train_steps_per_second": 0.114
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 17646,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 607478932832256.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}