diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,74597 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 53244,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1659.2270556222848,
+      "learning_rate": 1.8779342723004697e-09,
+      "loss": 16.4902,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1882.7565013569613,
+      "learning_rate": 9.389671361502349e-09,
+      "loss": 17.2082,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1713.0057743789032,
+      "learning_rate": 1.8779342723004698e-08,
+      "loss": 17.0727,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1843.3418373795246,
+      "learning_rate": 2.8169014084507045e-08,
+      "loss": 16.8226,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1813.3767547899452,
+      "learning_rate": 3.7558685446009395e-08,
+      "loss": 16.9255,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1678.219429048779,
+      "learning_rate": 4.694835680751174e-08,
+      "loss": 16.452,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1636.6897483112282,
+      "learning_rate": 5.633802816901409e-08,
+      "loss": 15.6666,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1318.3277535950876,
+      "learning_rate": 6.572769953051644e-08,
+      "loss": 15.4184,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1441.7068876937751,
+      "learning_rate": 7.511737089201879e-08,
+      "loss": 14.4912,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1292.3376716271725,
+      "learning_rate": 8.450704225352113e-08,
+      "loss": 13.7669,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1025.333612889984,
+      "learning_rate": 9.389671361502348e-08,
+      "loss": 12.0966,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1048.3141228926484,
+      "learning_rate": 1.0328638497652583e-07,
+      "loss": 11.3846,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 953.3406257911793,
+      "learning_rate": 1.1267605633802818e-07,
+      "loss": 10.3368,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 704.4452333942636,
+      "learning_rate": 1.2206572769953052e-07,
+      "loss": 9.1991,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 676.9230725364876,
+      "learning_rate": 1.3145539906103288e-07,
+      "loss": 8.0264,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 526.1271016517904,
+      "learning_rate": 1.4084507042253522e-07,
+      "loss": 7.0255,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 443.4611422903287,
+      "learning_rate": 1.5023474178403758e-07,
+      "loss": 6.3881,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 370.18793921907417,
+      "learning_rate": 1.5962441314553992e-07,
+      "loss": 5.8924,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 335.7407410048187,
+      "learning_rate": 1.6901408450704225e-07,
+      "loss": 5.5044,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 299.98221531514986,
+      "learning_rate": 1.7840375586854462e-07,
+      "loss": 5.195,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 236.48950000745847,
+      "learning_rate": 1.8779342723004696e-07,
+      "loss": 4.8934,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 230.23667875404382,
+      "learning_rate": 1.9718309859154932e-07,
+      "loss": 4.6712,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 281.23701010806394,
+      "learning_rate": 2.0657276995305166e-07,
+      "loss": 4.5184,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 193.95287916413935,
+      "learning_rate": 2.1596244131455402e-07,
+      "loss": 4.1266,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 185.3029747883216,
+      "learning_rate": 2.2535211267605636e-07,
+      "loss": 3.8772,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 173.66612187024688,
+      "learning_rate": 2.347417840375587e-07,
+      "loss": 3.7705,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 172.94071057332368,
+      "learning_rate": 2.4413145539906103e-07,
+      "loss": 3.4679,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 194.94757041376008,
+      "learning_rate": 2.535211267605634e-07,
+      "loss": 3.2841,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 166.56693420215467,
+      "learning_rate": 2.6291079812206576e-07,
+      "loss": 3.0015,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 143.23989029889705,
+      "learning_rate": 2.7230046948356807e-07,
+      "loss": 2.8706,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 107.66554276610727,
+      "learning_rate": 2.8169014084507043e-07,
+      "loss": 2.7523,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 98.96870952867148,
+      "learning_rate": 2.910798122065728e-07,
+      "loss": 2.4828,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 93.48328458386071,
+      "learning_rate": 3.0046948356807516e-07,
+      "loss": 2.5108,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 84.45473295814509,
+      "learning_rate": 3.0985915492957747e-07,
+      "loss": 2.382,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 72.72398209693705,
+      "learning_rate": 3.1924882629107983e-07,
+      "loss": 2.4057,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 59.77425684524275,
+      "learning_rate": 3.286384976525822e-07,
+      "loss": 2.4874,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 54.25859607523927,
+      "learning_rate": 3.380281690140845e-07,
+      "loss": 2.2608,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 62.59008628669073,
+      "learning_rate": 3.4741784037558687e-07,
+      "loss": 2.1518,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 61.748549974783444,
+      "learning_rate": 3.5680751173708924e-07,
+      "loss": 2.2213,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 53.17144499215813,
+      "learning_rate": 3.661971830985916e-07,
+      "loss": 2.2406,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 46.29529444440863,
+      "learning_rate": 3.755868544600939e-07,
+      "loss": 2.0733,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 46.584758261966314,
+      "learning_rate": 3.849765258215963e-07,
+      "loss": 2.0223,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 44.768233137428126,
+      "learning_rate": 3.9436619718309864e-07,
+      "loss": 2.0254,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 40.74425845277057,
+      "learning_rate": 4.0375586854460095e-07,
+      "loss": 2.0132,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 50.58164313855809,
+      "learning_rate": 4.131455399061033e-07,
+      "loss": 2.0735,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 36.21383505508657,
+      "learning_rate": 4.225352112676057e-07,
+      "loss": 2.0147,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 35.515462287966336,
+      "learning_rate": 4.3192488262910804e-07,
+      "loss": 2.0923,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 33.58221330168005,
+      "learning_rate": 4.413145539906104e-07,
+      "loss": 1.882,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.532017516063455,
+      "learning_rate": 4.507042253521127e-07,
+      "loss": 1.8617,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 37.96487393536598,
+      "learning_rate": 4.600938967136151e-07,
+      "loss": 1.8915,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.330095839009893,
+      "learning_rate": 4.694835680751174e-07,
+      "loss": 1.8524,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.94842775461955,
+      "learning_rate": 4.788732394366198e-07,
+      "loss": 1.715,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 26.766016728849173,
+      "learning_rate": 4.882629107981221e-07,
+      "loss": 1.7667,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.463242568186832,
+      "learning_rate": 4.976525821596245e-07,
+      "loss": 1.7974,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 25.118375474568996,
+      "learning_rate": 5.070422535211268e-07,
+      "loss": 1.698,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.72813701469559,
+      "learning_rate": 5.164319248826292e-07,
+      "loss": 1.8417,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.406810690811522,
+      "learning_rate": 5.258215962441315e-07,
+      "loss": 1.6796,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 23.125783926767337,
+      "learning_rate": 5.352112676056338e-07,
+      "loss": 1.6856,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 25.995319572484725,
+      "learning_rate": 5.446009389671361e-07,
+      "loss": 1.6726,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.243142035631116,
+      "learning_rate": 5.539906103286386e-07,
+      "loss": 1.7066,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.67820611891104,
+      "learning_rate": 5.633802816901409e-07,
+      "loss": 1.6953,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.376636526861923,
+      "learning_rate": 5.727699530516433e-07,
+      "loss": 1.7589,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.02065083625833,
+      "learning_rate": 5.821596244131456e-07,
+      "loss": 1.6449,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.58201361484835,
+      "learning_rate": 5.915492957746479e-07,
+      "loss": 1.6097,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.19599497843043,
+      "learning_rate": 6.009389671361503e-07,
+      "loss": 1.6031,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.046247584431306,
+      "learning_rate": 6.103286384976526e-07,
+      "loss": 1.6507,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 25.399110107235458,
+      "learning_rate": 6.197183098591549e-07,
+      "loss": 1.5415,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.82291205526455,
+      "learning_rate": 6.291079812206573e-07,
+      "loss": 1.6317,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.021949210423973,
+      "learning_rate": 6.384976525821597e-07,
+      "loss": 1.7103,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.150920968613,
+      "learning_rate": 6.47887323943662e-07,
+      "loss": 1.6881,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.333198402713048,
+      "learning_rate": 6.572769953051644e-07,
+      "loss": 1.6202,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.8010389586852,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 1.6292,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.00157539248203,
+      "learning_rate": 6.76056338028169e-07,
+      "loss": 1.5772,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.62853873047387,
+      "learning_rate": 6.854460093896714e-07,
+      "loss": 1.5247,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.193991835422885,
+      "learning_rate": 6.948356807511737e-07,
+      "loss": 1.6593,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 24.393541649217944,
+      "learning_rate": 7.042253521126762e-07,
+      "loss": 1.6113,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.31644211061861,
+      "learning_rate": 7.136150234741785e-07,
+      "loss": 1.6773,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.691680121843476,
+      "learning_rate": 7.230046948356809e-07,
+      "loss": 1.5878,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.04497999909259,
+      "learning_rate": 7.323943661971832e-07,
+      "loss": 1.5604,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.41186846019421,
+      "learning_rate": 7.417840375586854e-07,
+      "loss": 1.6153,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.83632590811489,
+      "learning_rate": 7.511737089201878e-07,
+      "loss": 1.435,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.278070590704782,
+      "learning_rate": 7.605633802816901e-07,
+      "loss": 1.5988,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.70233274567926,
+      "learning_rate": 7.699530516431925e-07,
+      "loss": 1.482,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.57338520068876,
+      "learning_rate": 7.793427230046949e-07,
+      "loss": 1.6485,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.885891113723483,
+      "learning_rate": 7.887323943661973e-07,
+      "loss": 1.5797,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.819451503329113,
+      "learning_rate": 7.981220657276996e-07,
+      "loss": 1.6249,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.976777618671832,
+      "learning_rate": 8.075117370892019e-07,
+      "loss": 1.6556,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.54294827649258,
+      "learning_rate": 8.169014084507043e-07,
+      "loss": 1.4836,
+      "step": 435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.521234608810484,
+      "learning_rate": 8.262910798122066e-07,
+      "loss": 1.4876,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.635846092765583,
+      "learning_rate": 8.35680751173709e-07,
+      "loss": 1.5898,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.053185201621453,
+      "learning_rate": 8.450704225352114e-07,
+      "loss": 1.5442,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.423021512749994,
+      "learning_rate": 8.544600938967138e-07,
+      "loss": 1.4988,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.495061538407052,
+      "learning_rate": 8.638497652582161e-07,
+      "loss": 1.4259,
+      "step": 460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.850361920318438,
+      "learning_rate": 8.732394366197183e-07,
+      "loss": 1.5186,
+      "step": 465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.578056106424471,
+      "learning_rate": 8.826291079812208e-07,
+      "loss": 1.4974,
+      "step": 470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.858639033286455,
+      "learning_rate": 8.92018779342723e-07,
+      "loss": 1.495,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.260815266203885,
+      "learning_rate": 9.014084507042254e-07,
+      "loss": 1.5365,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.54236055418119,
+      "learning_rate": 9.107981220657277e-07,
+      "loss": 1.5177,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.42644574502043,
+      "learning_rate": 9.201877934272302e-07,
+      "loss": 1.4785,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.272213121132413,
+      "learning_rate": 9.295774647887325e-07,
+      "loss": 1.4362,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.787703070931476,
+      "learning_rate": 9.389671361502348e-07,
+      "loss": 1.4147,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.290405211460172,
+      "learning_rate": 9.483568075117372e-07,
+      "loss": 1.5199,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 31.16683031656451,
+      "learning_rate": 9.577464788732395e-07,
+      "loss": 1.5129,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.05810144796746,
+      "learning_rate": 9.67136150234742e-07,
+      "loss": 1.4579,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.910797176778393,
+      "learning_rate": 9.765258215962441e-07,
+      "loss": 1.4825,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.828025692937706,
+      "learning_rate": 9.859154929577465e-07,
+      "loss": 1.4976,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.06711314660652,
+      "learning_rate": 9.95305164319249e-07,
+      "loss": 1.5023,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.902815752999173,
+      "learning_rate": 1.0046948356807512e-06,
+      "loss": 1.4582,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.299784889411022,
+      "learning_rate": 1.0140845070422536e-06,
+      "loss": 1.4677,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.224842088929567,
+      "learning_rate": 1.023474178403756e-06,
+      "loss": 1.4832,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.16512174398155,
+      "learning_rate": 1.0328638497652584e-06,
+      "loss": 1.4232,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.124144676501672,
+      "learning_rate": 1.0422535211267606e-06,
+      "loss": 1.4937,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.53966345408406,
+      "learning_rate": 1.051643192488263e-06,
+      "loss": 1.4997,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.057068878424882,
+      "learning_rate": 1.0610328638497655e-06,
+      "loss": 1.4607,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.222469721093784,
+      "learning_rate": 1.0704225352112677e-06,
+      "loss": 1.4737,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.482584766686639,
+      "learning_rate": 1.07981220657277e-06,
+      "loss": 1.4429,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.144605621844935,
+      "learning_rate": 1.0892018779342723e-06,
+      "loss": 1.4306,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.221183173617685,
+      "learning_rate": 1.098591549295775e-06,
+      "loss": 1.398,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.06786859392066,
+      "learning_rate": 1.1079812206572771e-06,
+      "loss": 1.3978,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.058932890422534,
+      "learning_rate": 1.1173708920187793e-06,
+      "loss": 1.4144,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.953379741736752,
+      "learning_rate": 1.1267605633802817e-06,
+      "loss": 1.371,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.427524454113836,
+      "learning_rate": 1.1361502347417841e-06,
+      "loss": 1.418,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.92374482063334,
+      "learning_rate": 1.1455399061032866e-06,
+      "loss": 1.3953,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.121410703430758,
+      "learning_rate": 1.1549295774647888e-06,
+      "loss": 1.423,
+      "step": 615
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.188058478295792,
+      "learning_rate": 1.1643192488262912e-06,
+      "loss": 1.409,
+      "step": 620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.25281743075917,
+      "learning_rate": 1.1737089201877936e-06,
+      "loss": 1.3714,
+      "step": 625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.277758266359262,
+      "learning_rate": 1.1830985915492958e-06,
+      "loss": 1.4606,
+      "step": 630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.10051728291961,
+      "learning_rate": 1.1924882629107982e-06,
+      "loss": 1.3395,
+      "step": 635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 21.941601579589136,
+      "learning_rate": 1.2018779342723006e-06,
+      "loss": 1.4602,
+      "step": 640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.800228963677368,
+      "learning_rate": 1.211267605633803e-06,
+      "loss": 1.396,
+      "step": 645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.188498623370233,
+      "learning_rate": 1.2206572769953053e-06,
+      "loss": 1.3942,
+      "step": 650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.377090250263134,
+      "learning_rate": 1.2300469483568075e-06,
+      "loss": 1.4094,
+      "step": 655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 28.301310304498266,
+      "learning_rate": 1.2394366197183099e-06,
+      "loss": 1.4329,
+      "step": 660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.47871877191197,
+      "learning_rate": 1.2488262910798123e-06,
+      "loss": 1.3971,
+      "step": 665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.595306071329194,
+      "learning_rate": 1.2582159624413145e-06,
+      "loss": 1.3817,
+      "step": 670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.595530388058712,
+      "learning_rate": 1.267605633802817e-06,
+      "loss": 1.3787,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 30.086937759318808,
+      "learning_rate": 1.2769953051643193e-06,
+      "loss": 1.3526,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.172621056361283,
+      "learning_rate": 1.2863849765258218e-06,
+      "loss": 1.4155,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 24.310858020676054,
+      "learning_rate": 1.295774647887324e-06,
+      "loss": 1.3586,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.095471048461118,
+      "learning_rate": 1.3051643192488264e-06,
+      "loss": 1.3316,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.345681006867164,
+      "learning_rate": 1.3145539906103288e-06,
+      "loss": 1.3864,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.472551573600576,
+      "learning_rate": 1.323943661971831e-06,
+      "loss": 1.4093,
+      "step": 705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.80859667842158,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.3963,
+      "step": 710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.629469421345675,
+      "learning_rate": 1.3427230046948358e-06,
+      "loss": 1.3077,
+      "step": 715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 28.724843874556054,
+      "learning_rate": 1.352112676056338e-06,
+      "loss": 1.3234,
+      "step": 720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.62788509343862,
+      "learning_rate": 1.3615023474178405e-06,
+      "loss": 1.3013,
+      "step": 725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 32.190687772388905,
+      "learning_rate": 1.3708920187793429e-06,
+      "loss": 1.3736,
+      "step": 730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 28.197369021160146,
+      "learning_rate": 1.3802816901408453e-06,
+      "loss": 1.3555,
+      "step": 735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.772548510655179,
+      "learning_rate": 1.3896713615023475e-06,
+      "loss": 1.2694,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.374151647085297,
+      "learning_rate": 1.39906103286385e-06,
+      "loss": 1.3748,
+      "step": 745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.473217747365887,
+      "learning_rate": 1.4084507042253523e-06,
+      "loss": 1.3575,
+      "step": 750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 36.7012252204732,
+      "learning_rate": 1.4178403755868545e-06,
+      "loss": 1.3715,
+      "step": 755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.154802725792006,
+      "learning_rate": 1.427230046948357e-06,
+      "loss": 1.4093,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.47266839815217,
+      "learning_rate": 1.4366197183098594e-06,
+      "loss": 1.3366,
+      "step": 765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.22464496371909,
+      "learning_rate": 1.4460093896713618e-06,
+      "loss": 1.3634,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.641809458970616,
+      "learning_rate": 1.455399061032864e-06,
+      "loss": 1.3799,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.751617149883694,
+      "learning_rate": 1.4647887323943664e-06,
+      "loss": 1.3247,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.155368277725412,
+      "learning_rate": 1.4741784037558688e-06,
+      "loss": 1.3073,
+      "step": 785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.460702773460593,
+      "learning_rate": 1.4835680751173708e-06,
+      "loss": 1.3435,
+      "step": 790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.662672711998084,
+      "learning_rate": 1.4929577464788732e-06,
+      "loss": 1.3248,
+      "step": 795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.459626455820352,
+      "learning_rate": 1.5023474178403756e-06,
+      "loss": 1.3241,
+      "step": 800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.92693756733971,
+      "learning_rate": 1.5117370892018783e-06,
+      "loss": 1.2749,
+      "step": 805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.724913694283131,
+      "learning_rate": 1.5211267605633803e-06,
+      "loss": 1.3404,
+      "step": 810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.532467582650584,
+      "learning_rate": 1.5305164319248827e-06,
+      "loss": 1.3283,
+      "step": 815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.90636117794833,
+      "learning_rate": 1.539906103286385e-06,
+      "loss": 1.2561,
+      "step": 820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.765187863382085,
+      "learning_rate": 1.5492957746478873e-06,
+      "loss": 1.3248,
+      "step": 825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.974253716946507,
+      "learning_rate": 1.5586854460093897e-06,
+      "loss": 1.3316,
+      "step": 830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.230876147011728,
+      "learning_rate": 1.5680751173708921e-06,
+      "loss": 1.3496,
+      "step": 835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.653346482896024,
+      "learning_rate": 1.5774647887323946e-06,
+      "loss": 1.3542,
+      "step": 840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.288384189626518,
+      "learning_rate": 1.5868544600938968e-06,
+      "loss": 1.3754,
+      "step": 845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.595063908973746,
+      "learning_rate": 1.5962441314553992e-06,
+      "loss": 1.327,
+      "step": 850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.707725833904247,
+      "learning_rate": 1.6056338028169016e-06,
+      "loss": 1.3276,
+      "step": 855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.575003230283633,
+      "learning_rate": 1.6150234741784038e-06,
+      "loss": 1.375,
+      "step": 860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.133078679937455,
+      "learning_rate": 1.6244131455399062e-06,
+      "loss": 1.2794,
+      "step": 865
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.396732713028415,
+      "learning_rate": 1.6338028169014086e-06,
+      "loss": 1.2786,
+      "step": 870
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.0038238204167,
+      "learning_rate": 1.643192488262911e-06,
+      "loss": 1.3261,
+      "step": 875
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.006940964261577,
+      "learning_rate": 1.6525821596244132e-06,
+      "loss": 1.2926,
+      "step": 880
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.93581106394731,
+      "learning_rate": 1.6619718309859157e-06,
+      "loss": 1.282,
+      "step": 885
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.65592428076664,
+      "learning_rate": 1.671361502347418e-06,
+      "loss": 1.2591,
+      "step": 890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.315244923529148,
+      "learning_rate": 1.6807511737089203e-06,
+      "loss": 1.4264,
+      "step": 895
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.74543685655779,
+      "learning_rate": 1.6901408450704227e-06,
+      "loss": 1.283,
+      "step": 900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.21183412560726,
+      "learning_rate": 1.6995305164319251e-06,
+      "loss": 1.3377,
+      "step": 905
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.587485547338511,
+      "learning_rate": 1.7089201877934275e-06,
+      "loss": 1.3355,
+      "step": 910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.570554898369355,
+      "learning_rate": 1.7183098591549297e-06,
+      "loss": 1.3419,
+      "step": 915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.080005496329896,
+      "learning_rate": 1.7276995305164322e-06,
+      "loss": 1.3719,
+      "step": 920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.601733886689647,
+      "learning_rate": 1.7370892018779346e-06,
+      "loss": 1.3305,
+      "step": 925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.182001359469993,
+      "learning_rate": 1.7464788732394366e-06,
+      "loss": 1.3133,
+      "step": 930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.654297993584537,
+      "learning_rate": 1.755868544600939e-06,
+      "loss": 1.2799,
+      "step": 935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.257829805726358,
+      "learning_rate": 1.7652582159624416e-06,
+      "loss": 1.2324,
+      "step": 940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.585485409291497,
+      "learning_rate": 1.774647887323944e-06,
+      "loss": 1.3184,
+      "step": 945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.159592380113939,
+      "learning_rate": 1.784037558685446e-06,
+      "loss": 1.2731,
+      "step": 950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.98072436891059,
+      "learning_rate": 1.7934272300469484e-06,
+      "loss": 1.3129,
+      "step": 955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.334950203134358,
+      "learning_rate": 1.8028169014084509e-06,
+      "loss": 1.3025,
+      "step": 960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.831771087571925,
+      "learning_rate": 1.812206572769953e-06,
+      "loss": 1.2471,
+      "step": 965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 25.594671065609298,
+      "learning_rate": 1.8215962441314555e-06,
+      "loss": 1.3591,
+      "step": 970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.09598181520699,
+      "learning_rate": 1.8309859154929579e-06,
+      "loss": 1.2491,
+      "step": 975
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.079393096266388,
+      "learning_rate": 1.8403755868544603e-06,
+      "loss": 1.3326,
+      "step": 980
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.672109157664112,
+      "learning_rate": 1.8497652582159625e-06,
+      "loss": 1.2313,
+      "step": 985
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.243721125514357,
+      "learning_rate": 1.859154929577465e-06,
+      "loss": 1.3036,
+      "step": 990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.324736332255661,
+      "learning_rate": 1.8685446009389673e-06,
+      "loss": 1.3095,
+      "step": 995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.27952026335669,
+      "learning_rate": 1.8779342723004696e-06,
+      "loss": 1.2832,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.979541899673272,
+      "learning_rate": 1.887323943661972e-06,
+      "loss": 1.3264,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.22324867764637,
+      "learning_rate": 1.8967136150234744e-06,
+      "loss": 1.2964,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 20.862442145452597,
+      "learning_rate": 1.9061032863849766e-06,
+      "loss": 1.2276,
+      "step": 1015
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.56689153802307,
+      "learning_rate": 1.915492957746479e-06,
+      "loss": 1.2277,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.59196128567854,
+      "learning_rate": 1.9248826291079814e-06,
+      "loss": 1.27,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.299128498391946,
+      "learning_rate": 1.934272300469484e-06,
+      "loss": 1.3295,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.20043989853521,
+      "learning_rate": 1.943661971830986e-06,
+      "loss": 1.3343,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.59109623708328,
+      "learning_rate": 1.9530516431924883e-06,
+      "loss": 1.3281,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.343332788368398,
+      "learning_rate": 1.9624413145539907e-06,
+      "loss": 1.339,
+      "step": 1045
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 34.38348243471212,
+      "learning_rate": 1.971830985915493e-06,
+      "loss": 1.2953,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 57.553178005750915,
+      "learning_rate": 1.9812206572769955e-06,
+      "loss": 1.2797,
+      "step": 1055
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.395373329350186,
+      "learning_rate": 1.990610328638498e-06,
+      "loss": 1.2543,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.18711726177598,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.2292,
+      "step": 1065
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.898228878702902,
+      "learning_rate": 2.0093896713615023e-06,
+      "loss": 1.3609,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.501589514924925,
+      "learning_rate": 2.0187793427230047e-06,
+      "loss": 1.2875,
+      "step": 1075
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.570857168438987,
+      "learning_rate": 2.028169014084507e-06,
+      "loss": 1.2705,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.011759356869026,
+      "learning_rate": 2.0375586854460096e-06,
+      "loss": 1.2501,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.36984363387079,
+      "learning_rate": 2.046948356807512e-06,
+      "loss": 1.2802,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.573554484180887,
+      "learning_rate": 2.0563380281690144e-06,
+      "loss": 1.2299,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.346613101864381,
+      "learning_rate": 2.065727699530517e-06,
+      "loss": 1.278,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.673556476995792,
+      "learning_rate": 2.075117370892019e-06,
+      "loss": 1.2857,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.743903049798025,
+      "learning_rate": 2.0845070422535212e-06,
+      "loss": 1.2731,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.095242668024651,
+      "learning_rate": 2.0938967136150237e-06,
+      "loss": 1.3167,
+      "step": 1115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.525888405537948,
+      "learning_rate": 2.103286384976526e-06,
+      "loss": 1.2914,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.939198161893481,
+      "learning_rate": 2.1126760563380285e-06,
+      "loss": 1.2053,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.462157973981705,
+      "learning_rate": 2.122065727699531e-06,
+      "loss": 1.2672,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.316075393990364,
+      "learning_rate": 2.1314553990610333e-06,
+      "loss": 1.2912,
+      "step": 1135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.059605944650713,
+      "learning_rate": 2.1408450704225353e-06,
+      "loss": 1.2916,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.315508373754005,
+      "learning_rate": 2.1502347417840377e-06,
+      "loss": 1.2437,
+      "step": 1145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.260117294160276,
+      "learning_rate": 2.15962441314554e-06,
+      "loss": 1.2363,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.201836066155025,
+      "learning_rate": 2.169014084507042e-06,
+      "loss": 1.3131,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.948713842229541,
+      "learning_rate": 2.1784037558685446e-06,
+      "loss": 1.277,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 32.52981277229426,
+      "learning_rate": 2.187793427230047e-06,
+      "loss": 1.3297,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 28.419378073727167,
+      "learning_rate": 2.19718309859155e-06,
+      "loss": 1.3176,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 36.13582450376201,
+      "learning_rate": 2.206572769953052e-06,
+      "loss": 1.1653,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 62.500714210346636,
+      "learning_rate": 2.2159624413145542e-06,
+      "loss": 1.2826,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 61.90731014306774,
+      "learning_rate": 2.2253521126760566e-06,
+      "loss": 1.3107,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 39.50365457516419,
+      "learning_rate": 2.2347417840375586e-06,
+      "loss": 1.2295,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 35.84862984982831,
+      "learning_rate": 2.244131455399061e-06,
+      "loss": 1.2772,
+      "step": 1195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.65318332059801,
+      "learning_rate": 2.2535211267605635e-06,
+      "loss": 1.3234,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.635645594041737,
+      "learning_rate": 2.262910798122066e-06,
+      "loss": 1.2615,
+      "step": 1205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.498192745981429,
+      "learning_rate": 2.2723004694835683e-06,
+      "loss": 1.3146,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.401059501962532,
+      "learning_rate": 2.2816901408450707e-06,
+      "loss": 1.2705,
+      "step": 1215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.747909461352078,
+      "learning_rate": 2.291079812206573e-06,
+      "loss": 1.3012,
+      "step": 1220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.72255084965526,
+      "learning_rate": 2.300469483568075e-06,
+      "loss": 1.2489,
+      "step": 1225
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.31235925255563,
+      "learning_rate": 2.3098591549295775e-06,
+      "loss": 1.2716,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.9954248660628,
+      "learning_rate": 2.31924882629108e-06,
+      "loss": 1.2991,
+      "step": 1235
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.42949763678413,
+      "learning_rate": 2.3286384976525824e-06,
+      "loss": 1.2765,
+      "step": 1240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 23.061281608940302,
+      "learning_rate": 2.338028169014085e-06,
+      "loss": 1.2494,
+      "step": 1245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.734814565335228,
+      "learning_rate": 2.347417840375587e-06,
+      "loss": 1.2396,
+      "step": 1250
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.540481170544485,
+      "learning_rate": 2.3568075117370896e-06,
+      "loss": 1.2321,
+      "step": 1255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.028057084580603,
+      "learning_rate": 2.3661971830985916e-06,
+      "loss": 1.2415,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.043226858203862,
+      "learning_rate": 2.375586854460094e-06,
+      "loss": 1.226,
+      "step": 1265
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.477510290358993,
+      "learning_rate": 2.3849765258215964e-06,
+      "loss": 1.2521,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.256647846745311,
+      "learning_rate": 2.3943661971830984e-06,
+      "loss": 1.2849,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.742208231148163,
+      "learning_rate": 2.4037558685446013e-06,
+      "loss": 1.1871,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.927219685267502,
+      "learning_rate": 2.4131455399061037e-06,
+      "loss": 1.2489,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.756776186947189,
+      "learning_rate": 2.422535211267606e-06,
+      "loss": 1.2231,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.616992125356624,
+      "learning_rate": 2.431924882629108e-06,
+      "loss": 1.2557,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.701257195294193,
+      "learning_rate": 2.4413145539906105e-06,
+      "loss": 1.2434,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.699896318555654,
+      "learning_rate": 2.450704225352113e-06,
+      "loss": 1.2262,
+      "step": 1305
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.530133028071965,
+      "learning_rate": 2.460093896713615e-06,
+      "loss": 1.3151,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.84474791051373,
+      "learning_rate": 2.4694835680751174e-06,
+      "loss": 1.269,
+      "step": 1315
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 36.51089454980534,
+      "learning_rate": 2.4788732394366198e-06,
+      "loss": 1.2982,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.941829354051336,
+      "learning_rate": 2.488262910798122e-06,
+      "loss": 1.2912,
+      "step": 1325
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 24.72393459100198,
+      "learning_rate": 2.4976525821596246e-06,
+      "loss": 1.3132,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.974596031712748,
+      "learning_rate": 2.507042253521127e-06,
+      "loss": 1.2502,
+      "step": 1335
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.752314054915557,
+      "learning_rate": 2.516431924882629e-06,
+      "loss": 1.2283,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.543313397681446,
+      "learning_rate": 2.525821596244132e-06,
+      "loss": 1.2551,
+      "step": 1345
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.87340019920864,
+      "learning_rate": 2.535211267605634e-06,
+      "loss": 1.1885,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.08675819665147,
+      "learning_rate": 2.5446009389671363e-06,
+      "loss": 1.2108,
+      "step": 1355
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 28.13883185096642,
+      "learning_rate": 2.5539906103286387e-06,
+      "loss": 1.2596,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 40.189244786733674,
+      "learning_rate": 2.563380281690141e-06,
+      "loss": 1.3154,
+      "step": 1365
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 42.33542568705795,
+      "learning_rate": 2.5727699530516435e-06,
+      "loss": 1.2612,
+      "step": 1370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 30.275801063302588,
+      "learning_rate": 2.582159624413146e-06,
+      "loss": 1.2936,
+      "step": 1375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 20.162098963082585,
+      "learning_rate": 2.591549295774648e-06,
+      "loss": 1.2896,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.498957249092086,
+      "learning_rate": 2.6009389671361508e-06,
+      "loss": 1.2507,
+      "step": 1385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.639486264516998,
+      "learning_rate": 2.6103286384976528e-06,
+      "loss": 1.2604,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 29.8483209419568,
+      "learning_rate": 2.619718309859155e-06,
+      "loss": 1.2366,
+      "step": 1395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.394341391555916,
+      "learning_rate": 2.6291079812206576e-06,
+      "loss": 1.2213,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.344509034756229,
+      "learning_rate": 2.63849765258216e-06,
+      "loss": 1.2865,
+      "step": 1405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.299775911148041,
+      "learning_rate": 2.647887323943662e-06,
+      "loss": 1.1872,
+      "step": 1410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 19.95285873093074,
+      "learning_rate": 2.657276995305165e-06,
+      "loss": 1.3064,
+      "step": 1415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 28.29931505654469,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.2422,
+      "step": 1420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.03172510028025,
+      "learning_rate": 2.676056338028169e-06,
+      "loss": 1.2986,
+      "step": 1425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.757275809832382,
+      "learning_rate": 2.6854460093896717e-06,
+      "loss": 1.1891,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.522088430581046,
+      "learning_rate": 2.6948356807511737e-06,
+      "loss": 1.29,
+      "step": 1435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 56.04452497266458,
+      "learning_rate": 2.704225352112676e-06,
+      "loss": 1.2318,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.328382728439792,
+      "learning_rate": 2.713615023474179e-06,
+      "loss": 1.2388,
+      "step": 1445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 27.67864633622398,
+      "learning_rate": 2.723004694835681e-06,
+      "loss": 1.2536,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 20.618348374946798,
+      "learning_rate": 2.7323943661971837e-06,
+      "loss": 1.2223,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 34.38731255019953,
+      "learning_rate": 2.7417840375586857e-06,
+      "loss": 1.2435,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.200577896742542,
+      "learning_rate": 2.7511737089201877e-06,
+      "loss": 1.2095,
+      "step": 1465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.6577772988377,
+      "learning_rate": 2.7605633802816906e-06,
+      "loss": 1.282,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.44746895708401,
+      "learning_rate": 2.7699530516431926e-06,
+      "loss": 1.2528,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 19.58629354717819,
+      "learning_rate": 2.779342723004695e-06,
+      "loss": 1.2114,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.911014872106671,
+      "learning_rate": 2.7887323943661974e-06,
+      "loss": 1.1756,
+      "step": 1485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.241297502880386,
+      "learning_rate": 2.7981220657277e-06,
+      "loss": 1.2957,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.997669865289506,
+      "learning_rate": 2.807511737089202e-06,
+      "loss": 1.2558,
+      "step": 1495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.127084753806834,
+      "learning_rate": 2.8169014084507046e-06,
+      "loss": 1.2417,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.97961856498744,
+      "learning_rate": 2.8262910798122066e-06,
+      "loss": 1.1799,
+      "step": 1505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.92322598523992,
+      "learning_rate": 2.835680751173709e-06,
+      "loss": 1.3069,
+      "step": 1510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.178134914863689,
+      "learning_rate": 2.8450704225352115e-06,
+      "loss": 1.2267,
+      "step": 1515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.73418912513778,
+      "learning_rate": 2.854460093896714e-06,
+      "loss": 1.1838,
+      "step": 1520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.57569030789101,
+      "learning_rate": 2.8638497652582163e-06,
+      "loss": 1.2699,
+      "step": 1525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.103205074818657,
+      "learning_rate": 2.8732394366197187e-06,
+      "loss": 1.2513,
+      "step": 1530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.74796858229263,
+      "learning_rate": 2.8826291079812207e-06,
+      "loss": 1.2233,
+      "step": 1535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.125121220128353,
+      "learning_rate": 2.8920187793427236e-06,
+      "loss": 1.2512,
+      "step": 1540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.923327802812725,
+      "learning_rate": 2.9014084507042255e-06,
+      "loss": 1.2217,
+      "step": 1545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 29.864230330956797,
+      "learning_rate": 2.910798122065728e-06,
+      "loss": 1.2279,
+      "step": 1550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.32678784754926,
+      "learning_rate": 2.9201877934272304e-06,
+      "loss": 1.2437,
+      "step": 1555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.743960758383423,
+      "learning_rate": 2.929577464788733e-06,
+      "loss": 1.2021,
+      "step": 1560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.866384456015291,
+      "learning_rate": 2.938967136150235e-06,
+      "loss": 1.1487,
+      "step": 1565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.098250571877362,
+      "learning_rate": 2.9483568075117376e-06,
+      "loss": 1.1885,
+      "step": 1570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.259031939504771,
+      "learning_rate": 2.9577464788732396e-06,
+      "loss": 1.2476,
+      "step": 1575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.472449698302071,
+      "learning_rate": 2.9671361502347416e-06,
+      "loss": 1.2519,
+      "step": 1580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.30621641804569,
+      "learning_rate": 2.9765258215962445e-06,
+      "loss": 1.2108,
+      "step": 1585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.804495414027024,
+      "learning_rate": 2.9859154929577465e-06,
+      "loss": 1.2356,
+      "step": 1590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 79.8771308486058,
+      "learning_rate": 2.9953051643192493e-06,
+      "loss": 1.2927,
+      "step": 1595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 36.98912216957134,
+      "learning_rate": 3.0046948356807513e-06,
+      "loss": 1.2825,
+      "step": 1600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 24.22949511816613,
+      "learning_rate": 3.0140845070422537e-06,
+      "loss": 1.2393,
+      "step": 1605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 43.48801816298305,
+      "learning_rate": 3.0234741784037565e-06,
+      "loss": 1.2168,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 69.02114420972799,
+      "learning_rate": 3.0328638497652585e-06,
+      "loss": 1.2701,
+      "step": 1615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 51.13066642076986,
+      "learning_rate": 3.0422535211267605e-06,
+      "loss": 1.1947,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 41.12245034419979,
+      "learning_rate": 3.0516431924882634e-06,
+      "loss": 1.2376,
+      "step": 1625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.919946134065007,
+      "learning_rate": 3.0610328638497654e-06,
+      "loss": 1.2236,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.287405241698428,
+      "learning_rate": 3.0704225352112678e-06,
+      "loss": 1.1929,
+      "step": 1635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.292933429820003,
+      "learning_rate": 3.07981220657277e-06,
+      "loss": 1.2267,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 21.71127998573997,
+      "learning_rate": 3.0892018779342726e-06,
+      "loss": 1.2475,
+      "step": 1645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 31.074374172436762,
+      "learning_rate": 3.0985915492957746e-06,
+      "loss": 1.2199,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.182189207241336,
+      "learning_rate": 3.1079812206572774e-06,
+      "loss": 1.1956,
+      "step": 1655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.695901024833002,
+      "learning_rate": 3.1173708920187794e-06,
+      "loss": 1.265,
+      "step": 1660
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 23.652515760504574,
+      "learning_rate": 3.1267605633802823e-06,
+      "loss": 1.2191,
+      "step": 1665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.510446962166238,
+      "learning_rate": 3.1361502347417843e-06,
+      "loss": 1.2269,
+      "step": 1670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.55534207393913,
+      "learning_rate": 3.1455399061032867e-06,
+      "loss": 1.1944,
+      "step": 1675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.834353219589225,
+      "learning_rate": 3.154929577464789e-06,
+      "loss": 1.2206,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.347622734688182,
+      "learning_rate": 3.1643192488262915e-06,
+      "loss": 1.2037,
+      "step": 1685
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.082052478565581,
+      "learning_rate": 3.1737089201877935e-06,
+      "loss": 1.214,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.235478472594236,
+      "learning_rate": 3.1830985915492964e-06,
+      "loss": 1.2618,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.51370136559126,
+      "learning_rate": 3.1924882629107983e-06,
+      "loss": 1.1912,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.518871002073665,
+      "learning_rate": 3.2018779342723003e-06,
+      "loss": 1.2252,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.183774407336283,
+      "learning_rate": 3.211267605633803e-06,
+      "loss": 1.1821,
+      "step": 1710
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.953755616187795,
+      "learning_rate": 3.2206572769953056e-06,
+      "loss": 1.178,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.16947579342062,
+      "learning_rate": 3.2300469483568076e-06,
+      "loss": 1.2226,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.102934787903266,
+      "learning_rate": 3.2394366197183104e-06,
+      "loss": 1.1678,
+      "step": 1725
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.20945662610384,
+      "learning_rate": 3.2488262910798124e-06,
+      "loss": 1.2281,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.459902129059378,
+      "learning_rate": 3.2582159624413144e-06,
+      "loss": 1.2358,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.087525019755466,
+      "learning_rate": 3.2676056338028173e-06,
+      "loss": 1.2747,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.388260393020143,
+      "learning_rate": 3.2769953051643192e-06,
+      "loss": 1.1767,
+      "step": 1745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.354232491977962,
+      "learning_rate": 3.286384976525822e-06,
+      "loss": 1.2726,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.134829932865085,
+      "learning_rate": 3.295774647887324e-06,
+      "loss": 1.2588,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.99587325617737,
+      "learning_rate": 3.3051643192488265e-06,
+      "loss": 1.2264,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.734943211600395,
+      "learning_rate": 3.314553990610329e-06,
+      "loss": 1.256,
+      "step": 1765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.34217967087882,
+      "learning_rate": 3.3239436619718313e-06,
+      "loss": 1.2513,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.042606148479086,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.1657,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.693454161873957,
+      "learning_rate": 3.342723004694836e-06,
+      "loss": 1.2002,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 19.002447138761273,
+      "learning_rate": 3.352112676056338e-06,
+      "loss": 1.2263,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.60649729452257,
+      "learning_rate": 3.3615023474178406e-06,
+      "loss": 1.1948,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.752199113163702,
+      "learning_rate": 3.370892018779343e-06,
+      "loss": 1.225,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.912569900789103,
+      "learning_rate": 3.3802816901408454e-06,
+      "loss": 1.1964,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 28.776006217385422,
+      "learning_rate": 3.3896713615023474e-06,
+      "loss": 1.2135,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.445070304376742,
+      "learning_rate": 3.3990610328638502e-06,
+      "loss": 1.1712,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 21.66514653599388,
+      "learning_rate": 3.4084507042253522e-06,
+      "loss": 1.2671,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.12605318731465,
+      "learning_rate": 3.417840375586855e-06,
+      "loss": 1.2047,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 22.66942288865685,
+      "learning_rate": 3.427230046948357e-06,
+      "loss": 1.1977,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 28.517348201476594,
+      "learning_rate": 3.4366197183098595e-06,
+      "loss": 1.2314,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.236614878342387,
+      "learning_rate": 3.446009389671362e-06,
+      "loss": 1.2479,
+      "step": 1835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.58188772698804,
+      "learning_rate": 3.4553990610328643e-06,
+      "loss": 1.1897,
+      "step": 1840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.450810792750605,
+      "learning_rate": 3.4647887323943663e-06,
+      "loss": 1.2657,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.328144068914792,
+      "learning_rate": 3.474178403755869e-06,
+      "loss": 1.1788,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.646147429445598,
+      "learning_rate": 3.483568075117371e-06,
+      "loss": 1.1874,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.046538853977292,
+      "learning_rate": 3.492957746478873e-06,
+      "loss": 1.1592,
+      "step": 1860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.704241336750403,
+      "learning_rate": 3.502347417840376e-06,
+      "loss": 1.1885,
+      "step": 1865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.4346558829419,
+      "learning_rate": 3.511737089201878e-06,
+      "loss": 1.2097,
+      "step": 1870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.397004458626839,
+      "learning_rate": 3.5211267605633804e-06,
+      "loss": 1.168,
+      "step": 1875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.039403557104384,
+      "learning_rate": 3.5305164319248832e-06,
+      "loss": 1.2138,
+      "step": 1880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.033316856532485,
+      "learning_rate": 3.5399061032863852e-06,
+      "loss": 1.2104,
+      "step": 1885
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.581742073717887,
+      "learning_rate": 3.549295774647888e-06,
+      "loss": 1.1657,
+      "step": 1890
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.193936813759473,
+      "learning_rate": 3.55868544600939e-06,
+      "loss": 1.1598,
+      "step": 1895
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.783341696812734,
+      "learning_rate": 3.568075117370892e-06,
+      "loss": 1.1809,
+      "step": 1900
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.53705773501876,
+      "learning_rate": 3.577464788732395e-06,
+      "loss": 1.2116,
+      "step": 1905
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 32.46653112162618,
+      "learning_rate": 3.586854460093897e-06,
+      "loss": 1.1735,
+      "step": 1910
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.29347797370921,
+      "learning_rate": 3.5962441314553993e-06,
+      "loss": 1.1923,
+      "step": 1915
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 49.30618204917337,
+      "learning_rate": 3.6056338028169017e-06,
+      "loss": 1.216,
+      "step": 1920
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 19.760430970051083,
+      "learning_rate": 3.615023474178404e-06,
+      "loss": 1.2522,
+      "step": 1925
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.109595738192525,
+      "learning_rate": 3.624413145539906e-06,
+      "loss": 1.1739,
+      "step": 1930
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.032334502759433,
+      "learning_rate": 3.633802816901409e-06,
+      "loss": 1.2265,
+      "step": 1935
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.839563942638703,
+      "learning_rate": 3.643192488262911e-06,
+      "loss": 1.2752,
+      "step": 1940
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.061383020324275,
+      "learning_rate": 3.6525821596244134e-06,
+      "loss": 1.2062,
+      "step": 1945
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.490899045262893,
+      "learning_rate": 3.6619718309859158e-06,
+      "loss": 1.2212,
+      "step": 1950
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.176378163025321,
+      "learning_rate": 3.671361502347418e-06,
+      "loss": 1.2657,
+      "step": 1955
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.368210617302188,
+      "learning_rate": 3.6807511737089206e-06,
+      "loss": 1.1647,
+      "step": 1960
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 30.91571927783136,
+      "learning_rate": 3.690140845070423e-06,
+      "loss": 1.2092,
+      "step": 1965
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 29.255455789708254,
+      "learning_rate": 3.699530516431925e-06,
+      "loss": 1.2392,
+      "step": 1970
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 36.16139916768643,
+      "learning_rate": 3.708920187793428e-06,
+      "loss": 1.1945,
+      "step": 1975
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 29.10675274421538,
+      "learning_rate": 3.71830985915493e-06,
+      "loss": 1.2732,
+      "step": 1980
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 21.275163471293087,
+      "learning_rate": 3.7276995305164323e-06,
+      "loss": 1.2389,
+      "step": 1985
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.208871298142192,
+      "learning_rate": 3.7370892018779347e-06,
+      "loss": 1.2168,
+      "step": 1990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.469318357293972,
+      "learning_rate": 3.746478873239437e-06,
+      "loss": 1.1905,
+      "step": 1995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.688363452103964,
+      "learning_rate": 3.755868544600939e-06,
+      "loss": 1.2247,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 22.47927123788488,
+      "learning_rate": 3.765258215962442e-06,
+      "loss": 1.2028,
+      "step": 2005
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.822752014295487,
+      "learning_rate": 3.774647887323944e-06,
+      "loss": 1.2249,
+      "step": 2010
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.775832776295033,
+      "learning_rate": 3.784037558685446e-06,
+      "loss": 1.1848,
+      "step": 2015
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.403695439707086,
+      "learning_rate": 3.7934272300469488e-06,
+      "loss": 1.2687,
+      "step": 2020
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 25.544886300422906,
+      "learning_rate": 3.8028169014084508e-06,
+      "loss": 1.191,
+      "step": 2025
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 26.852316151220336,
+      "learning_rate": 3.812206572769953e-06,
+      "loss": 1.215,
+      "step": 2030
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.691188545436885,
+      "learning_rate": 3.821596244131456e-06,
+      "loss": 1.2523,
+      "step": 2035
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.774750676679698,
+      "learning_rate": 3.830985915492958e-06,
+      "loss": 1.1613,
+      "step": 2040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.23137836953399,
+      "learning_rate": 3.8403755868544604e-06,
+      "loss": 1.2546,
+      "step": 2045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.4462130773395,
+      "learning_rate": 3.849765258215963e-06,
+      "loss": 1.196,
+      "step": 2050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.07146144561959,
+      "learning_rate": 3.859154929577465e-06,
+      "loss": 1.1759,
+      "step": 2055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.880854867288102,
+      "learning_rate": 3.868544600938968e-06,
+      "loss": 1.1976,
+      "step": 2060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.23859917864838,
+      "learning_rate": 3.87793427230047e-06,
+      "loss": 1.1506,
+      "step": 2065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.261931534727905,
+      "learning_rate": 3.887323943661972e-06,
+      "loss": 1.1614,
+      "step": 2070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.990337812660432,
+      "learning_rate": 3.896713615023475e-06,
+      "loss": 1.189,
+      "step": 2075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.297729793250264,
+      "learning_rate": 3.9061032863849765e-06,
+      "loss": 1.1841,
+      "step": 2080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.809849895984062,
+      "learning_rate": 3.915492957746479e-06,
+      "loss": 1.1932,
+      "step": 2085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.644849837301333,
+      "learning_rate": 3.924882629107981e-06,
+      "loss": 1.251,
+      "step": 2090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.654947680600644,
+      "learning_rate": 3.934272300469484e-06,
+      "loss": 1.1413,
+      "step": 2095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.666717543340287,
+      "learning_rate": 3.943661971830986e-06,
+      "loss": 1.2533,
+      "step": 2100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.23515916827622,
+      "learning_rate": 3.953051643192489e-06,
+      "loss": 1.1855,
+      "step": 2105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.611998617092816,
+      "learning_rate": 3.962441314553991e-06,
+      "loss": 1.1686,
+      "step": 2110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 75.18546876934055,
+      "learning_rate": 3.971830985915493e-06,
+      "loss": 1.211,
+      "step": 2115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 34.48381540632781,
+      "learning_rate": 3.981220657276996e-06,
+      "loss": 1.265,
+      "step": 2120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 24.173242774205907,
+      "learning_rate": 3.990610328638498e-06,
+      "loss": 1.2204,
+      "step": 2125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.0891214249118,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.1923,
+      "step": 2130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 41.64291555315773,
+      "learning_rate": 4.009389671361503e-06,
+      "loss": 1.2128,
+      "step": 2135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.281938539315025,
+      "learning_rate": 4.018779342723005e-06,
+      "loss": 1.2072,
+      "step": 2140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.413266774979027,
+      "learning_rate": 4.028169014084508e-06,
+      "loss": 1.2549,
+      "step": 2145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 30.710934721912487,
+      "learning_rate": 4.0375586854460095e-06,
+      "loss": 1.1913,
+      "step": 2150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.110650332435146,
+      "learning_rate": 4.046948356807512e-06,
+      "loss": 1.199,
+      "step": 2155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.431332214312636,
+      "learning_rate": 4.056338028169014e-06,
+      "loss": 1.2076,
+      "step": 2160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.946223670549578,
+      "learning_rate": 4.065727699530517e-06,
+      "loss": 1.2505,
+      "step": 2165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.049396887988486,
+      "learning_rate": 4.075117370892019e-06,
+      "loss": 1.2853,
+      "step": 2170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.991821965340616,
+      "learning_rate": 4.0845070422535216e-06,
+      "loss": 1.2469,
+      "step": 2175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 21.132829306364883,
+      "learning_rate": 4.093896713615024e-06,
+      "loss": 1.1638,
+      "step": 2180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.734974513523571,
+      "learning_rate": 4.103286384976526e-06,
+      "loss": 1.2183,
+      "step": 2185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 22.656682822121244,
+      "learning_rate": 4.112676056338029e-06,
+      "loss": 1.185,
+      "step": 2190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 55.96848982649205,
+      "learning_rate": 4.12206572769953e-06,
+      "loss": 1.2009,
+      "step": 2195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.454671520496518,
+      "learning_rate": 4.131455399061034e-06,
+      "loss": 1.1985,
+      "step": 2200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.289250236753556,
+      "learning_rate": 4.140845070422535e-06,
+      "loss": 1.2101,
+      "step": 2205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.676033470734506,
+      "learning_rate": 4.150234741784038e-06,
+      "loss": 1.1882,
+      "step": 2210
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.254135510272167,
+      "learning_rate": 4.15962441314554e-06,
+      "loss": 1.1654,
+      "step": 2215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.98453780071096,
+      "learning_rate": 4.1690140845070425e-06,
+      "loss": 1.2611,
+      "step": 2220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.549573390214084,
+      "learning_rate": 4.178403755868545e-06,
+      "loss": 1.1565,
+      "step": 2225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.58385597390127,
+      "learning_rate": 4.187793427230047e-06,
+      "loss": 1.1752,
+      "step": 2230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.392989782907804,
+      "learning_rate": 4.19718309859155e-06,
+      "loss": 1.2213,
+      "step": 2235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.978724708307432,
+      "learning_rate": 4.206572769953052e-06,
+      "loss": 1.1658,
+      "step": 2240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.119250016932348,
+      "learning_rate": 4.2159624413145546e-06,
+      "loss": 1.1651,
+      "step": 2245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.990008208774727,
+      "learning_rate": 4.225352112676057e-06,
+      "loss": 1.2203,
+      "step": 2250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.797731311304023,
+      "learning_rate": 4.2347417840375585e-06,
+      "loss": 1.2147,
+      "step": 2255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.574322167429386,
+      "learning_rate": 4.244131455399062e-06,
+      "loss": 1.2288,
+      "step": 2260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 67.25446012983973,
+      "learning_rate": 4.253521126760563e-06,
+      "loss": 1.2248,
+      "step": 2265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.087015270076762,
+      "learning_rate": 4.262910798122067e-06,
+      "loss": 1.251,
+      "step": 2270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 32.14474547185929,
+      "learning_rate": 4.272300469483568e-06,
+      "loss": 1.1826,
+      "step": 2275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 45.47076478471937,
+      "learning_rate": 4.281690140845071e-06,
+      "loss": 1.2227,
+      "step": 2280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 36.72117557644853,
+      "learning_rate": 4.291079812206573e-06,
+      "loss": 1.2262,
+      "step": 2285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.1574966708121,
+      "learning_rate": 4.3004694835680755e-06,
+      "loss": 1.1448,
+      "step": 2290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.615420459592887,
+      "learning_rate": 4.309859154929578e-06,
+      "loss": 1.1637,
+      "step": 2295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 32.35427176558952,
+      "learning_rate": 4.31924882629108e-06,
+      "loss": 1.2204,
+      "step": 2300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.4522668863733,
+      "learning_rate": 4.328638497652583e-06,
+      "loss": 1.2138,
+      "step": 2305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.937973098253103,
+      "learning_rate": 4.338028169014084e-06,
+      "loss": 1.2189,
+      "step": 2310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.770652084639305,
+      "learning_rate": 4.3474178403755875e-06,
+      "loss": 1.1874,
+      "step": 2315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 31.721192010565304,
+      "learning_rate": 4.356807511737089e-06,
+      "loss": 1.1749,
+      "step": 2320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.921912666981186,
+      "learning_rate": 4.3661971830985915e-06,
+      "loss": 1.2121,
+      "step": 2325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.3296836830859,
+      "learning_rate": 4.375586854460094e-06,
+      "loss": 1.2265,
+      "step": 2330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 19.536602882796448,
+      "learning_rate": 4.384976525821596e-06,
+      "loss": 1.1435,
+      "step": 2335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.331085645634388,
+      "learning_rate": 4.3943661971831e-06,
+      "loss": 1.1882,
+      "step": 2340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 22.917501240167,
+      "learning_rate": 4.403755868544601e-06,
+      "loss": 1.1389,
+      "step": 2345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 25.79328209161115,
+      "learning_rate": 4.413145539906104e-06,
+      "loss": 1.2164,
+      "step": 2350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.358192281594208,
+      "learning_rate": 4.422535211267606e-06,
+      "loss": 1.176,
+      "step": 2355
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.43037237322513,
+      "learning_rate": 4.4319248826291084e-06,
+      "loss": 1.194,
+      "step": 2360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 26.736859965173707,
+      "learning_rate": 4.441314553990611e-06,
+      "loss": 1.2026,
+      "step": 2365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.52324590542694,
+      "learning_rate": 4.450704225352113e-06,
+      "loss": 1.1399,
+      "step": 2370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.92299270355069,
+      "learning_rate": 4.460093896713616e-06,
+      "loss": 1.219,
+      "step": 2375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.095558611381136,
+      "learning_rate": 4.469483568075117e-06,
+      "loss": 1.1769,
+      "step": 2380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.923688201653198,
+      "learning_rate": 4.4788732394366205e-06,
+      "loss": 1.1507,
+      "step": 2385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 31.590920243735336,
+      "learning_rate": 4.488262910798122e-06,
+      "loss": 1.1989,
+      "step": 2390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.408685623776789,
+      "learning_rate": 4.4976525821596245e-06,
+      "loss": 1.1727,
+      "step": 2395
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 22.466982177506797,
+      "learning_rate": 4.507042253521127e-06,
+      "loss": 1.1682,
+      "step": 2400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 15.601061830318697,
+      "learning_rate": 4.516431924882629e-06,
+      "loss": 1.1787,
+      "step": 2405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.385013048034264,
+      "learning_rate": 4.525821596244132e-06,
+      "loss": 1.2546,
+      "step": 2410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 47.0796561622599,
+      "learning_rate": 4.535211267605634e-06,
+      "loss": 1.2035,
+      "step": 2415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 69.181673678111,
+      "learning_rate": 4.544600938967137e-06,
+      "loss": 1.2772,
+      "step": 2420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.158561027487735,
+      "learning_rate": 4.553990610328639e-06,
+      "loss": 1.1987,
+      "step": 2425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 32.647291631523196,
+      "learning_rate": 4.5633802816901414e-06,
+      "loss": 1.2427,
+      "step": 2430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 54.632444722814455,
+      "learning_rate": 4.572769953051643e-06,
+      "loss": 1.2666,
+      "step": 2435
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 37.02072504555358,
+      "learning_rate": 4.582159624413146e-06,
+      "loss": 1.2341,
+      "step": 2440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.220404511205434,
+      "learning_rate": 4.591549295774648e-06,
+      "loss": 1.2271,
+      "step": 2445
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.732674018835615,
+      "learning_rate": 4.60093896713615e-06,
+      "loss": 1.2335,
+      "step": 2450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.631415654924684,
+      "learning_rate": 4.6103286384976535e-06,
+      "loss": 1.2299,
+      "step": 2455
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 34.927007022453836,
+      "learning_rate": 4.619718309859155e-06,
+      "loss": 1.1715,
+      "step": 2460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 22.940986170699468,
+      "learning_rate": 4.6291079812206575e-06,
+      "loss": 1.1948,
+      "step": 2465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.41628824982717,
+      "learning_rate": 4.63849765258216e-06,
+      "loss": 1.1882,
+      "step": 2470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.632035301928894,
+      "learning_rate": 4.647887323943662e-06,
+      "loss": 1.2135,
+      "step": 2475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.609853090326421,
+      "learning_rate": 4.657276995305165e-06,
+      "loss": 1.2727,
+      "step": 2480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.105528217898113,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.1884,
+      "step": 2485
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 26.294778056136835,
+      "learning_rate": 4.67605633802817e-06,
+      "loss": 1.1862,
+      "step": 2490
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.718429733984017,
+      "learning_rate": 4.685446009389672e-06,
+      "loss": 1.169,
+      "step": 2495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.600669653065827,
+      "learning_rate": 4.694835680751174e-06,
+      "loss": 1.1823,
+      "step": 2500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 40.15389187905186,
+      "learning_rate": 4.704225352112676e-06,
+      "loss": 1.1924,
+      "step": 2505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 23.998923051536682,
+      "learning_rate": 4.713615023474179e-06,
+      "loss": 1.2393,
+      "step": 2510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.442591209952589,
+      "learning_rate": 4.723004694835681e-06,
+      "loss": 1.1873,
+      "step": 2515
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.96855325104703,
+      "learning_rate": 4.732394366197183e-06,
+      "loss": 1.2309,
+      "step": 2520
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 29.129323397220617,
+      "learning_rate": 4.741784037558686e-06,
+      "loss": 1.1697,
+      "step": 2525
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.672294663258253,
+      "learning_rate": 4.751173708920188e-06,
+      "loss": 1.2323,
+      "step": 2530
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.913564909324933,
+      "learning_rate": 4.7605633802816905e-06,
+      "loss": 1.237,
+      "step": 2535
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.38302347120668,
+      "learning_rate": 4.769953051643193e-06,
+      "loss": 1.1793,
+      "step": 2540
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.257323163907138,
+      "learning_rate": 4.779342723004695e-06,
+      "loss": 1.178,
+      "step": 2545
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.812584778699634,
+      "learning_rate": 4.788732394366197e-06,
+      "loss": 1.147,
+      "step": 2550
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.65982397347087,
+      "learning_rate": 4.7981220657277e-06,
+      "loss": 1.1151,
+      "step": 2555
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.73998551226896,
+      "learning_rate": 4.8075117370892026e-06,
+      "loss": 1.2453,
+      "step": 2560
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 28.096313844818898,
+      "learning_rate": 4.816901408450705e-06,
+      "loss": 1.1797,
+      "step": 2565
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.786439715810394,
+      "learning_rate": 4.826291079812207e-06,
+      "loss": 1.2389,
+      "step": 2570
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.557288723402566,
+      "learning_rate": 4.835680751173709e-06,
+      "loss": 1.1726,
+      "step": 2575
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.217934859465894,
+      "learning_rate": 4.845070422535212e-06,
+      "loss": 1.2189,
+      "step": 2580
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.935243637488893,
+      "learning_rate": 4.854460093896714e-06,
+      "loss": 1.1678,
+      "step": 2585
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 62.22877930390806,
+      "learning_rate": 4.863849765258216e-06,
+      "loss": 1.19,
+      "step": 2590
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 23.008126923573162,
+      "learning_rate": 4.873239436619719e-06,
+      "loss": 1.2024,
+      "step": 2595
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 49.50461424928968,
+      "learning_rate": 4.882629107981221e-06,
+      "loss": 1.2202,
+      "step": 2600
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 31.757856449655563,
+      "learning_rate": 4.8920187793427235e-06,
+      "loss": 1.2014,
+      "step": 2605
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 17.261547220674743,
+      "learning_rate": 4.901408450704226e-06,
+      "loss": 1.2066,
+      "step": 2610
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 18.268289450107485,
+      "learning_rate": 4.910798122065728e-06,
+      "loss": 1.186,
+      "step": 2615
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.307824070308255,
+      "learning_rate": 4.92018779342723e-06,
+      "loss": 1.189,
+      "step": 2620
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.637796115657476,
+      "learning_rate": 4.929577464788733e-06,
+      "loss": 1.1956,
+      "step": 2625
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.755645306820854,
+      "learning_rate": 4.938967136150235e-06,
+      "loss": 1.1665,
+      "step": 2630
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.468304785143085,
+      "learning_rate": 4.948356807511738e-06,
+      "loss": 1.2118,
+      "step": 2635
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.799015010559174,
+      "learning_rate": 4.9577464788732395e-06,
+      "loss": 1.2311,
+      "step": 2640
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.87197709226322,
+      "learning_rate": 4.967136150234742e-06,
+      "loss": 1.2427,
+      "step": 2645
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.099632777132724,
+      "learning_rate": 4.976525821596244e-06,
+      "loss": 1.1823,
+      "step": 2650
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 34.29335759964042,
+      "learning_rate": 4.985915492957747e-06,
+      "loss": 1.2207,
+      "step": 2655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 21.877054854846023,
+      "learning_rate": 4.995305164319249e-06,
+      "loss": 1.1786,
+      "step": 2660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.579004116388244,
+      "learning_rate": 5.004694835680752e-06,
+      "loss": 1.215,
+      "step": 2665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.894708596586472,
+      "learning_rate": 5.014084507042254e-06,
+      "loss": 1.1824,
+      "step": 2670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.768192973718882,
+      "learning_rate": 5.0234741784037565e-06,
+      "loss": 1.1863,
+      "step": 2675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.184374234203284,
+      "learning_rate": 5.032863849765258e-06,
+      "loss": 1.1987,
+      "step": 2680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 27.69630520473007,
+      "learning_rate": 5.042253521126761e-06,
+      "loss": 1.2174,
+      "step": 2685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.14666148453873,
+      "learning_rate": 5.051643192488264e-06,
+      "loss": 1.1656,
+      "step": 2690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 51.86245200534382,
+      "learning_rate": 5.061032863849765e-06,
+      "loss": 1.2327,
+      "step": 2695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.08166458034953,
+      "learning_rate": 5.070422535211268e-06,
+      "loss": 1.1426,
+      "step": 2700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 30.672451641803097,
+      "learning_rate": 5.079812206572771e-06,
+      "loss": 1.228,
+      "step": 2705
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.681858098471906,
+      "learning_rate": 5.0892018779342725e-06,
+      "loss": 1.2094,
+      "step": 2710
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.582748731923179,
+      "learning_rate": 5.098591549295775e-06,
+      "loss": 1.1459,
+      "step": 2715
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.526169998045273,
+      "learning_rate": 5.107981220657277e-06,
+      "loss": 1.1561,
+      "step": 2720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.182816069511833,
+      "learning_rate": 5.117370892018779e-06,
+      "loss": 1.1646,
+      "step": 2725
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.13347585327498,
+      "learning_rate": 5.126760563380282e-06,
+      "loss": 1.1673,
+      "step": 2730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.92383247295312,
+      "learning_rate": 5.136150234741785e-06,
+      "loss": 1.0964,
+      "step": 2735
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 25.238958262104113,
+      "learning_rate": 5.145539906103287e-06,
+      "loss": 1.126,
+      "step": 2740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.727915817681522,
+      "learning_rate": 5.154929577464789e-06,
+      "loss": 1.184,
+      "step": 2745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.07815234664372,
+      "learning_rate": 5.164319248826292e-06,
+      "loss": 1.1962,
+      "step": 2750
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.580965380449397,
+      "learning_rate": 5.173708920187794e-06,
+      "loss": 1.1812,
+      "step": 2755
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.957695684884207,
+      "learning_rate": 5.183098591549296e-06,
+      "loss": 1.2402,
+      "step": 2760
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.005195755748117,
+      "learning_rate": 5.192488262910798e-06,
+      "loss": 1.1686,
+      "step": 2765
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.136526342545483,
+      "learning_rate": 5.2018779342723015e-06,
+      "loss": 1.1726,
+      "step": 2770
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.085901815839362,
+      "learning_rate": 5.211267605633803e-06,
+      "loss": 1.1847,
+      "step": 2775
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 21.964623577454045,
+      "learning_rate": 5.2206572769953055e-06,
+      "loss": 1.2073,
+      "step": 2780
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 34.18386913088435,
+      "learning_rate": 5.230046948356809e-06,
+      "loss": 1.2156,
+      "step": 2785
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.006982124236956,
+      "learning_rate": 5.23943661971831e-06,
+      "loss": 1.2176,
+      "step": 2790
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 29.671873936742475,
+      "learning_rate": 5.248826291079813e-06,
+      "loss": 1.1959,
+      "step": 2795
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.956018503582134,
+      "learning_rate": 5.258215962441315e-06,
+      "loss": 1.2051,
+      "step": 2800
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.745510165265399,
+      "learning_rate": 5.267605633802817e-06,
+      "loss": 1.1691,
+      "step": 2805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 17.64201493072601,
+      "learning_rate": 5.27699530516432e-06,
+      "loss": 1.2276,
+      "step": 2810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.769176937775635,
+      "learning_rate": 5.286384976525822e-06,
+      "loss": 1.235,
+      "step": 2815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.63246771406025,
+      "learning_rate": 5.295774647887324e-06,
+      "loss": 1.2138,
+      "step": 2820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.948650740908132,
+      "learning_rate": 5.305164319248826e-06,
+      "loss": 1.1787,
+      "step": 2825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.360167189109927,
+      "learning_rate": 5.31455399061033e-06,
+      "loss": 1.2318,
+      "step": 2830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 30.88292172945298,
+      "learning_rate": 5.323943661971831e-06,
+      "loss": 1.2123,
+      "step": 2835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 37.37611584358783,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.1862,
+      "step": 2840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.77094118061188,
+      "learning_rate": 5.342723004694836e-06,
+      "loss": 1.1575,
+      "step": 2845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 79.92042130900299,
+      "learning_rate": 5.352112676056338e-06,
+      "loss": 1.2249,
+      "step": 2850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.350216749270025,
+      "learning_rate": 5.361502347417841e-06,
+      "loss": 1.1811,
+      "step": 2855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 61.20128382809511,
+      "learning_rate": 5.370892018779343e-06,
+      "loss": 1.2,
+      "step": 2860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 15.102967578060602,
+      "learning_rate": 5.380281690140845e-06,
+      "loss": 1.1781,
+      "step": 2865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.080028854178368,
+      "learning_rate": 5.389671361502347e-06,
+      "loss": 1.2154,
+      "step": 2870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.67126903643239,
+      "learning_rate": 5.3990610328638506e-06,
+      "loss": 1.2048,
+      "step": 2875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 59.06522160392222,
+      "learning_rate": 5.408450704225352e-06,
+      "loss": 1.2839,
+      "step": 2880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.612084203697622,
+      "learning_rate": 5.4178403755868546e-06,
+      "loss": 1.1689,
+      "step": 2885
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 34.25826185544695,
+      "learning_rate": 5.427230046948358e-06,
+      "loss": 1.1436,
+      "step": 2890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.800326936769686,
+      "learning_rate": 5.43661971830986e-06,
+      "loss": 1.2112,
+      "step": 2895
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 23.10092238106868,
+      "learning_rate": 5.446009389671362e-06,
+      "loss": 1.1718,
+      "step": 2900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 30.954692739996624,
+      "learning_rate": 5.455399061032864e-06,
+      "loss": 1.1675,
+      "step": 2905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.944793887450846,
+      "learning_rate": 5.4647887323943675e-06,
+      "loss": 1.1885,
+      "step": 2910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 44.80214963552076,
+      "learning_rate": 5.474178403755869e-06,
+      "loss": 1.2333,
+      "step": 2915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.42679087897064,
+      "learning_rate": 5.4835680751173715e-06,
+      "loss": 1.1682,
+      "step": 2920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 37.40246134344992,
+      "learning_rate": 5.492957746478874e-06,
+      "loss": 1.1665,
+      "step": 2925
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.787998396878661,
+      "learning_rate": 5.5023474178403755e-06,
+      "loss": 1.2112,
+      "step": 2930
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.379855721125532,
+      "learning_rate": 5.511737089201879e-06,
+      "loss": 1.1584,
+      "step": 2935
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 24.23506920135781,
+      "learning_rate": 5.521126760563381e-06,
+      "loss": 1.1276,
+      "step": 2940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.665725698283438,
+      "learning_rate": 5.530516431924883e-06,
+      "loss": 1.1764,
+      "step": 2945
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.039102772157442,
+      "learning_rate": 5.539906103286385e-06,
+      "loss": 1.1508,
+      "step": 2950
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.457190549365503,
+      "learning_rate": 5.549295774647888e-06,
+      "loss": 1.1823,
+      "step": 2955
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.476738254085967,
+      "learning_rate": 5.55868544600939e-06,
+      "loss": 1.2079,
+      "step": 2960
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.458033287032747,
+      "learning_rate": 5.568075117370892e-06,
+      "loss": 1.1699,
+      "step": 2965
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.712976133528741,
+      "learning_rate": 5.577464788732395e-06,
+      "loss": 1.1868,
+      "step": 2970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.417131558431354,
+      "learning_rate": 5.586854460093896e-06,
+      "loss": 1.1654,
+      "step": 2975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 28.59571408785586,
+      "learning_rate": 5.5962441314554e-06,
+      "loss": 1.2117,
+      "step": 2980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.036729284240412,
+      "learning_rate": 5.605633802816902e-06,
+      "loss": 1.17,
+      "step": 2985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.856664652849746,
+      "learning_rate": 5.615023474178404e-06,
+      "loss": 1.2174,
+      "step": 2990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.003741305513707,
+      "learning_rate": 5.624413145539907e-06,
+      "loss": 1.1538,
+      "step": 2995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.1943600029335,
+      "learning_rate": 5.633802816901409e-06,
+      "loss": 1.2039,
+      "step": 3000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 31.453900964586012,
+      "learning_rate": 5.643192488262911e-06,
+      "loss": 1.2086,
+      "step": 3005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.353598012487927,
+      "learning_rate": 5.652582159624413e-06,
+      "loss": 1.1823,
+      "step": 3010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.542089610369638,
+      "learning_rate": 5.6619718309859165e-06,
+      "loss": 1.1715,
+      "step": 3015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 22.616104263459135,
+      "learning_rate": 5.671361502347418e-06,
+      "loss": 1.1886,
+      "step": 3020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.306014250423694,
+      "learning_rate": 5.6807511737089205e-06,
+      "loss": 1.1822,
+      "step": 3025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.440575640440233,
+      "learning_rate": 5.690140845070423e-06,
+      "loss": 1.1809,
+      "step": 3030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.727589981071375,
+      "learning_rate": 5.699530516431926e-06,
+      "loss": 1.2242,
+      "step": 3035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.2126417115623,
+      "learning_rate": 5.708920187793428e-06,
+      "loss": 1.242,
+      "step": 3040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.82700181003469,
+      "learning_rate": 5.71830985915493e-06,
+      "loss": 1.2324,
+      "step": 3045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 28.032261575246643,
+      "learning_rate": 5.727699530516433e-06,
+      "loss": 1.1834,
+      "step": 3050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 18.608062314257065,
+      "learning_rate": 5.737089201877934e-06,
+      "loss": 1.1929,
+      "step": 3055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 18.57639194191237,
+      "learning_rate": 5.7464788732394374e-06,
+      "loss": 1.1984,
+      "step": 3060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 22.635698991745947,
+      "learning_rate": 5.75586854460094e-06,
+      "loss": 1.1496,
+      "step": 3065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.601298823552913,
+      "learning_rate": 5.7652582159624414e-06,
+      "loss": 1.2219,
+      "step": 3070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 34.14981338727205,
+      "learning_rate": 5.774647887323944e-06,
+      "loss": 1.2354,
+      "step": 3075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 25.367596079631312,
+      "learning_rate": 5.784037558685447e-06,
+      "loss": 1.2353,
+      "step": 3080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.420422834221338,
+      "learning_rate": 5.793427230046949e-06,
+      "loss": 1.2228,
+      "step": 3085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.91873655885533,
+      "learning_rate": 5.802816901408451e-06,
+      "loss": 1.1709,
+      "step": 3090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.844350103128894,
+      "learning_rate": 5.8122065727699535e-06,
+      "loss": 1.1639,
+      "step": 3095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.342796919264774,
+      "learning_rate": 5.821596244131456e-06,
+      "loss": 1.1326,
+      "step": 3100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 18.855294721496623,
+      "learning_rate": 5.830985915492958e-06,
+      "loss": 1.2328,
+      "step": 3105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.427431240511645,
+      "learning_rate": 5.840375586854461e-06,
+      "loss": 1.2116,
+      "step": 3110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 45.723532604141056,
+      "learning_rate": 5.849765258215962e-06,
+      "loss": 1.1256,
+      "step": 3115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 33.542083457268305,
+      "learning_rate": 5.859154929577466e-06,
+      "loss": 1.1882,
+      "step": 3120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.129904502261516,
+      "learning_rate": 5.868544600938968e-06,
+      "loss": 1.162,
+      "step": 3125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 18.476932976901118,
+      "learning_rate": 5.87793427230047e-06,
+      "loss": 1.1973,
+      "step": 3130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.902035615766474,
+      "learning_rate": 5.887323943661972e-06,
+      "loss": 1.1472,
+      "step": 3135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.704225491764763,
+      "learning_rate": 5.896713615023475e-06,
+      "loss": 1.2177,
+      "step": 3140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.953895764161967,
+      "learning_rate": 5.906103286384977e-06,
+      "loss": 1.2006,
+      "step": 3145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.109934859779493,
+      "learning_rate": 5.915492957746479e-06,
+      "loss": 1.1376,
+      "step": 3150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.455425666662528,
+      "learning_rate": 5.924882629107982e-06,
+      "loss": 1.14,
+      "step": 3155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.751153243299727,
+      "learning_rate": 5.934272300469483e-06,
+      "loss": 1.1672,
+      "step": 3160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.578442734134143,
+      "learning_rate": 5.9436619718309865e-06,
+      "loss": 1.1984,
+      "step": 3165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.27194019530656,
+      "learning_rate": 5.953051643192489e-06,
+      "loss": 1.1612,
+      "step": 3170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.663340387913973,
+      "learning_rate": 5.9624413145539905e-06,
+      "loss": 1.1789,
+      "step": 3175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.588367068745,
+      "learning_rate": 5.971830985915493e-06,
+      "loss": 1.1218,
+      "step": 3180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.293006787691757,
+      "learning_rate": 5.981220657276996e-06,
+      "loss": 1.2613,
+      "step": 3185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.822416797957528,
+      "learning_rate": 5.990610328638499e-06,
+      "loss": 1.1899,
+      "step": 3190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.880043593701103,
+      "learning_rate": 6e-06,
+      "loss": 1.2044,
+      "step": 3195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.928209500542499,
+      "learning_rate": 6.0093896713615026e-06,
+      "loss": 1.232,
+      "step": 3200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 25.473220694504985,
+      "learning_rate": 6.018779342723006e-06,
+      "loss": 1.1904,
+      "step": 3205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.182348972445453,
+      "learning_rate": 6.028169014084507e-06,
+      "loss": 1.1145,
+      "step": 3210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.93176421759847,
+      "learning_rate": 6.03755868544601e-06,
+      "loss": 1.2053,
+      "step": 3215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 33.311478654152445,
+      "learning_rate": 6.046948356807513e-06,
+      "loss": 1.23,
+      "step": 3220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 34.910563964566165,
+      "learning_rate": 6.056338028169015e-06,
+      "loss": 1.2042,
+      "step": 3225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.305916726699248,
+      "learning_rate": 6.065727699530517e-06,
+      "loss": 1.2107,
+      "step": 3230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.925555694783906,
+      "learning_rate": 6.0751173708920195e-06,
+      "loss": 1.1915,
+      "step": 3235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.72001978234508,
+      "learning_rate": 6.084507042253521e-06,
+      "loss": 1.1603,
+      "step": 3240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.731287901642945,
+      "learning_rate": 6.093896713615024e-06,
+      "loss": 1.2402,
+      "step": 3245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.023217803642202,
+      "learning_rate": 6.103286384976527e-06,
+      "loss": 1.1384,
+      "step": 3250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 43.72845432383127,
+      "learning_rate": 6.112676056338028e-06,
+      "loss": 1.1864,
+      "step": 3255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.88664324844327,
+      "learning_rate": 6.122065727699531e-06,
+      "loss": 1.1742,
+      "step": 3260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 22.062745353495536,
+      "learning_rate": 6.131455399061034e-06,
+      "loss": 1.2107,
+      "step": 3265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 59.88662294125921,
+      "learning_rate": 6.1408450704225356e-06,
+      "loss": 1.2874,
+      "step": 3270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 29.659605027540636,
+      "learning_rate": 6.150234741784038e-06,
+      "loss": 1.2323,
+      "step": 3275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.246236154429443,
+      "learning_rate": 6.15962441314554e-06,
+      "loss": 1.1871,
+      "step": 3280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.487902777975865,
+      "learning_rate": 6.169014084507042e-06,
+      "loss": 1.1903,
+      "step": 3285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.190816034075878,
+      "learning_rate": 6.178403755868545e-06,
+      "loss": 1.1471,
+      "step": 3290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.895203886752387,
+      "learning_rate": 6.187793427230048e-06,
+      "loss": 1.1858,
+      "step": 3295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.52950883704903,
+      "learning_rate": 6.197183098591549e-06,
+      "loss": 1.2423,
+      "step": 3300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.835986404636852,
+      "learning_rate": 6.206572769953052e-06,
+      "loss": 1.1984,
+      "step": 3305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.283584871546616,
+      "learning_rate": 6.215962441314555e-06,
+      "loss": 1.1808,
+      "step": 3310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.592837492965831,
+      "learning_rate": 6.2253521126760565e-06,
+      "loss": 1.1685,
+      "step": 3315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.482807347446647,
+      "learning_rate": 6.234741784037559e-06,
+      "loss": 1.1492,
+      "step": 3320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.836520719453812,
+      "learning_rate": 6.244131455399062e-06,
+      "loss": 1.174,
+      "step": 3325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.145387454025895,
+      "learning_rate": 6.2535211267605646e-06,
+      "loss": 1.1839,
+      "step": 3330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.74980245451608,
+      "learning_rate": 6.262910798122066e-06,
+      "loss": 1.213,
+      "step": 3335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.0051182281824,
+      "learning_rate": 6.2723004694835685e-06,
+      "loss": 1.1656,
+      "step": 3340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 26.544198001647317,
+      "learning_rate": 6.281690140845072e-06,
+      "loss": 1.2225,
+      "step": 3345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 21.05461943725882,
+      "learning_rate": 6.291079812206573e-06,
+      "loss": 1.1576,
+      "step": 3350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 46.837709334348844,
+      "learning_rate": 6.300469483568076e-06,
+      "loss": 1.2523,
+      "step": 3355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.130518117204725,
+      "learning_rate": 6.309859154929578e-06,
+      "loss": 1.1609,
+      "step": 3360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.90081126517263,
+      "learning_rate": 6.31924882629108e-06,
+      "loss": 1.1375,
+      "step": 3365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.902814665858642,
+      "learning_rate": 6.328638497652583e-06,
+      "loss": 1.21,
+      "step": 3370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.0678477939345,
+      "learning_rate": 6.3380281690140855e-06,
+      "loss": 1.158,
+      "step": 3375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.45480025089843,
+      "learning_rate": 6.347417840375587e-06,
+      "loss": 1.2198,
+      "step": 3380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.318387694667846,
+      "learning_rate": 6.3568075117370894e-06,
+      "loss": 1.1965,
+      "step": 3385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.84241615890925,
+      "learning_rate": 6.366197183098593e-06,
+      "loss": 1.1983,
+      "step": 3390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.25212373036805,
+      "learning_rate": 6.375586854460094e-06,
+      "loss": 1.1699,
+      "step": 3395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.21689707128868,
+      "learning_rate": 6.384976525821597e-06,
+      "loss": 1.2123,
+      "step": 3400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.160465138136907,
+      "learning_rate": 6.394366197183099e-06,
+      "loss": 1.1336,
+      "step": 3405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.25062850522987,
+      "learning_rate": 6.403755868544601e-06,
+      "loss": 1.1507,
+      "step": 3410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.759040977103274,
+      "learning_rate": 6.413145539906104e-06,
+      "loss": 1.2328,
+      "step": 3415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.202392125250093,
+      "learning_rate": 6.422535211267606e-06,
+      "loss": 1.1934,
+      "step": 3420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 22.395804601771438,
+      "learning_rate": 6.431924882629108e-06,
+      "loss": 1.1812,
+      "step": 3425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.776606046029634,
+      "learning_rate": 6.441314553990611e-06,
+      "loss": 1.1571,
+      "step": 3430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 25.10005593935468,
+      "learning_rate": 6.450704225352114e-06,
+      "loss": 1.2581,
+      "step": 3435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.36646332934216,
+      "learning_rate": 6.460093896713615e-06,
+      "loss": 1.2021,
+      "step": 3440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.836089777946722,
+      "learning_rate": 6.469483568075118e-06,
+      "loss": 1.173,
+      "step": 3445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.904518395390559,
+      "learning_rate": 6.478873239436621e-06,
+      "loss": 1.1393,
+      "step": 3450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.856781617568297,
+      "learning_rate": 6.4882629107981224e-06,
+      "loss": 1.1876,
+      "step": 3455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.892170004163315,
+      "learning_rate": 6.497652582159625e-06,
+      "loss": 1.1603,
+      "step": 3460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.992435890958625,
+      "learning_rate": 6.507042253521127e-06,
+      "loss": 1.1057,
+      "step": 3465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.727581752017006,
+      "learning_rate": 6.516431924882629e-06,
+      "loss": 1.1899,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.205494216931335,
+      "learning_rate": 6.525821596244132e-06,
+      "loss": 1.1475,
+      "step": 3475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 21.09712047658045,
+      "learning_rate": 6.5352112676056345e-06,
+      "loss": 1.2599,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.585943164007832,
+      "learning_rate": 6.544600938967137e-06,
+      "loss": 1.1485,
+      "step": 3485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.70113802638682,
+      "learning_rate": 6.5539906103286385e-06,
+      "loss": 1.1576,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.476299406363472,
+      "learning_rate": 6.563380281690142e-06,
+      "loss": 1.2452,
+      "step": 3495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 35.03789912898917,
+      "learning_rate": 6.572769953051644e-06,
+      "loss": 1.161,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 27.36668377715275,
+      "learning_rate": 6.582159624413146e-06,
+      "loss": 1.1931,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.136014877955109,
+      "learning_rate": 6.591549295774648e-06,
+      "loss": 1.161,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.84250751227113,
+      "learning_rate": 6.6009389671361514e-06,
+      "loss": 1.1951,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.378067004203,
+      "learning_rate": 6.610328638497653e-06,
+      "loss": 1.1747,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.81809857296866,
+      "learning_rate": 6.619718309859155e-06,
+      "loss": 1.1838,
+      "step": 3525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.566027578446029,
+      "learning_rate": 6.629107981220658e-06,
+      "loss": 1.2146,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 22.616110729334892,
+      "learning_rate": 6.63849765258216e-06,
+      "loss": 1.217,
+      "step": 3535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.777553651375694,
+      "learning_rate": 6.647887323943663e-06,
+      "loss": 1.1952,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.766432365714778,
+      "learning_rate": 6.657276995305165e-06,
+      "loss": 1.1913,
+      "step": 3545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 23.87892511875046,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.1854,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.134182584773212,
+      "learning_rate": 6.67605633802817e-06,
+      "loss": 1.2105,
+      "step": 3555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.472091344972434,
+      "learning_rate": 6.685446009389672e-06,
+      "loss": 1.17,
+      "step": 3560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.822251781811772,
+      "learning_rate": 6.694835680751174e-06,
+      "loss": 1.1933,
+      "step": 3565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.129273978116814,
+      "learning_rate": 6.704225352112676e-06,
+      "loss": 1.2379,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 23.817218249527652,
+      "learning_rate": 6.71361502347418e-06,
+      "loss": 1.1708,
+      "step": 3575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 28.537208165380388,
+      "learning_rate": 6.723004694835681e-06,
+      "loss": 1.161,
+      "step": 3580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.494696977137288,
+      "learning_rate": 6.7323943661971836e-06,
+      "loss": 1.1503,
+      "step": 3585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 68.31195592819469,
+      "learning_rate": 6.741784037558686e-06,
+      "loss": 1.243,
+      "step": 3590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 40.88593787189487,
+      "learning_rate": 6.7511737089201875e-06,
+      "loss": 1.2053,
+      "step": 3595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.129499892476897,
+      "learning_rate": 6.760563380281691e-06,
+      "loss": 1.2732,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 42.19342123857229,
+      "learning_rate": 6.769953051643193e-06,
+      "loss": 1.1678,
+      "step": 3605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.406888645661365,
+      "learning_rate": 6.779342723004695e-06,
+      "loss": 1.2387,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 24.730586011039748,
+      "learning_rate": 6.788732394366197e-06,
+      "loss": 1.2213,
+      "step": 3615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.977379864150745,
+      "learning_rate": 6.7981220657277005e-06,
+      "loss": 1.1901,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 61.8058816801069,
+      "learning_rate": 6.807511737089203e-06,
+      "loss": 1.1543,
+      "step": 3625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 29.135776663159877,
+      "learning_rate": 6.8169014084507045e-06,
+      "loss": 1.2084,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.69234399232136,
+      "learning_rate": 6.826291079812207e-06,
+      "loss": 1.1876,
+      "step": 3635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.404150099415487,
+      "learning_rate": 6.83568075117371e-06,
+      "loss": 1.1912,
+      "step": 3640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.070747134810834,
+      "learning_rate": 6.845070422535212e-06,
+      "loss": 1.2465,
+      "step": 3645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.393409730076835,
+      "learning_rate": 6.854460093896714e-06,
+      "loss": 1.1301,
+      "step": 3650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.319997724780428,
+      "learning_rate": 6.863849765258217e-06,
+      "loss": 1.1826,
+      "step": 3655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.14056625190511,
+      "learning_rate": 6.873239436619719e-06,
+      "loss": 1.188,
+      "step": 3660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.06116512876983,
+      "learning_rate": 6.882629107981221e-06,
+      "loss": 1.1528,
+      "step": 3665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.17917024296913,
+      "learning_rate": 6.892018779342724e-06,
+      "loss": 1.2301,
+      "step": 3670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 39.495521701693804,
+      "learning_rate": 6.901408450704225e-06,
+      "loss": 1.2008,
+      "step": 3675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 33.18895419124601,
+      "learning_rate": 6.910798122065729e-06,
+      "loss": 1.2165,
+      "step": 3680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 28.62316229900056,
+      "learning_rate": 6.920187793427231e-06,
+      "loss": 1.2534,
+      "step": 3685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 54.32211781057552,
+      "learning_rate": 6.929577464788733e-06,
+      "loss": 1.2292,
+      "step": 3690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 56.82054736290158,
+      "learning_rate": 6.938967136150235e-06,
+      "loss": 1.2068,
+      "step": 3695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.029008545931667,
+      "learning_rate": 6.948356807511738e-06,
+      "loss": 1.2286,
+      "step": 3700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.402768156419866,
+      "learning_rate": 6.95774647887324e-06,
+      "loss": 1.2406,
+      "step": 3705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.536206448928514,
+      "learning_rate": 6.967136150234742e-06,
+      "loss": 1.1751,
+      "step": 3710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.748296040716767,
+      "learning_rate": 6.976525821596245e-06,
+      "loss": 1.2088,
+      "step": 3715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.62386252533626,
+      "learning_rate": 6.985915492957746e-06,
+      "loss": 1.2141,
+      "step": 3720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 30.942116402328708,
+      "learning_rate": 6.9953051643192495e-06,
+      "loss": 1.2274,
+      "step": 3725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 40.26536107697977,
+      "learning_rate": 7.004694835680752e-06,
+      "loss": 1.2088,
+      "step": 3730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.619462153605463,
+      "learning_rate": 7.0140845070422535e-06,
+      "loss": 1.1983,
+      "step": 3735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 20.964427584238834,
+      "learning_rate": 7.023474178403756e-06,
+      "loss": 1.238,
+      "step": 3740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 19.65296109542318,
+      "learning_rate": 7.032863849765259e-06,
+      "loss": 1.1932,
+      "step": 3745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.22501001260899,
+      "learning_rate": 7.042253521126761e-06,
+      "loss": 1.1857,
+      "step": 3750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.360921962093933,
+      "learning_rate": 7.051643192488263e-06,
+      "loss": 1.1414,
+      "step": 3755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.304445795952342,
+      "learning_rate": 7.0610328638497664e-06,
+      "loss": 1.2094,
+      "step": 3760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.736799655817833,
+      "learning_rate": 7.070422535211268e-06,
+      "loss": 1.1651,
+      "step": 3765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.340443257491685,
+      "learning_rate": 7.0798122065727704e-06,
+      "loss": 1.2106,
+      "step": 3770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 57.75088553318209,
+      "learning_rate": 7.089201877934273e-06,
+      "loss": 1.2133,
+      "step": 3775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.102570078350915,
+      "learning_rate": 7.098591549295776e-06,
+      "loss": 1.206,
+      "step": 3780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.34966176346707,
+      "learning_rate": 7.107981220657278e-06,
+      "loss": 1.2293,
+      "step": 3785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.013929238073546,
+      "learning_rate": 7.11737089201878e-06,
+      "loss": 1.244,
+      "step": 3790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.017428219682708,
+      "learning_rate": 7.1267605633802825e-06,
+      "loss": 1.2131,
+      "step": 3795
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 38.2059808102528,
+      "learning_rate": 7.136150234741784e-06,
+      "loss": 1.1732,
+      "step": 3800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.18166957051223,
+      "learning_rate": 7.145539906103287e-06,
+      "loss": 1.1776,
+      "step": 3805
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.630825940208817,
+      "learning_rate": 7.15492957746479e-06,
+      "loss": 1.2183,
+      "step": 3810
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.973979962381074,
+      "learning_rate": 7.164319248826291e-06,
+      "loss": 1.2085,
+      "step": 3815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 18.371477416055765,
+      "learning_rate": 7.173708920187794e-06,
+      "loss": 1.2123,
+      "step": 3820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 22.95078187102011,
+      "learning_rate": 7.183098591549297e-06,
+      "loss": 1.2317,
+      "step": 3825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.633831261862134,
+      "learning_rate": 7.192488262910799e-06,
+      "loss": 1.1827,
+      "step": 3830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 34.9900080578578,
+      "learning_rate": 7.201877934272301e-06,
+      "loss": 1.1642,
+      "step": 3835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 31.593510530964537,
+      "learning_rate": 7.211267605633803e-06,
+      "loss": 1.1815,
+      "step": 3840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.554527659078145,
+      "learning_rate": 7.220657276995305e-06,
+      "loss": 1.2266,
+      "step": 3845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 27.523021720743884,
+      "learning_rate": 7.230046948356808e-06,
+      "loss": 1.195,
+      "step": 3850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 84.84982009791756,
+      "learning_rate": 7.239436619718311e-06,
+      "loss": 1.2443,
+      "step": 3855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.432189680659697,
+      "learning_rate": 7.248826291079812e-06,
+      "loss": 1.165,
+      "step": 3860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.854220309725344,
+      "learning_rate": 7.2582159624413155e-06,
+      "loss": 1.2511,
+      "step": 3865
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 31.319517601510057,
+      "learning_rate": 7.267605633802818e-06,
+      "loss": 1.2311,
+      "step": 3870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.750601797395683,
+      "learning_rate": 7.2769953051643195e-06,
+      "loss": 1.2083,
+      "step": 3875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.113609084691047,
+      "learning_rate": 7.286384976525822e-06,
+      "loss": 1.1807,
+      "step": 3880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 76.31055625879928,
+      "learning_rate": 7.295774647887325e-06,
+      "loss": 1.22,
+      "step": 3885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 46.4921319101062,
+      "learning_rate": 7.305164319248827e-06,
+      "loss": 1.2519,
+      "step": 3890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 34.476256756362176,
+      "learning_rate": 7.314553990610329e-06,
+      "loss": 1.2219,
+      "step": 3895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 47.99845942276182,
+      "learning_rate": 7.3239436619718316e-06,
+      "loss": 1.1994,
+      "step": 3900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 29.45466850995247,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 1.179,
+      "step": 3905
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.558244967085193,
+      "learning_rate": 7.342723004694836e-06,
+      "loss": 1.1744,
+      "step": 3910
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.34414112113531,
+      "learning_rate": 7.352112676056339e-06,
+      "loss": 1.1872,
+      "step": 3915
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.416129737879526,
+      "learning_rate": 7.361502347417841e-06,
+      "loss": 1.1791,
+      "step": 3920
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.076925129853986,
+      "learning_rate": 7.370892018779343e-06,
+      "loss": 1.2536,
+      "step": 3925
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.459658344319436,
+      "learning_rate": 7.380281690140846e-06,
+      "loss": 1.1564,
+      "step": 3930
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.418908912329496,
+      "learning_rate": 7.3896713615023485e-06,
+      "loss": 1.2002,
+      "step": 3935
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.350250261591546,
+      "learning_rate": 7.39906103286385e-06,
+      "loss": 1.2101,
+      "step": 3940
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.054635065283787,
+      "learning_rate": 7.4084507042253525e-06,
+      "loss": 1.18,
+      "step": 3945
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.66569204829225,
+      "learning_rate": 7.417840375586856e-06,
+      "loss": 1.1737,
+      "step": 3950
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.125203143270163,
+      "learning_rate": 7.427230046948357e-06,
+      "loss": 1.2223,
+      "step": 3955
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 23.73720907789374,
+      "learning_rate": 7.43661971830986e-06,
+      "loss": 1.161,
+      "step": 3960
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.332365920244538,
+      "learning_rate": 7.446009389671362e-06,
+      "loss": 1.2081,
+      "step": 3965
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 41.037071426492446,
+      "learning_rate": 7.4553990610328646e-06,
+      "loss": 1.1583,
+      "step": 3970
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 31.54174964649289,
+      "learning_rate": 7.464788732394367e-06,
+      "loss": 1.1936,
+      "step": 3975
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.638670956253598,
+      "learning_rate": 7.474178403755869e-06,
+      "loss": 1.1246,
+      "step": 3980
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 47.33858300112234,
+      "learning_rate": 7.483568075117371e-06,
+      "loss": 1.2131,
+      "step": 3985
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.125222564039044,
+      "learning_rate": 7.492957746478874e-06,
+      "loss": 1.2149,
+      "step": 3990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.060015672819873,
+      "learning_rate": 7.502347417840377e-06,
+      "loss": 1.1817,
+      "step": 3995
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.717886632697339,
+      "learning_rate": 7.511737089201878e-06,
+      "loss": 1.1792,
+      "step": 4000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 22.735666868633807,
+      "learning_rate": 7.521126760563381e-06,
+      "loss": 1.2003,
+      "step": 4005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.679319560312683,
+      "learning_rate": 7.530516431924884e-06,
+      "loss": 1.1944,
+      "step": 4010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.321591404409501,
+      "learning_rate": 7.5399061032863855e-06,
+      "loss": 1.1637,
+      "step": 4015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.569842640812825,
+      "learning_rate": 7.549295774647888e-06,
+      "loss": 1.1933,
+      "step": 4020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.966539846253095,
+      "learning_rate": 7.55868544600939e-06,
+      "loss": 1.1931,
+      "step": 4025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.298177602548586,
+      "learning_rate": 7.568075117370892e-06,
+      "loss": 1.2147,
+      "step": 4030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.695716274478213,
+      "learning_rate": 7.577464788732395e-06,
+      "loss": 1.1568,
+      "step": 4035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.76956762071812,
+      "learning_rate": 7.5868544600938975e-06,
+      "loss": 1.2203,
+      "step": 4040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.74207568799746,
+      "learning_rate": 7.596244131455399e-06,
+      "loss": 1.1174,
+      "step": 4045
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.746951692900796,
+      "learning_rate": 7.6056338028169015e-06,
+      "loss": 1.2095,
+      "step": 4050
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 25.823990537237783,
+      "learning_rate": 7.615023474178405e-06,
+      "loss": 1.1828,
+      "step": 4055
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.522634108230067,
+      "learning_rate": 7.624413145539906e-06,
+      "loss": 1.219,
+      "step": 4060
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.862682594839754,
+      "learning_rate": 7.633802816901409e-06,
+      "loss": 1.2119,
+      "step": 4065
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 37.50414447404458,
+      "learning_rate": 7.643192488262911e-06,
+      "loss": 1.1571,
+      "step": 4070
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 29.026688227016276,
+      "learning_rate": 7.652582159624414e-06,
+      "loss": 1.1998,
+      "step": 4075
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.831027066123458,
+      "learning_rate": 7.661971830985916e-06,
+      "loss": 1.2157,
+      "step": 4080
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.160512019870453,
+      "learning_rate": 7.671361502347418e-06,
+      "loss": 1.1936,
+      "step": 4085
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.509104337919787,
+      "learning_rate": 7.680751173708921e-06,
+      "loss": 1.2712,
+      "step": 4090
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.264105952348327,
+      "learning_rate": 7.690140845070423e-06,
+      "loss": 1.2473,
+      "step": 4095
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.324056959006983,
+      "learning_rate": 7.699530516431926e-06,
+      "loss": 1.23,
+      "step": 4100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.00961819776812,
+      "learning_rate": 7.708920187793428e-06,
+      "loss": 1.1555,
+      "step": 4105
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.830030393314683,
+      "learning_rate": 7.71830985915493e-06,
+      "loss": 1.1994,
+      "step": 4110
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 24.400722597596747,
+      "learning_rate": 7.727699530516433e-06,
+      "loss": 1.2221,
+      "step": 4115
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.517134784716312,
+      "learning_rate": 7.737089201877935e-06,
+      "loss": 1.2175,
+      "step": 4120
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.65491023895828,
+      "learning_rate": 7.746478873239436e-06,
+      "loss": 1.1672,
+      "step": 4125
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.7455423499664615,
+      "learning_rate": 7.75586854460094e-06,
+      "loss": 1.1662,
+      "step": 4130
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.862721002479682,
+      "learning_rate": 7.765258215962443e-06,
+      "loss": 1.203,
+      "step": 4135
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 21.376653945251896,
+      "learning_rate": 7.774647887323943e-06,
+      "loss": 1.254,
+      "step": 4140
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.882981877205165,
+      "learning_rate": 7.784037558685447e-06,
+      "loss": 1.1537,
+      "step": 4145
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 33.72581253134319,
+      "learning_rate": 7.79342723004695e-06,
+      "loss": 1.2069,
+      "step": 4150
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.53421172681963,
+      "learning_rate": 7.80281690140845e-06,
+      "loss": 1.1842,
+      "step": 4155
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 28.955238680225012,
+      "learning_rate": 7.812206572769953e-06,
+      "loss": 1.1971,
+      "step": 4160
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 53.484123103401174,
+      "learning_rate": 7.821596244131457e-06,
+      "loss": 1.19,
+      "step": 4165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 56.311802418144374,
+      "learning_rate": 7.830985915492958e-06,
+      "loss": 1.2309,
+      "step": 4170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 54.745201532177774,
+      "learning_rate": 7.84037558685446e-06,
+      "loss": 1.2278,
+      "step": 4175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.370907547871159,
+      "learning_rate": 7.849765258215963e-06,
+      "loss": 1.2722,
+      "step": 4180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 46.65810052536865,
+      "learning_rate": 7.859154929577465e-06,
+      "loss": 1.2049,
+      "step": 4185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 31.81831041104849,
+      "learning_rate": 7.868544600938967e-06,
+      "loss": 1.2209,
+      "step": 4190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 32.47601424145794,
+      "learning_rate": 7.87793427230047e-06,
+      "loss": 1.2545,
+      "step": 4195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 31.787518737832805,
+      "learning_rate": 7.887323943661972e-06,
+      "loss": 1.2616,
+      "step": 4200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 26.705372881127936,
+      "learning_rate": 7.896713615023475e-06,
+      "loss": 1.1764,
+      "step": 4205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 42.69334237596915,
+      "learning_rate": 7.906103286384977e-06,
+      "loss": 1.1996,
+      "step": 4210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 27.09679778149449,
+      "learning_rate": 7.91549295774648e-06,
+      "loss": 1.2149,
+      "step": 4215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 32.73831121920157,
+      "learning_rate": 7.924882629107982e-06,
+      "loss": 1.2362,
+      "step": 4220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 21.04835758853057,
+      "learning_rate": 7.934272300469484e-06,
+      "loss": 1.194,
+      "step": 4225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 22.251811150037636,
+      "learning_rate": 7.943661971830987e-06,
+      "loss": 1.1803,
+      "step": 4230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.281136896647453,
+      "learning_rate": 7.95305164319249e-06,
+      "loss": 1.1946,
+      "step": 4235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.126753471815968,
+      "learning_rate": 7.962441314553992e-06,
+      "loss": 1.2082,
+      "step": 4240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 28.520843363187964,
+      "learning_rate": 7.971830985915494e-06,
+      "loss": 1.2242,
+      "step": 4245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.517790290601727,
+      "learning_rate": 7.981220657276996e-06,
+      "loss": 1.2475,
+      "step": 4250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 26.652401187780274,
+      "learning_rate": 7.990610328638499e-06,
+      "loss": 1.2229,
+      "step": 4255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.409431570584443,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.2134,
+      "step": 4260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.12748565038928,
+      "learning_rate": 8.009389671361502e-06,
+      "loss": 1.2264,
+      "step": 4265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.312110984657279,
+      "learning_rate": 8.018779342723006e-06,
+      "loss": 1.182,
+      "step": 4270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 52.16608985314976,
+      "learning_rate": 8.028169014084509e-06,
+      "loss": 1.1355,
+      "step": 4275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.120160867118052,
+      "learning_rate": 8.03755868544601e-06,
+      "loss": 1.2416,
+      "step": 4280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.830831717537642,
+      "learning_rate": 8.046948356807512e-06,
+      "loss": 1.225,
+      "step": 4285
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 43.23822029068536,
+      "learning_rate": 8.056338028169016e-06,
+      "loss": 1.2252,
+      "step": 4290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.278674301492558,
+      "learning_rate": 8.065727699530517e-06,
+      "loss": 1.2109,
+      "step": 4295
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.288013498220494,
+      "learning_rate": 8.075117370892019e-06,
+      "loss": 1.1789,
+      "step": 4300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.26449336507538,
+      "learning_rate": 8.084507042253521e-06,
+      "loss": 1.2125,
+      "step": 4305
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 28.22555996148371,
+      "learning_rate": 8.093896713615024e-06,
+      "loss": 1.2431,
+      "step": 4310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.867982626962476,
+      "learning_rate": 8.103286384976526e-06,
+      "loss": 1.2525,
+      "step": 4315
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.028491223313324,
+      "learning_rate": 8.112676056338029e-06,
+      "loss": 1.1883,
+      "step": 4320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.2026817833496,
+      "learning_rate": 8.122065727699531e-06,
+      "loss": 1.1986,
+      "step": 4325
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 27.554321300631596,
+      "learning_rate": 8.131455399061033e-06,
+      "loss": 1.1934,
+      "step": 4330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.581434368028852,
+      "learning_rate": 8.140845070422536e-06,
+      "loss": 1.2165,
+      "step": 4335
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.80090735090302,
+      "learning_rate": 8.150234741784038e-06,
+      "loss": 1.1882,
+      "step": 4340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.10623653337208,
+      "learning_rate": 8.15962441314554e-06,
+      "loss": 1.2051,
+      "step": 4345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 36.58103521997585,
+      "learning_rate": 8.169014084507043e-06,
+      "loss": 1.165,
+      "step": 4350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.14362401499682,
+      "learning_rate": 8.178403755868546e-06,
+      "loss": 1.1963,
+      "step": 4355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 23.915175786758955,
+      "learning_rate": 8.187793427230048e-06,
+      "loss": 1.1793,
+      "step": 4360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 83.0767168465954,
+      "learning_rate": 8.19718309859155e-06,
+      "loss": 1.217,
+      "step": 4365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 60.3543840448068,
+      "learning_rate": 8.206572769953053e-06,
+      "loss": 1.2637,
+      "step": 4370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 48.38956559363697,
+      "learning_rate": 8.215962441314555e-06,
+      "loss": 1.1145,
+      "step": 4375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.06381918741665,
+      "learning_rate": 8.225352112676058e-06,
+      "loss": 1.2502,
+      "step": 4380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.613602776471925,
+      "learning_rate": 8.23474178403756e-06,
+      "loss": 1.253,
+      "step": 4385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 67.98537964903069,
+      "learning_rate": 8.24413145539906e-06,
+      "loss": 1.2557,
+      "step": 4390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 44.88995190776058,
+      "learning_rate": 8.253521126760565e-06,
+      "loss": 1.2989,
+      "step": 4395
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 28.976224526658328,
+      "learning_rate": 8.262910798122067e-06,
+      "loss": 1.1992,
+      "step": 4400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.817326532997889,
+      "learning_rate": 8.272300469483568e-06,
+      "loss": 1.1621,
+      "step": 4405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.74428012005997,
+      "learning_rate": 8.28169014084507e-06,
+      "loss": 1.1887,
+      "step": 4410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.572505703699525,
+      "learning_rate": 8.291079812206575e-06,
+      "loss": 1.2101,
+      "step": 4415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 25.415921386403607,
+      "learning_rate": 8.300469483568075e-06,
+      "loss": 1.2156,
+      "step": 4420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.61583806645396,
+      "learning_rate": 8.309859154929578e-06,
+      "loss": 1.1939,
+      "step": 4425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.863954359581307,
+      "learning_rate": 8.31924882629108e-06,
+      "loss": 1.193,
+      "step": 4430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 38.64725509136549,
+      "learning_rate": 8.328638497652583e-06,
+      "loss": 1.2823,
+      "step": 4435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.766446850354708,
+      "learning_rate": 8.338028169014085e-06,
+      "loss": 1.1997,
+      "step": 4440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 49.93259454328654,
+      "learning_rate": 8.347417840375587e-06,
+      "loss": 1.2411,
+      "step": 4445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 58.99810320942265,
+      "learning_rate": 8.35680751173709e-06,
+      "loss": 1.2283,
+      "step": 4450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.671889593603145,
+      "learning_rate": 8.366197183098592e-06,
+      "loss": 1.2491,
+      "step": 4455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.782965260071162,
+      "learning_rate": 8.375586854460095e-06,
+      "loss": 1.18,
+      "step": 4460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.14152789516212,
+      "learning_rate": 8.384976525821597e-06,
+      "loss": 1.2081,
+      "step": 4465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.434041552947308,
+      "learning_rate": 8.3943661971831e-06,
+      "loss": 1.1511,
+      "step": 4470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.056764649773962,
+      "learning_rate": 8.403755868544602e-06,
+      "loss": 1.2038,
+      "step": 4475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.02810844754425,
+      "learning_rate": 8.413145539906104e-06,
+      "loss": 1.2548,
+      "step": 4480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 24.719945686128177,
+      "learning_rate": 8.422535211267607e-06,
+      "loss": 1.2403,
+      "step": 4485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 46.08082388148565,
+      "learning_rate": 8.431924882629109e-06,
+      "loss": 1.2058,
+      "step": 4490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 26.814904159754715,
+      "learning_rate": 8.44131455399061e-06,
+      "loss": 1.1815,
+      "step": 4495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 35.51873251767294,
+      "learning_rate": 8.450704225352114e-06,
+      "loss": 1.1935,
+      "step": 4500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.390534518678628,
+      "learning_rate": 8.460093896713616e-06,
+      "loss": 1.261,
+      "step": 4505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.526978010888968,
+      "learning_rate": 8.469483568075117e-06,
+      "loss": 1.1904,
+      "step": 4510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.487866155998266,
+      "learning_rate": 8.47887323943662e-06,
+      "loss": 1.2252,
+      "step": 4515
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 28.22177824337065,
+      "learning_rate": 8.488262910798124e-06,
+      "loss": 1.2275,
+      "step": 4520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 22.611165472988805,
+      "learning_rate": 8.497652582159626e-06,
+      "loss": 1.2108,
+      "step": 4525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.378918799576965,
+      "learning_rate": 8.507042253521127e-06,
+      "loss": 1.2636,
+      "step": 4530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.497233258422098,
+      "learning_rate": 8.51643192488263e-06,
+      "loss": 1.1891,
+      "step": 4535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 22.49226598940317,
+      "learning_rate": 8.525821596244133e-06,
+      "loss": 1.1832,
+      "step": 4540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 19.168432021952608,
+      "learning_rate": 8.535211267605634e-06,
+      "loss": 1.2072,
+      "step": 4545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 26.166076586005975,
+      "learning_rate": 8.544600938967136e-06,
+      "loss": 1.2226,
+      "step": 4550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 32.03204570802435,
+      "learning_rate": 8.553990610328639e-06,
+      "loss": 1.1851,
+      "step": 4555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.092581131426625,
+      "learning_rate": 8.563380281690141e-06,
+      "loss": 1.1864,
+      "step": 4560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.669659448564795,
+      "learning_rate": 8.572769953051644e-06,
+      "loss": 1.2309,
+      "step": 4565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 19.334502010788857,
+      "learning_rate": 8.582159624413146e-06,
+      "loss": 1.2154,
+      "step": 4570
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.563671018584959,
+      "learning_rate": 8.591549295774648e-06,
+      "loss": 1.2385,
+      "step": 4575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.915338118752953,
+      "learning_rate": 8.600938967136151e-06,
+      "loss": 1.1964,
+      "step": 4580
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 26.404055809630883,
+      "learning_rate": 8.610328638497653e-06,
+      "loss": 1.1839,
+      "step": 4585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.222126968100023,
+      "learning_rate": 8.619718309859156e-06,
+      "loss": 1.2581,
+      "step": 4590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.35747409325916,
+      "learning_rate": 8.629107981220658e-06,
+      "loss": 1.2179,
+      "step": 4595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.990756284594863,
+      "learning_rate": 8.63849765258216e-06,
+      "loss": 1.2231,
+      "step": 4600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.257116674804113,
+      "learning_rate": 8.647887323943663e-06,
+      "loss": 1.1873,
+      "step": 4605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 28.933897560856717,
+      "learning_rate": 8.657276995305165e-06,
+      "loss": 1.2032,
+      "step": 4610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.215272714102335,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 1.2329,
+      "step": 4615
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.913604246367948,
+      "learning_rate": 8.676056338028169e-06,
+      "loss": 1.1643,
+      "step": 4620
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.192712631352123,
+      "learning_rate": 8.685446009389673e-06,
+      "loss": 1.206,
+      "step": 4625
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 20.39635890815786,
+      "learning_rate": 8.694835680751175e-06,
+      "loss": 1.2078,
+      "step": 4630
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.542066863681356,
+      "learning_rate": 8.704225352112676e-06,
+      "loss": 1.1961,
+      "step": 4635
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.022672362985384,
+      "learning_rate": 8.713615023474178e-06,
+      "loss": 1.1985,
+      "step": 4640
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.92321634589073,
+      "learning_rate": 8.723004694835682e-06,
+      "loss": 1.2378,
+      "step": 4645
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.357184906029232,
+      "learning_rate": 8.732394366197183e-06,
+      "loss": 1.2266,
+      "step": 4650
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.056095828904821,
+      "learning_rate": 8.741784037558685e-06,
+      "loss": 1.1735,
+      "step": 4655
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.816915017876703,
+      "learning_rate": 8.751173708920188e-06,
+      "loss": 1.1604,
+      "step": 4660
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.612070891630275,
+      "learning_rate": 8.760563380281692e-06,
+      "loss": 1.2305,
+      "step": 4665
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.759724594659053,
+      "learning_rate": 8.769953051643193e-06,
+      "loss": 1.1982,
+      "step": 4670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.514505344613759,
+      "learning_rate": 8.779342723004695e-06,
+      "loss": 1.1367,
+      "step": 4675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.03064258087366,
+      "learning_rate": 8.7887323943662e-06,
+      "loss": 1.2106,
+      "step": 4680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 22.814805988832852,
+      "learning_rate": 8.7981220657277e-06,
+      "loss": 1.1732,
+      "step": 4685
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.906934924379284,
+      "learning_rate": 8.807511737089202e-06,
+      "loss": 1.2103,
+      "step": 4690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 25.346428870841898,
+      "learning_rate": 8.816901408450705e-06,
+      "loss": 1.2158,
+      "step": 4695
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 93.89877917994153,
+      "learning_rate": 8.826291079812207e-06,
+      "loss": 1.2685,
+      "step": 4700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 37.7336374532386,
+      "learning_rate": 8.83568075117371e-06,
+      "loss": 1.2076,
+      "step": 4705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 21.686061239154387,
+      "learning_rate": 8.845070422535212e-06,
+      "loss": 1.1559,
+      "step": 4710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.26863873637512,
+      "learning_rate": 8.854460093896714e-06,
+      "loss": 1.262,
+      "step": 4715
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.399365209935816,
+      "learning_rate": 8.863849765258217e-06,
+      "loss": 1.1888,
+      "step": 4720
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 28.833717861525898,
+      "learning_rate": 8.87323943661972e-06,
+      "loss": 1.193,
+      "step": 4725
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.17236565349994,
+      "learning_rate": 8.882629107981222e-06,
+      "loss": 1.228,
+      "step": 4730
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.49984344740007,
+      "learning_rate": 8.892018779342724e-06,
+      "loss": 1.1805,
+      "step": 4735
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 23.133834074648217,
+      "learning_rate": 8.901408450704227e-06,
+      "loss": 1.231,
+      "step": 4740
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.390859731210462,
+      "learning_rate": 8.910798122065727e-06,
+      "loss": 1.2167,
+      "step": 4745
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.52393152719337,
+      "learning_rate": 8.920187793427231e-06,
+      "loss": 1.2012,
+      "step": 4750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.63053260827425,
+      "learning_rate": 8.929577464788734e-06,
+      "loss": 1.1925,
+      "step": 4755
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 32.73148648410042,
+      "learning_rate": 8.938967136150235e-06,
+      "loss": 1.2605,
+      "step": 4760
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.894352059864662,
+      "learning_rate": 8.948356807511737e-06,
+      "loss": 1.2162,
+      "step": 4765
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.552047195004757,
+      "learning_rate": 8.957746478873241e-06,
+      "loss": 1.2326,
+      "step": 4770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.151321061078818,
+      "learning_rate": 8.967136150234742e-06,
+      "loss": 1.23,
+      "step": 4775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.886873360044373,
+      "learning_rate": 8.976525821596244e-06,
+      "loss": 1.1822,
+      "step": 4780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.398381585266354,
+      "learning_rate": 8.985915492957748e-06,
+      "loss": 1.218,
+      "step": 4785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.457273127763088,
+      "learning_rate": 8.995305164319249e-06,
+      "loss": 1.1844,
+      "step": 4790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.001274637047235,
+      "learning_rate": 9.004694835680751e-06,
+      "loss": 1.2374,
+      "step": 4795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.323113614516511,
+      "learning_rate": 9.014084507042254e-06,
+      "loss": 1.1926,
+      "step": 4800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.424173138049385,
+      "learning_rate": 9.023474178403756e-06,
+      "loss": 1.1746,
+      "step": 4805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.750633924642878,
+      "learning_rate": 9.032863849765259e-06,
+      "loss": 1.1695,
+      "step": 4810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.126106427680092,
+      "learning_rate": 9.042253521126761e-06,
+      "loss": 1.2219,
+      "step": 4815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.440550076998624,
+      "learning_rate": 9.051643192488264e-06,
+      "loss": 1.197,
+      "step": 4820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.805224265781256,
+      "learning_rate": 9.061032863849766e-06,
+      "loss": 1.151,
+      "step": 4825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.839277125488668,
+      "learning_rate": 9.070422535211268e-06,
+      "loss": 1.1953,
+      "step": 4830
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.487407125805589,
+      "learning_rate": 9.07981220657277e-06,
+      "loss": 1.1639,
+      "step": 4835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 15.147288333159524,
+      "learning_rate": 9.089201877934273e-06,
+      "loss": 1.1562,
+      "step": 4840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.170524064010127,
+      "learning_rate": 9.098591549295776e-06,
+      "loss": 1.2053,
+      "step": 4845
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 21.159133945462376,
+      "learning_rate": 9.107981220657278e-06,
+      "loss": 1.2215,
+      "step": 4850
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.317344425477746,
+      "learning_rate": 9.11737089201878e-06,
+      "loss": 1.2007,
+      "step": 4855
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.153987978554724,
+      "learning_rate": 9.126760563380283e-06,
+      "loss": 1.2317,
+      "step": 4860
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 33.34311167199715,
+      "learning_rate": 9.136150234741785e-06,
+      "loss": 1.188,
+      "step": 4865
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.06773661122188,
+      "learning_rate": 9.145539906103286e-06,
+      "loss": 1.138,
+      "step": 4870
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 53.31008122043855,
+      "learning_rate": 9.15492957746479e-06,
+      "loss": 1.223,
+      "step": 4875
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 97.84584817125534,
+      "learning_rate": 9.164319248826293e-06,
+      "loss": 1.2707,
+      "step": 4880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 63.0543719233712,
+      "learning_rate": 9.173708920187793e-06,
+      "loss": 1.2382,
+      "step": 4885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 25.27796698046594,
+      "learning_rate": 9.183098591549296e-06,
+      "loss": 1.2052,
+      "step": 4890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.754193419573788,
+      "learning_rate": 9.1924882629108e-06,
+      "loss": 1.2351,
+      "step": 4895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 75.82582834495221,
+      "learning_rate": 9.2018779342723e-06,
+      "loss": 1.2637,
+      "step": 4900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 31.605845668635858,
+      "learning_rate": 9.211267605633803e-06,
+      "loss": 1.2442,
+      "step": 4905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 24.00807115841491,
+      "learning_rate": 9.220657276995307e-06,
+      "loss": 1.264,
+      "step": 4910
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 46.746809972706465,
+      "learning_rate": 9.230046948356808e-06,
+      "loss": 1.2684,
+      "step": 4915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.930482339991464,
+      "learning_rate": 9.23943661971831e-06,
+      "loss": 1.2334,
+      "step": 4920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.376896059644336,
+      "learning_rate": 9.248826291079813e-06,
+      "loss": 1.1814,
+      "step": 4925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.253694936032222,
+      "learning_rate": 9.258215962441315e-06,
+      "loss": 1.2407,
+      "step": 4930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.872345619849714,
+      "learning_rate": 9.267605633802817e-06,
+      "loss": 1.1661,
+      "step": 4935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 17.877162737493244,
+      "learning_rate": 9.27699530516432e-06,
+      "loss": 1.2197,
+      "step": 4940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 25.009862820708754,
+      "learning_rate": 9.286384976525822e-06,
+      "loss": 1.1932,
+      "step": 4945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.191903839893259,
+      "learning_rate": 9.295774647887325e-06,
+      "loss": 1.207,
+      "step": 4950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.3620995658234,
+      "learning_rate": 9.305164319248827e-06,
+      "loss": 1.1951,
+      "step": 4955
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 42.848416867624465,
+      "learning_rate": 9.31455399061033e-06,
+      "loss": 1.1801,
+      "step": 4960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.973034679820064,
+      "learning_rate": 9.323943661971832e-06,
+      "loss": 1.2264,
+      "step": 4965
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.464394544881255,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 1.2056,
+      "step": 4970
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.694987514410068,
+      "learning_rate": 9.342723004694837e-06,
+      "loss": 1.2022,
+      "step": 4975
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.62508281391501,
+      "learning_rate": 9.35211267605634e-06,
+      "loss": 1.1698,
+      "step": 4980
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.597224579554815,
+      "learning_rate": 9.361502347417842e-06,
+      "loss": 1.1909,
+      "step": 4985
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 50.200478722221405,
+      "learning_rate": 9.370892018779344e-06,
+      "loss": 1.212,
+      "step": 4990
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 23.166968312343474,
+      "learning_rate": 9.380281690140845e-06,
+      "loss": 1.2054,
+      "step": 4995
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 27.50082609891482,
+      "learning_rate": 9.389671361502349e-06,
+      "loss": 1.2143,
+      "step": 5000
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.526418712566558,
+      "learning_rate": 9.399061032863851e-06,
+      "loss": 1.2128,
+      "step": 5005
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.575803570472214,
+      "learning_rate": 9.408450704225352e-06,
+      "loss": 1.2097,
+      "step": 5010
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 22.375062956236498,
+      "learning_rate": 9.417840375586856e-06,
+      "loss": 1.1992,
+      "step": 5015
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 43.71697187903528,
+      "learning_rate": 9.427230046948358e-06,
+      "loss": 1.2348,
+      "step": 5020
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 20.416802698197525,
+      "learning_rate": 9.43661971830986e-06,
+      "loss": 1.2097,
+      "step": 5025
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 17.384073176517635,
+      "learning_rate": 9.446009389671362e-06,
+      "loss": 1.2415,
+      "step": 5030
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.513791754954687,
+      "learning_rate": 9.455399061032866e-06,
+      "loss": 1.2087,
+      "step": 5035
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 37.451133021295504,
+      "learning_rate": 9.464788732394366e-06,
+      "loss": 1.2042,
+      "step": 5040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 25.269686324868847,
+      "learning_rate": 9.474178403755869e-06,
+      "loss": 1.2279,
+      "step": 5045
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.323648410174899,
+      "learning_rate": 9.483568075117371e-06,
+      "loss": 1.2422,
+      "step": 5050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.894795533491129,
+      "learning_rate": 9.492957746478874e-06,
+      "loss": 1.1976,
+      "step": 5055
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.497996133238441,
+      "learning_rate": 9.502347417840376e-06,
+      "loss": 1.1788,
+      "step": 5060
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 24.26837565399261,
+      "learning_rate": 9.511737089201879e-06,
+      "loss": 1.1912,
+      "step": 5065
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.685763569894704,
+      "learning_rate": 9.521126760563381e-06,
+      "loss": 1.2126,
+      "step": 5070
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.914181612884889,
+      "learning_rate": 9.530516431924883e-06,
+      "loss": 1.2282,
+      "step": 5075
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 34.564605352546806,
+      "learning_rate": 9.539906103286386e-06,
+      "loss": 1.1896,
+      "step": 5080
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.217776506429574,
+      "learning_rate": 9.549295774647888e-06,
+      "loss": 1.2424,
+      "step": 5085
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.717742410152061,
+      "learning_rate": 9.55868544600939e-06,
+      "loss": 1.2019,
+      "step": 5090
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.549766784551434,
+      "learning_rate": 9.568075117370893e-06,
+      "loss": 1.2022,
+      "step": 5095
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.441506824486499,
+      "learning_rate": 9.577464788732394e-06,
+      "loss": 1.2492,
+      "step": 5100
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 20.704859865177966,
+      "learning_rate": 9.586854460093898e-06,
+      "loss": 1.2659,
+      "step": 5105
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.938016303757815,
+      "learning_rate": 9.5962441314554e-06,
+      "loss": 1.2315,
+      "step": 5110
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.955300844308784,
+      "learning_rate": 9.605633802816903e-06,
+      "loss": 1.1809,
+      "step": 5115
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.783693029149658,
+      "learning_rate": 9.615023474178405e-06,
+      "loss": 1.2424,
+      "step": 5120
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.820839527288985,
+      "learning_rate": 9.624413145539908e-06,
+      "loss": 1.1924,
+      "step": 5125
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.242449432636181,
+      "learning_rate": 9.63380281690141e-06,
+      "loss": 1.1465,
+      "step": 5130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 25.358715380622098,
+      "learning_rate": 9.64319248826291e-06,
+      "loss": 1.2252,
+      "step": 5135
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.684351273646044,
+      "learning_rate": 9.652582159624415e-06,
+      "loss": 1.1993,
+      "step": 5140
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.680154433781327,
+      "learning_rate": 9.661971830985917e-06,
+      "loss": 1.2464,
+      "step": 5145
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.177913733583562,
+      "learning_rate": 9.671361502347418e-06,
+      "loss": 1.2024,
+      "step": 5150
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.0362502634753,
+      "learning_rate": 9.68075117370892e-06,
+      "loss": 1.1835,
+      "step": 5155
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 26.9558842363334,
+      "learning_rate": 9.690140845070424e-06,
+      "loss": 1.1635,
+      "step": 5160
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.942749649374964,
+      "learning_rate": 9.699530516431925e-06,
+      "loss": 1.2109,
+      "step": 5165
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.829725606056792,
+      "learning_rate": 9.708920187793428e-06,
+      "loss": 1.1928,
+      "step": 5170
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.25680118822684,
+      "learning_rate": 9.71830985915493e-06,
+      "loss": 1.2693,
+      "step": 5175
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 24.313307859101393,
+      "learning_rate": 9.727699530516432e-06,
+      "loss": 1.1852,
+      "step": 5180
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 23.367700122245164,
+      "learning_rate": 9.737089201877935e-06,
+      "loss": 1.2617,
+      "step": 5185
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.692227606630539,
+      "learning_rate": 9.746478873239437e-06,
+      "loss": 1.1946,
+      "step": 5190
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.188404391780125,
+      "learning_rate": 9.75586854460094e-06,
+      "loss": 1.1914,
+      "step": 5195
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.923435624679081,
+      "learning_rate": 9.765258215962442e-06,
+      "loss": 1.2436,
+      "step": 5200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.620573114568305,
+      "learning_rate": 9.774647887323945e-06,
+      "loss": 1.2035,
+      "step": 5205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.537849595936141,
+      "learning_rate": 9.784037558685447e-06,
+      "loss": 1.1962,
+      "step": 5210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 19.793103112018994,
+      "learning_rate": 9.79342723004695e-06,
+      "loss": 1.1887,
+      "step": 5215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.759387206168986,
+      "learning_rate": 9.802816901408452e-06,
+      "loss": 1.2161,
+      "step": 5220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 33.568472541916144,
+      "learning_rate": 9.812206572769954e-06,
+      "loss": 1.2223,
+      "step": 5225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 32.03329398432156,
+      "learning_rate": 9.821596244131457e-06,
+      "loss": 1.1922,
+      "step": 5230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.986653092221317,
+      "learning_rate": 9.830985915492959e-06,
+      "loss": 1.1862,
+      "step": 5235
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 34.68362616331035,
+      "learning_rate": 9.84037558685446e-06,
+      "loss": 1.1731,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.757833240331694,
+      "learning_rate": 9.849765258215964e-06,
+      "loss": 1.2275,
+      "step": 5245
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.732448948046146,
+      "learning_rate": 9.859154929577466e-06,
+      "loss": 1.2045,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 27.160606468261093,
+      "learning_rate": 9.868544600938969e-06,
+      "loss": 1.2143,
+      "step": 5255
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 32.5303111400796,
+      "learning_rate": 9.87793427230047e-06,
+      "loss": 1.2168,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.339242928379411,
+      "learning_rate": 9.887323943661974e-06,
+      "loss": 1.2535,
+      "step": 5265
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 38.25669067612021,
+      "learning_rate": 9.896713615023476e-06,
+      "loss": 1.1996,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 29.544270638998693,
+      "learning_rate": 9.906103286384977e-06,
+      "loss": 1.2133,
+      "step": 5275
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 32.02082441843353,
+      "learning_rate": 9.915492957746479e-06,
+      "loss": 1.2179,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 53.503063665296374,
+      "learning_rate": 9.924882629107983e-06,
+      "loss": 1.218,
+      "step": 5285
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.44623121392335,
+      "learning_rate": 9.934272300469484e-06,
+      "loss": 1.1835,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.714919664090878,
+      "learning_rate": 9.943661971830986e-06,
+      "loss": 1.2108,
+      "step": 5295
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.31137408950519,
+      "learning_rate": 9.953051643192489e-06,
+      "loss": 1.1877,
+      "step": 5300
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.843465143372992,
+      "learning_rate": 9.962441314553991e-06,
+      "loss": 1.1946,
+      "step": 5305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.691963663826979,
+      "learning_rate": 9.971830985915494e-06,
+      "loss": 1.1819,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.847767715476808,
+      "learning_rate": 9.981220657276996e-06,
+      "loss": 1.2201,
+      "step": 5315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 59.0052613075288,
+      "learning_rate": 9.990610328638498e-06,
+      "loss": 1.1814,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 42.90524030242737,
+      "learning_rate": 1e-05,
+      "loss": 1.2171,
+      "step": 5325
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.642478887489734,
+      "learning_rate": 9.999999731363967e-06,
+      "loss": 1.2269,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 35.97618179675625,
+      "learning_rate": 9.999998925455896e-06,
+      "loss": 1.2143,
+      "step": 5335
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.759564398963768,
+      "learning_rate": 9.999997582275873e-06,
+      "loss": 1.2483,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 19.71902916847821,
+      "learning_rate": 9.999995701824045e-06,
+      "loss": 1.1916,
+      "step": 5345
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.70424631708859,
+      "learning_rate": 9.99999328410061e-06,
+      "loss": 1.1702,
+      "step": 5350
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.6804488590414,
+      "learning_rate": 9.99999032910583e-06,
+      "loss": 1.1901,
+      "step": 5355
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.783159093383655,
+      "learning_rate": 9.999986836840025e-06,
+      "loss": 1.213,
+      "step": 5360
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.091343304753137,
+      "learning_rate": 9.999982807303565e-06,
+      "loss": 1.1593,
+      "step": 5365
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.762379753402147,
+      "learning_rate": 9.999978240496888e-06,
+      "loss": 1.1877,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.131886057956466,
+      "learning_rate": 9.99997313642048e-06,
+      "loss": 1.17,
+      "step": 5375
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.58416480541799,
+      "learning_rate": 9.999967495074893e-06,
+      "loss": 1.2139,
+      "step": 5380
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.13355735589607,
+      "learning_rate": 9.999961316460734e-06,
+      "loss": 1.2132,
+      "step": 5385
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.716110353909023,
+      "learning_rate": 9.999954600578664e-06,
+      "loss": 1.2026,
+      "step": 5390
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.14182793444988,
+      "learning_rate": 9.999947347429404e-06,
+      "loss": 1.1717,
+      "step": 5395
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.543748536284797,
+      "learning_rate": 9.999939557013735e-06,
+      "loss": 1.2499,
+      "step": 5400
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.874892555428925,
+      "learning_rate": 9.999931229332497e-06,
+      "loss": 1.2488,
+      "step": 5405
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.052901422161222,
+      "learning_rate": 9.99992236438658e-06,
+      "loss": 1.2287,
+      "step": 5410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.42883514807317,
+      "learning_rate": 9.99991296217694e-06,
+      "loss": 1.232,
+      "step": 5415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.873226141550873,
+      "learning_rate": 9.999903022704585e-06,
+      "loss": 1.194,
+      "step": 5420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 27.301087086132636,
+      "learning_rate": 9.999892545970584e-06,
+      "loss": 1.2601,
+      "step": 5425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 32.91165160043027,
+      "learning_rate": 9.999881531976061e-06,
+      "loss": 1.2243,
+      "step": 5430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.174456466038595,
+      "learning_rate": 9.999869980722203e-06,
+      "loss": 1.2251,
+      "step": 5435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.547114510797721,
+      "learning_rate": 9.99985789221025e-06,
+      "loss": 1.2368,
+      "step": 5440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.863205681315343,
+      "learning_rate": 9.9998452664415e-06,
+      "loss": 1.2096,
+      "step": 5445
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.244074030599695,
+      "learning_rate": 9.99983210341731e-06,
+      "loss": 1.2346,
+      "step": 5450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 44.47088945480686,
+      "learning_rate": 9.999818403139095e-06,
+      "loss": 1.2125,
+      "step": 5455
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 40.326348389319726,
+      "learning_rate": 9.999804165608324e-06,
+      "loss": 1.1811,
+      "step": 5460
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 48.70966946626771,
+      "learning_rate": 9.999789390826533e-06,
+      "loss": 1.2393,
+      "step": 5465
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.254712614925019,
+      "learning_rate": 9.999774078795305e-06,
+      "loss": 1.2123,
+      "step": 5470
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.37011727537402,
+      "learning_rate": 9.999758229516285e-06,
+      "loss": 1.2538,
+      "step": 5475
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 19.1875008874081,
+      "learning_rate": 9.99974184299118e-06,
+      "loss": 1.1837,
+      "step": 5480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 30.704990829622357,
+      "learning_rate": 9.999724919221748e-06,
+      "loss": 1.2427,
+      "step": 5485
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.034428590584625,
+      "learning_rate": 9.999707458209806e-06,
+      "loss": 1.1904,
+      "step": 5490
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.073346430131144,
+      "learning_rate": 9.999689459957235e-06,
+      "loss": 1.2091,
+      "step": 5495
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 25.739096234399756,
+      "learning_rate": 9.999670924465965e-06,
+      "loss": 1.2148,
+      "step": 5500
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 25.476229200823596,
+      "learning_rate": 9.999651851737988e-06,
+      "loss": 1.2147,
+      "step": 5505
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 20.755577749329202,
+      "learning_rate": 9.999632241775356e-06,
+      "loss": 1.2629,
+      "step": 5510
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.853478498323742,
+      "learning_rate": 9.999612094580173e-06,
+      "loss": 1.2378,
+      "step": 5515
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.94071427053803,
+      "learning_rate": 9.999591410154607e-06,
+      "loss": 1.1723,
+      "step": 5520
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.987992075584888,
+      "learning_rate": 9.999570188500878e-06,
+      "loss": 1.2089,
+      "step": 5525
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.367646734886607,
+      "learning_rate": 9.999548429621268e-06,
+      "loss": 1.173,
+      "step": 5530
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 28.945020269537835,
+      "learning_rate": 9.999526133518114e-06,
+      "loss": 1.2657,
+      "step": 5535
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 20.0141231729636,
+      "learning_rate": 9.999503300193812e-06,
+      "loss": 1.1477,
+      "step": 5540
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 35.39670647732335,
+      "learning_rate": 9.999479929650817e-06,
+      "loss": 1.1829,
+      "step": 5545
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.426164209134736,
+      "learning_rate": 9.999456021891639e-06,
+      "loss": 1.1728,
+      "step": 5550
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.157225159006694,
+      "learning_rate": 9.999431576918846e-06,
+      "loss": 1.2227,
+      "step": 5555
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.200093175825737,
+      "learning_rate": 9.999406594735069e-06,
+      "loss": 1.195,
+      "step": 5560
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.359635002109798,
+      "learning_rate": 9.999381075342985e-06,
+      "loss": 1.1369,
+      "step": 5565
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.130749881958652,
+      "learning_rate": 9.999355018745342e-06,
+      "loss": 1.2477,
+      "step": 5570
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 37.31754887801666,
+      "learning_rate": 9.999328424944941e-06,
+      "loss": 1.2176,
+      "step": 5575
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.99169364129181,
+      "learning_rate": 9.999301293944634e-06,
+      "loss": 1.1923,
+      "step": 5580
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 39.25216367288758,
+      "learning_rate": 9.999273625747342e-06,
+      "loss": 1.2632,
+      "step": 5585
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 20.378364338026277,
+      "learning_rate": 9.999245420356033e-06,
+      "loss": 1.2102,
+      "step": 5590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 26.550975094038357,
+      "learning_rate": 9.999216677773741e-06,
+      "loss": 1.2125,
+      "step": 5595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.447423120345595,
+      "learning_rate": 9.999187398003553e-06,
+      "loss": 1.1979,
+      "step": 5600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.126437570958707,
+      "learning_rate": 9.999157581048618e-06,
+      "loss": 1.1757,
+      "step": 5605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.668495802312965,
+      "learning_rate": 9.999127226912137e-06,
+      "loss": 1.2428,
+      "step": 5610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.988932733946095,
+      "learning_rate": 9.999096335597372e-06,
+      "loss": 1.2566,
+      "step": 5615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.905705436481886,
+      "learning_rate": 9.999064907107643e-06,
+      "loss": 1.2127,
+      "step": 5620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.778779821183123,
+      "learning_rate": 9.999032941446327e-06,
+      "loss": 1.228,
+      "step": 5625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 32.87174609874904,
+      "learning_rate": 9.99900043861686e-06,
+      "loss": 1.2092,
+      "step": 5630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 115.50877638594692,
+      "learning_rate": 9.998967398622734e-06,
+      "loss": 1.4483,
+      "step": 5635
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 881.4583633594746,
+      "learning_rate": 9.998933821467498e-06,
+      "loss": 5.8272,
+      "step": 5640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3625.864863542416,
+      "learning_rate": 9.99889970715476e-06,
+      "loss": 31.1807,
+      "step": 5645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 305.2992238370064,
+      "learning_rate": 9.998865055688189e-06,
+      "loss": 15.8795,
+      "step": 5650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 220.90265182525806,
+      "learning_rate": 9.998829867071505e-06,
+      "loss": 11.2603,
+      "step": 5655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 202.50617493743596,
+      "learning_rate": 9.998794141308489e-06,
+      "loss": 9.4577,
+      "step": 5660
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 127.08734520183457,
+      "learning_rate": 9.998757878402982e-06,
+      "loss": 8.3117,
+      "step": 5665
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 46.60949031393593,
+      "learning_rate": 9.998721078358881e-06,
+      "loss": 7.9507,
+      "step": 5670
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 96.18974630241713,
+      "learning_rate": 9.998683741180138e-06,
+      "loss": 7.5956,
+      "step": 5675
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 96.12491744486918,
+      "learning_rate": 9.998645866870766e-06,
+      "loss": 7.315,
+      "step": 5680
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 167.2630103539481,
+      "learning_rate": 9.998607455434835e-06,
+      "loss": 7.3938,
+      "step": 5685
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 222.38158996981866,
+      "learning_rate": 9.998568506876473e-06,
+      "loss": 7.3077,
+      "step": 5690
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 151.93538832351118,
+      "learning_rate": 9.998529021199863e-06,
+      "loss": 6.9725,
+      "step": 5695
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 103.61752098498127,
+      "learning_rate": 9.998488998409251e-06,
+      "loss": 6.874,
+      "step": 5700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 296.2096876645142,
+      "learning_rate": 9.998448438508937e-06,
+      "loss": 6.7711,
+      "step": 5705
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 200.06188781368692,
+      "learning_rate": 9.998407341503276e-06,
+      "loss": 6.8071,
+      "step": 5710
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 56.562649527634086,
+      "learning_rate": 9.998365707396688e-06,
+      "loss": 6.5955,
+      "step": 5715
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 131.0841260135904,
+      "learning_rate": 9.998323536193643e-06,
+      "loss": 6.1754,
+      "step": 5720
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 91.185524542102,
+      "learning_rate": 9.998280827898679e-06,
+      "loss": 6.5431,
+      "step": 5725
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 177.60728612407365,
+      "learning_rate": 9.998237582516377e-06,
+      "loss": 6.5339,
+      "step": 5730
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 92.57721147555587,
+      "learning_rate": 9.99819380005139e-06,
+      "loss": 6.3031,
+      "step": 5735
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 85.28429333005862,
+      "learning_rate": 9.99814948050842e-06,
+      "loss": 5.9641,
+      "step": 5740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 46.15466298428571,
+      "learning_rate": 9.99810462389223e-06,
+      "loss": 5.73,
+      "step": 5745
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 108.04987142422614,
+      "learning_rate": 9.998059230207639e-06,
+      "loss": 5.5082,
+      "step": 5750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 71.01705772068496,
+      "learning_rate": 9.998013299459525e-06,
+      "loss": 5.1304,
+      "step": 5755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 166.47184090502537,
+      "learning_rate": 9.997966831652826e-06,
+      "loss": 4.7207,
+      "step": 5760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 108.34809596443222,
+      "learning_rate": 9.997919826792532e-06,
+      "loss": 4.4064,
+      "step": 5765
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 203.25488572574156,
+      "learning_rate": 9.997872284883697e-06,
+      "loss": 4.1995,
+      "step": 5770
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 76.96913596718933,
+      "learning_rate": 9.997824205931426e-06,
+      "loss": 3.918,
+      "step": 5775
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 46.789318270837775,
+      "learning_rate": 9.997775589940886e-06,
+      "loss": 3.2998,
+      "step": 5780
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 150.9068237438903,
+      "learning_rate": 9.997726436917303e-06,
+      "loss": 2.9199,
+      "step": 5785
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 101.81714399310493,
+      "learning_rate": 9.997676746865958e-06,
+      "loss": 2.8614,
+      "step": 5790
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 39.513324392664025,
+      "learning_rate": 9.997626519792192e-06,
+      "loss": 2.3969,
+      "step": 5795
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 110.3570993391848,
+      "learning_rate": 9.997575755701397e-06,
+      "loss": 3.1355,
+      "step": 5800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 39.52089431069558,
+      "learning_rate": 9.997524454599033e-06,
+      "loss": 3.2513,
+      "step": 5805
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 33.58047088835631,
+      "learning_rate": 9.997472616490608e-06,
+      "loss": 2.8184,
+      "step": 5810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 65.7301556926572,
+      "learning_rate": 9.9974202413817e-06,
+      "loss": 2.6076,
+      "step": 5815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 49.74937374199039,
+      "learning_rate": 9.997367329277927e-06,
+      "loss": 2.2297,
+      "step": 5820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 56.065182542415954,
+      "learning_rate": 9.99731388018498e-06,
+      "loss": 2.1355,
+      "step": 5825
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 109.14082554172863,
+      "learning_rate": 9.997259894108604e-06,
+      "loss": 2.0288,
+      "step": 5830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 33.05771008468508,
+      "learning_rate": 9.997205371054595e-06,
+      "loss": 2.0012,
+      "step": 5835
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 57.27210785048184,
+      "learning_rate": 9.997150311028817e-06,
+      "loss": 1.9041,
+      "step": 5840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.765176528339722,
+      "learning_rate": 9.997094714037184e-06,
+      "loss": 1.7866,
+      "step": 5845
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.644104953020037,
+      "learning_rate": 9.997038580085667e-06,
+      "loss": 1.7795,
+      "step": 5850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.885425575621817,
+      "learning_rate": 9.996981909180302e-06,
+      "loss": 1.6351,
+      "step": 5855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.47464971728927,
+      "learning_rate": 9.996924701327181e-06,
+      "loss": 1.7488,
+      "step": 5860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 35.885926326951555,
+      "learning_rate": 9.996866956532445e-06,
+      "loss": 1.6965,
+      "step": 5865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 30.187527884353738,
+      "learning_rate": 9.996808674802302e-06,
+      "loss": 1.5814,
+      "step": 5870
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.593526588901867,
+      "learning_rate": 9.996749856143015e-06,
+      "loss": 1.5712,
+      "step": 5875
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.543105853979235,
+      "learning_rate": 9.996690500560901e-06,
+      "loss": 1.6264,
+      "step": 5880
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 19.130086679293516,
+      "learning_rate": 9.996630608062343e-06,
+      "loss": 1.5903,
+      "step": 5885
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 24.336913613381725,
+      "learning_rate": 9.996570178653775e-06,
+      "loss": 1.5608,
+      "step": 5890
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.863603621841438,
+      "learning_rate": 9.996509212341688e-06,
+      "loss": 1.5058,
+      "step": 5895
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.92143656057317,
+      "learning_rate": 9.996447709132636e-06,
+      "loss": 1.5812,
+      "step": 5900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 27.622210473749963,
+      "learning_rate": 9.996385669033226e-06,
+      "loss": 1.4991,
+      "step": 5905
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.14597569004163,
+      "learning_rate": 9.996323092050124e-06,
+      "loss": 1.3941,
+      "step": 5910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.666020322647581,
+      "learning_rate": 9.996259978190057e-06,
+      "loss": 1.4521,
+      "step": 5915
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.34972346254299,
+      "learning_rate": 9.996196327459805e-06,
+      "loss": 1.4744,
+      "step": 5920
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.553647830129735,
+      "learning_rate": 9.996132139866209e-06,
+      "loss": 1.4585,
+      "step": 5925
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.795663010068404,
+      "learning_rate": 9.996067415416162e-06,
+      "loss": 1.4352,
+      "step": 5930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.003136410264096,
+      "learning_rate": 9.996002154116623e-06,
+      "loss": 1.453,
+      "step": 5935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 22.519047791976174,
+      "learning_rate": 9.995936355974604e-06,
+      "loss": 1.4043,
+      "step": 5940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 22.01002822942101,
+      "learning_rate": 9.995870020997174e-06,
+      "loss": 1.4325,
+      "step": 5945
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.875887721533456,
+      "learning_rate": 9.995803149191464e-06,
+      "loss": 1.3588,
+      "step": 5950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.485119099960336,
+      "learning_rate": 9.995735740564654e-06,
+      "loss": 1.3962,
+      "step": 5955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.37483936114496,
+      "learning_rate": 9.995667795123993e-06,
+      "loss": 1.4035,
+      "step": 5960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.990121869896935,
+      "learning_rate": 9.99559931287678e-06,
+      "loss": 1.3746,
+      "step": 5965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.323034006028854,
+      "learning_rate": 9.995530293830373e-06,
+      "loss": 1.4105,
+      "step": 5970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 25.63935213087236,
+      "learning_rate": 9.995460737992189e-06,
+      "loss": 1.3679,
+      "step": 5975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.770733129860528,
+      "learning_rate": 9.995390645369702e-06,
+      "loss": 1.3796,
+      "step": 5980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.724713081643543,
+      "learning_rate": 9.995320015970444e-06,
+      "loss": 1.3579,
+      "step": 5985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.353192099760806,
+      "learning_rate": 9.995248849802005e-06,
+      "loss": 1.3298,
+      "step": 5990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 41.078784542354725,
+      "learning_rate": 9.99517714687203e-06,
+      "loss": 1.3807,
+      "step": 5995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.714724279848866,
+      "learning_rate": 9.995104907188224e-06,
+      "loss": 1.3605,
+      "step": 6000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.188363629473503,
+      "learning_rate": 9.995032130758352e-06,
+      "loss": 1.3698,
+      "step": 6005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.443782220759342,
+      "learning_rate": 9.994958817590235e-06,
+      "loss": 1.356,
+      "step": 6010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 48.070755393626605,
+      "learning_rate": 9.994884967691746e-06,
+      "loss": 1.3474,
+      "step": 6015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.701750321001162,
+      "learning_rate": 9.994810581070824e-06,
+      "loss": 1.2907,
+      "step": 6020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 21.92830765227391,
+      "learning_rate": 9.994735657735461e-06,
+      "loss": 1.3804,
+      "step": 6025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.06167115759183,
+      "learning_rate": 9.994660197693709e-06,
+      "loss": 1.306,
+      "step": 6030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.024618993003601,
+      "learning_rate": 9.994584200953675e-06,
+      "loss": 1.3732,
+      "step": 6035
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 47.32997844861397,
+      "learning_rate": 9.994507667523525e-06,
+      "loss": 1.3245,
+      "step": 6040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.72115080743169,
+      "learning_rate": 9.994430597411485e-06,
+      "loss": 1.3369,
+      "step": 6045
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 31.174924094202126,
+      "learning_rate": 9.994352990625834e-06,
+      "loss": 1.3063,
+      "step": 6050
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.511141474095467,
+      "learning_rate": 9.994274847174914e-06,
+      "loss": 1.316,
+      "step": 6055
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 39.562695270008206,
+      "learning_rate": 9.994196167067119e-06,
+      "loss": 1.3749,
+      "step": 6060
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 30.851507876936136,
+      "learning_rate": 9.994116950310906e-06,
+      "loss": 1.3287,
+      "step": 6065
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.80246557600141,
+      "learning_rate": 9.994037196914783e-06,
+      "loss": 1.303,
+      "step": 6070
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.735455033227984,
+      "learning_rate": 9.993956906887325e-06,
+      "loss": 1.3022,
+      "step": 6075
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.090530761996225,
+      "learning_rate": 9.993876080237156e-06,
+      "loss": 1.3235,
+      "step": 6080
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.250823813128243,
+      "learning_rate": 9.993794716972965e-06,
+      "loss": 1.2754,
+      "step": 6085
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.751606006352525,
+      "learning_rate": 9.993712817103492e-06,
+      "loss": 1.2513,
+      "step": 6090
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.24144172013427,
+      "learning_rate": 9.993630380637535e-06,
+      "loss": 1.2602,
+      "step": 6095
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.462754315735199,
+      "learning_rate": 9.993547407583959e-06,
+      "loss": 1.2914,
+      "step": 6100
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.361145071242321,
+      "learning_rate": 9.993463897951674e-06,
+      "loss": 1.2964,
+      "step": 6105
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.5607624591136,
+      "learning_rate": 9.993379851749654e-06,
+      "loss": 1.294,
+      "step": 6110
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.514105902885473,
+      "learning_rate": 9.993295268986933e-06,
+      "loss": 1.3002,
+      "step": 6115
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.428444692759303,
+      "learning_rate": 9.993210149672599e-06,
+      "loss": 1.2937,
+      "step": 6120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.3766845403608885,
+      "learning_rate": 9.993124493815795e-06,
+      "loss": 1.2976,
+      "step": 6125
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.438019509579105,
+      "learning_rate": 9.99303830142573e-06,
+      "loss": 1.2802,
+      "step": 6130
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.211470289024709,
+      "learning_rate": 9.992951572511663e-06,
+      "loss": 1.2915,
+      "step": 6135
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.094169692326695,
+      "learning_rate": 9.992864307082913e-06,
+      "loss": 1.3278,
+      "step": 6140
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.0882662818744,
+      "learning_rate": 9.99277650514886e-06,
+      "loss": 1.3465,
+      "step": 6145
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.03194416213248,
+      "learning_rate": 9.992688166718936e-06,
+      "loss": 1.2896,
+      "step": 6150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.655716652474666,
+      "learning_rate": 9.992599291802632e-06,
+      "loss": 1.2799,
+      "step": 6155
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.493938727180848,
+      "learning_rate": 9.992509880409502e-06,
+      "loss": 1.248,
+      "step": 6160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.579512459797152,
+      "learning_rate": 9.992419932549153e-06,
+      "loss": 1.3345,
+      "step": 6165
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.364772561346149,
+      "learning_rate": 9.992329448231247e-06,
+      "loss": 1.2394,
+      "step": 6170
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.170354962495536,
+      "learning_rate": 9.992238427465509e-06,
+      "loss": 1.2742,
+      "step": 6175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.749806841645428,
+      "learning_rate": 9.99214687026172e-06,
+      "loss": 1.2616,
+      "step": 6180
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.81401302120067,
+      "learning_rate": 9.992054776629718e-06,
+      "loss": 1.2788,
+      "step": 6185
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.618846351763926,
+      "learning_rate": 9.991962146579398e-06,
+      "loss": 1.2593,
+      "step": 6190
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.48171112079907,
+      "learning_rate": 9.991868980120715e-06,
+      "loss": 1.2739,
+      "step": 6195
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.382895403840644,
+      "learning_rate": 9.991775277263679e-06,
+      "loss": 1.2356,
+      "step": 6200
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.339821057005265,
+      "learning_rate": 9.991681038018358e-06,
+      "loss": 1.2174,
+      "step": 6205
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.43383633184435,
+      "learning_rate": 9.99158626239488e-06,
+      "loss": 1.252,
+      "step": 6210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.853426960471305,
+      "learning_rate": 9.99149095040343e-06,
+      "loss": 1.2671,
+      "step": 6215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.38063312361502,
+      "learning_rate": 9.991395102054247e-06,
+      "loss": 1.3231,
+      "step": 6220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.293696295463246,
+      "learning_rate": 9.99129871735763e-06,
+      "loss": 1.2648,
+      "step": 6225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.106918476852423,
+      "learning_rate": 9.991201796323942e-06,
+      "loss": 1.2838,
+      "step": 6230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.306838698078574,
+      "learning_rate": 9.991104338963589e-06,
+      "loss": 1.213,
+      "step": 6235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.798825223650576,
+      "learning_rate": 9.99100634528705e-06,
+      "loss": 1.2426,
+      "step": 6240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.919194363165933,
+      "learning_rate": 9.990907815304852e-06,
+      "loss": 1.2312,
+      "step": 6245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.478546964961088,
+      "learning_rate": 9.99080874902758e-06,
+      "loss": 1.2159,
+      "step": 6250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.07205165223581,
+      "learning_rate": 9.990709146465883e-06,
+      "loss": 1.2935,
+      "step": 6255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.683757685547498,
+      "learning_rate": 9.990609007630463e-06,
+      "loss": 1.2632,
+      "step": 6260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.307852796498807,
+      "learning_rate": 9.990508332532081e-06,
+      "loss": 1.2277,
+      "step": 6265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.457185329870928,
+      "learning_rate": 9.990407121181552e-06,
+      "loss": 1.2405,
+      "step": 6270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.918195701636762,
+      "learning_rate": 9.990305373589755e-06,
+      "loss": 1.2337,
+      "step": 6275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.652700093334805,
+      "learning_rate": 9.990203089767621e-06,
+      "loss": 1.2668,
+      "step": 6280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.710269812484548,
+      "learning_rate": 9.990100269726142e-06,
+      "loss": 1.2457,
+      "step": 6285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.886033872146542,
+      "learning_rate": 9.989996913476365e-06,
+      "loss": 1.2653,
+      "step": 6290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.875680118840527,
+      "learning_rate": 9.989893021029397e-06,
+      "loss": 1.2257,
+      "step": 6295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.031372531354194,
+      "learning_rate": 9.989788592396403e-06,
+      "loss": 1.2417,
+      "step": 6300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.029512932873606,
+      "learning_rate": 9.989683627588602e-06,
+      "loss": 1.2226,
+      "step": 6305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.366888709665293,
+      "learning_rate": 9.989578126617275e-06,
+      "loss": 1.2136,
+      "step": 6310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.228820263441918,
+      "learning_rate": 9.989472089493757e-06,
+      "loss": 1.2434,
+      "step": 6315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.872041024498126,
+      "learning_rate": 9.989365516229443e-06,
+      "loss": 1.1995,
+      "step": 6320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.195402871756778,
+      "learning_rate": 9.989258406835784e-06,
+      "loss": 1.26,
+      "step": 6325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.746412979279807,
+      "learning_rate": 9.989150761324291e-06,
+      "loss": 1.2123,
+      "step": 6330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.677966169342557,
+      "learning_rate": 9.989042579706529e-06,
+      "loss": 1.2162,
+      "step": 6335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.547617666929987,
+      "learning_rate": 9.988933861994126e-06,
+      "loss": 1.1848,
+      "step": 6340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.536929004723213,
+      "learning_rate": 9.988824608198759e-06,
+      "loss": 1.2127,
+      "step": 6345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.949556034222528,
+      "learning_rate": 9.98871481833217e-06,
+      "loss": 1.2518,
+      "step": 6350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.790578838003556,
+      "learning_rate": 9.988604492406158e-06,
+      "loss": 1.1982,
+      "step": 6355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.176996419741183,
+      "learning_rate": 9.988493630432577e-06,
+      "loss": 1.2282,
+      "step": 6360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.317991360196196,
+      "learning_rate": 9.988382232423338e-06,
+      "loss": 1.24,
+      "step": 6365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.271672244857991,
+      "learning_rate": 9.988270298390415e-06,
+      "loss": 1.2182,
+      "step": 6370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.743481707909398,
+      "learning_rate": 9.988157828345833e-06,
+      "loss": 1.2571,
+      "step": 6375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.888077544382435,
+      "learning_rate": 9.988044822301678e-06,
+      "loss": 1.2328,
+      "step": 6380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.160903762627039,
+      "learning_rate": 9.987931280270093e-06,
+      "loss": 1.1933,
+      "step": 6385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.00407774737748,
+      "learning_rate": 9.987817202263276e-06,
+      "loss": 1.2325,
+      "step": 6390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.029209848814627,
+      "learning_rate": 9.98770258829349e-06,
+      "loss": 1.1658,
+      "step": 6395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 26.30664677401515,
+      "learning_rate": 9.98758743837305e-06,
+      "loss": 1.1992,
+      "step": 6400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.125595188626505,
+      "learning_rate": 9.987471752514326e-06,
+      "loss": 1.2889,
+      "step": 6405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.211557792670988,
+      "learning_rate": 9.987355530729749e-06,
+      "loss": 1.2021,
+      "step": 6410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.557012792283704,
+      "learning_rate": 9.987238773031813e-06,
+      "loss": 1.1957,
+      "step": 6415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.691877776216604,
+      "learning_rate": 9.987121479433058e-06,
+      "loss": 1.2452,
+      "step": 6420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.072706922358888,
+      "learning_rate": 9.98700364994609e-06,
+      "loss": 1.2601,
+      "step": 6425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 26.229296615583486,
+      "learning_rate": 9.986885284583571e-06,
+      "loss": 1.2652,
+      "step": 6430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.646765147718657,
+      "learning_rate": 9.98676638335822e-06,
+      "loss": 1.3181,
+      "step": 6435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.255804620174242,
+      "learning_rate": 9.986646946282813e-06,
+      "loss": 1.414,
+      "step": 6440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.07395174063236,
+      "learning_rate": 9.986526973370183e-06,
+      "loss": 1.3736,
+      "step": 6445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.26355634975153,
+      "learning_rate": 9.986406464633223e-06,
+      "loss": 1.2844,
+      "step": 6450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 22.164598787057212,
+      "learning_rate": 9.986285420084882e-06,
+      "loss": 1.2579,
+      "step": 6455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.467313692602483,
+      "learning_rate": 9.986163839738166e-06,
+      "loss": 1.193,
+      "step": 6460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.508513702137008,
+      "learning_rate": 9.98604172360614e-06,
+      "loss": 1.2054,
+      "step": 6465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 23.286058715981014,
+      "learning_rate": 9.985919071701924e-06,
+      "loss": 1.2745,
+      "step": 6470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.995764903647617,
+      "learning_rate": 9.985795884038702e-06,
+      "loss": 1.2495,
+      "step": 6475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 21.336127048366873,
+      "learning_rate": 9.985672160629707e-06,
+      "loss": 1.2266,
+      "step": 6480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.059099352692346,
+      "learning_rate": 9.985547901488234e-06,
+      "loss": 1.2429,
+      "step": 6485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.811121913527234,
+      "learning_rate": 9.985423106627637e-06,
+      "loss": 1.2583,
+      "step": 6490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.09505169249471,
+      "learning_rate": 9.985297776061325e-06,
+      "loss": 1.1985,
+      "step": 6495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.115540489699223,
+      "learning_rate": 9.985171909802764e-06,
+      "loss": 1.1897,
+      "step": 6500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.421280212127511,
+      "learning_rate": 9.985045507865483e-06,
+      "loss": 1.1876,
+      "step": 6505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.42788968584019,
+      "learning_rate": 9.984918570263057e-06,
+      "loss": 1.2182,
+      "step": 6510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.183407227421018,
+      "learning_rate": 9.984791097009135e-06,
+      "loss": 1.1955,
+      "step": 6515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.311443818316873,
+      "learning_rate": 9.984663088117408e-06,
+      "loss": 1.2622,
+      "step": 6520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.801014423396312,
+      "learning_rate": 9.984534543601633e-06,
+      "loss": 1.1979,
+      "step": 6525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.702329591648484,
+      "learning_rate": 9.984405463475623e-06,
+      "loss": 1.2319,
+      "step": 6530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.6443117734074635,
+      "learning_rate": 9.984275847753249e-06,
+      "loss": 1.1722,
+      "step": 6535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.961602409349189,
+      "learning_rate": 9.984145696448436e-06,
+      "loss": 1.1836,
+      "step": 6540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.1218714975184945,
+      "learning_rate": 9.984015009575173e-06,
+      "loss": 1.2184,
+      "step": 6545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.770273509403772,
+      "learning_rate": 9.9838837871475e-06,
+      "loss": 1.2763,
+      "step": 6550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.412295504959258,
+      "learning_rate": 9.983752029179519e-06,
+      "loss": 1.2296,
+      "step": 6555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.005358612462016,
+      "learning_rate": 9.983619735685388e-06,
+      "loss": 1.2663,
+      "step": 6560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.303838874266845,
+      "learning_rate": 9.98348690667932e-06,
+      "loss": 1.3097,
+      "step": 6565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.482381417791832,
+      "learning_rate": 9.983353542175593e-06,
+      "loss": 1.1933,
+      "step": 6570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.8825771764517185,
+      "learning_rate": 9.983219642188534e-06,
+      "loss": 1.2223,
+      "step": 6575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.0285012987186,
+      "learning_rate": 9.98308520673253e-06,
+      "loss": 1.1972,
+      "step": 6580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.451173223452615,
+      "learning_rate": 9.982950235822029e-06,
+      "loss": 1.2012,
+      "step": 6585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.26244826372432,
+      "learning_rate": 9.982814729471534e-06,
+      "loss": 1.227,
+      "step": 6590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.762539474418195,
+      "learning_rate": 9.982678687695606e-06,
+      "loss": 1.1729,
+      "step": 6595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.948620552003382,
+      "learning_rate": 9.982542110508861e-06,
+      "loss": 1.2146,
+      "step": 6600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.969768118573688,
+      "learning_rate": 9.982404997925979e-06,
+      "loss": 1.1989,
+      "step": 6605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.409753394542134,
+      "learning_rate": 9.98226734996169e-06,
+      "loss": 1.186,
+      "step": 6610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.606136956491269,
+      "learning_rate": 9.982129166630784e-06,
+      "loss": 1.1183,
+      "step": 6615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.052044206488178,
+      "learning_rate": 9.981990447948115e-06,
+      "loss": 1.2006,
+      "step": 6620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.276486519041882,
+      "learning_rate": 9.981851193928582e-06,
+      "loss": 1.1967,
+      "step": 6625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.725980893011915,
+      "learning_rate": 9.981711404587154e-06,
+      "loss": 1.1951,
+      "step": 6630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.543421111360324,
+      "learning_rate": 9.981571079938848e-06,
+      "loss": 1.1752,
+      "step": 6635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.42120363304379,
+      "learning_rate": 9.981430219998745e-06,
+      "loss": 1.189,
+      "step": 6640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.31891174061556,
+      "learning_rate": 9.981288824781979e-06,
+      "loss": 1.179,
+      "step": 6645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.73574880272908,
+      "learning_rate": 9.981146894303746e-06,
+      "loss": 1.221,
+      "step": 6650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.625795137531074,
+      "learning_rate": 9.981004428579295e-06,
+      "loss": 1.2337,
+      "step": 6655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.926641347856304,
+      "learning_rate": 9.980861427623936e-06,
+      "loss": 1.1927,
+      "step": 6660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.027132323335643,
+      "learning_rate": 9.980717891453033e-06,
+      "loss": 1.229,
+      "step": 6665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.515724171980387,
+      "learning_rate": 9.98057382008201e-06,
+      "loss": 1.1977,
+      "step": 6670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.04002557342166,
+      "learning_rate": 9.980429213526353e-06,
+      "loss": 1.2101,
+      "step": 6675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.2277641416974,
+      "learning_rate": 9.980284071801593e-06,
+      "loss": 1.2256,
+      "step": 6680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.385543940602545,
+      "learning_rate": 9.980138394923332e-06,
+      "loss": 1.207,
+      "step": 6685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.407538260156455,
+      "learning_rate": 9.979992182907221e-06,
+      "loss": 1.2228,
+      "step": 6690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.668419958029871,
+      "learning_rate": 9.979845435768972e-06,
+      "loss": 1.1601,
+      "step": 6695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.056901129620707,
+      "learning_rate": 9.979698153524352e-06,
+      "loss": 1.1835,
+      "step": 6700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.809242809918025,
+      "learning_rate": 9.97955033618919e-06,
+      "loss": 1.1895,
+      "step": 6705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.7023950580463545,
+      "learning_rate": 9.979401983779365e-06,
+      "loss": 1.2104,
+      "step": 6710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.585746717283113,
+      "learning_rate": 9.979253096310823e-06,
+      "loss": 1.2242,
+      "step": 6715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.002910629314679,
+      "learning_rate": 9.97910367379956e-06,
+      "loss": 1.1486,
+      "step": 6720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.570248464117552,
+      "learning_rate": 9.978953716261635e-06,
+      "loss": 1.2567,
+      "step": 6725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.381962150302656,
+      "learning_rate": 9.978803223713156e-06,
+      "loss": 1.2097,
+      "step": 6730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.2509666853499715,
+      "learning_rate": 9.9786521961703e-06,
+      "loss": 1.1832,
+      "step": 6735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.526298456096228,
+      "learning_rate": 9.978500633649293e-06,
+      "loss": 1.1829,
+      "step": 6740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.6947050563226576,
+      "learning_rate": 9.97834853616642e-06,
+      "loss": 1.1662,
+      "step": 6745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.66838934819671,
+      "learning_rate": 9.978195903738025e-06,
+      "loss": 1.1819,
+      "step": 6750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.851742927177732,
+      "learning_rate": 9.978042736380512e-06,
+      "loss": 1.2557,
+      "step": 6755
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.781792167301317,
+      "learning_rate": 9.977889034110335e-06,
+      "loss": 1.1885,
+      "step": 6760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.408764618901928,
+      "learning_rate": 9.977734796944014e-06,
+      "loss": 1.1605,
+      "step": 6765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.7193806777456,
+      "learning_rate": 9.97758002489812e-06,
+      "loss": 1.1567,
+      "step": 6770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.112051037999008,
+      "learning_rate": 9.977424717989283e-06,
+      "loss": 1.2238,
+      "step": 6775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.4197189642230095,
+      "learning_rate": 9.977268876234194e-06,
+      "loss": 1.1709,
+      "step": 6780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.464930167865641,
+      "learning_rate": 9.9771124996496e-06,
+      "loss": 1.1785,
+      "step": 6785
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.589860503526791,
+      "learning_rate": 9.9769555882523e-06,
+      "loss": 1.2352,
+      "step": 6790
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.092556393247447,
+      "learning_rate": 9.976798142059157e-06,
+      "loss": 1.2366,
+      "step": 6795
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.813574251657148,
+      "learning_rate": 9.97664016108709e-06,
+      "loss": 1.197,
+      "step": 6800
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.022672747157394,
+      "learning_rate": 9.976481645353072e-06,
+      "loss": 1.2248,
+      "step": 6805
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.438375149101976,
+      "learning_rate": 9.976322594874142e-06,
+      "loss": 1.2111,
+      "step": 6810
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.544602325550182,
+      "learning_rate": 9.976163009667385e-06,
+      "loss": 1.1698,
+      "step": 6815
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.727415785040562,
+      "learning_rate": 9.976002889749952e-06,
+      "loss": 1.2037,
+      "step": 6820
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.383405370863493,
+      "learning_rate": 9.975842235139047e-06,
+      "loss": 1.167,
+      "step": 6825
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.669874700594074,
+      "learning_rate": 9.975681045851935e-06,
+      "loss": 1.2254,
+      "step": 6830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.957958807511885,
+      "learning_rate": 9.975519321905934e-06,
+      "loss": 1.1434,
+      "step": 6835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.459036766792778,
+      "learning_rate": 9.975357063318423e-06,
+      "loss": 1.2177,
+      "step": 6840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.745614678291854,
+      "learning_rate": 9.975194270106838e-06,
+      "loss": 1.192,
+      "step": 6845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.764960663693426,
+      "learning_rate": 9.97503094228867e-06,
+      "loss": 1.1663,
+      "step": 6850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.76346246690597,
+      "learning_rate": 9.974867079881473e-06,
+      "loss": 1.2545,
+      "step": 6855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.340690318549726,
+      "learning_rate": 9.974702682902853e-06,
+      "loss": 1.1791,
+      "step": 6860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.636852289243506,
+      "learning_rate": 9.974537751370473e-06,
+      "loss": 1.1937,
+      "step": 6865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.576039674025727,
+      "learning_rate": 9.974372285302056e-06,
+      "loss": 1.1739,
+      "step": 6870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.079931641374136,
+      "learning_rate": 9.974206284715387e-06,
+      "loss": 1.2428,
+      "step": 6875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.4279107914430185,
+      "learning_rate": 9.974039749628297e-06,
+      "loss": 1.1797,
+      "step": 6880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.246179295928092,
+      "learning_rate": 9.973872680058685e-06,
+      "loss": 1.1558,
+      "step": 6885
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.78910949664825,
+      "learning_rate": 9.973705076024503e-06,
+      "loss": 1.198,
+      "step": 6890
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.0796416496715215,
+      "learning_rate": 9.973536937543758e-06,
+      "loss": 1.1151,
+      "step": 6895
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.01726678119124,
+      "learning_rate": 9.97336826463452e-06,
+      "loss": 1.1033,
+      "step": 6900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.893076369727662,
+      "learning_rate": 9.973199057314913e-06,
+      "loss": 1.1359,
+      "step": 6905
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.065964614117026,
+      "learning_rate": 9.973029315603118e-06,
+      "loss": 1.1998,
+      "step": 6910
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.56045272028396,
+      "learning_rate": 9.972859039517377e-06,
+      "loss": 1.1522,
+      "step": 6915
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.657183858925124,
+      "learning_rate": 9.972688229075984e-06,
+      "loss": 1.1471,
+      "step": 6920
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.611161738437362,
+      "learning_rate": 9.972516884297295e-06,
+      "loss": 1.1458,
+      "step": 6925
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.304912127698508,
+      "learning_rate": 9.972345005199721e-06,
+      "loss": 1.2101,
+      "step": 6930
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.495415165322395,
+      "learning_rate": 9.972172591801731e-06,
+      "loss": 1.2116,
+      "step": 6935
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.805608571776813,
+      "learning_rate": 9.971999644121855e-06,
+      "loss": 1.1854,
+      "step": 6940
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.142397334356656,
+      "learning_rate": 9.971826162178672e-06,
+      "loss": 1.1331,
+      "step": 6945
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.769205911723637,
+      "learning_rate": 9.971652145990826e-06,
+      "loss": 1.1581,
+      "step": 6950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.606059039236939,
+      "learning_rate": 9.971477595577015e-06,
+      "loss": 1.1466,
+      "step": 6955
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.909953088389111,
+      "learning_rate": 9.971302510955996e-06,
+      "loss": 1.1509,
+      "step": 6960
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.326337758290452,
+      "learning_rate": 9.97112689214658e-06,
+      "loss": 1.156,
+      "step": 6965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.277565895609338,
+      "learning_rate": 9.970950739167645e-06,
+      "loss": 1.1802,
+      "step": 6970
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.777569819012663,
+      "learning_rate": 9.97077405203811e-06,
+      "loss": 1.2291,
+      "step": 6975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.434539473523682,
+      "learning_rate": 9.970596830776967e-06,
+      "loss": 1.1833,
+      "step": 6980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.67024664056154,
+      "learning_rate": 9.970419075403256e-06,
+      "loss": 1.1416,
+      "step": 6985
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.134607097455263,
+      "learning_rate": 9.970240785936082e-06,
+      "loss": 1.1764,
+      "step": 6990
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.460788448567936,
+      "learning_rate": 9.970061962394599e-06,
+      "loss": 1.1595,
+      "step": 6995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.894085190018938,
+      "learning_rate": 9.969882604798025e-06,
+      "loss": 1.1614,
+      "step": 7000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.99500506951188,
+      "learning_rate": 9.96970271316563e-06,
+      "loss": 1.127,
+      "step": 7005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.801642701082123,
+      "learning_rate": 9.969522287516744e-06,
+      "loss": 1.1715,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.321390118992963,
+      "learning_rate": 9.969341327870758e-06,
+      "loss": 1.1757,
+      "step": 7015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.9823147197924635,
+      "learning_rate": 9.969159834247116e-06,
+      "loss": 1.0909,
+      "step": 7020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.693311659686022,
+      "learning_rate": 9.968977806665316e-06,
+      "loss": 1.1536,
+      "step": 7025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.082207239457722,
+      "learning_rate": 9.968795245144923e-06,
+      "loss": 1.1878,
+      "step": 7030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.624350030905093,
+      "learning_rate": 9.968612149705553e-06,
+      "loss": 1.2021,
+      "step": 7035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.263920645404605,
+      "learning_rate": 9.968428520366877e-06,
+      "loss": 1.1464,
+      "step": 7040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.201694449389493,
+      "learning_rate": 9.96824435714863e-06,
+      "loss": 1.2044,
+      "step": 7045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.152478309158454,
+      "learning_rate": 9.9680596600706e-06,
+      "loss": 1.207,
+      "step": 7050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.621848561179025,
+      "learning_rate": 9.967874429152635e-06,
+      "loss": 1.1679,
+      "step": 7055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.26393855465338,
+      "learning_rate": 9.967688664414636e-06,
+      "loss": 1.1895,
+      "step": 7060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.808273309055244,
+      "learning_rate": 9.967502365876567e-06,
+      "loss": 1.136,
+      "step": 7065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.1364997458048345,
+      "learning_rate": 9.967315533558445e-06,
+      "loss": 1.1617,
+      "step": 7070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.255141481393444,
+      "learning_rate": 9.967128167480346e-06,
+      "loss": 1.1395,
+      "step": 7075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.404988482522825,
+      "learning_rate": 9.966940267662407e-06,
+      "loss": 1.1422,
+      "step": 7080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.162902349791633,
+      "learning_rate": 9.966751834124812e-06,
+      "loss": 1.2098,
+      "step": 7085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.472829080961775,
+      "learning_rate": 9.966562866887813e-06,
+      "loss": 1.1637,
+      "step": 7090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.065553020026686,
+      "learning_rate": 9.966373365971716e-06,
+      "loss": 1.1489,
+      "step": 7095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.926763823743649,
+      "learning_rate": 9.966183331396882e-06,
+      "loss": 1.1942,
+      "step": 7100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.670843399481209,
+      "learning_rate": 9.965992763183732e-06,
+      "loss": 1.1627,
+      "step": 7105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.252928513189513,
+      "learning_rate": 9.965801661352742e-06,
+      "loss": 1.1862,
+      "step": 7110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.935201089724359,
+      "learning_rate": 9.965610025924449e-06,
+      "loss": 1.1522,
+      "step": 7115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.763877416311463,
+      "learning_rate": 9.965417856919442e-06,
+      "loss": 1.1204,
+      "step": 7120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.404757273478117,
+      "learning_rate": 9.965225154358374e-06,
+      "loss": 1.1906,
+      "step": 7125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.131156521871148,
+      "learning_rate": 9.965031918261949e-06,
+      "loss": 1.1561,
+      "step": 7130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.646992235178871,
+      "learning_rate": 9.964838148650932e-06,
+      "loss": 1.1779,
+      "step": 7135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.351613934878443,
+      "learning_rate": 9.964643845546144e-06,
+      "loss": 1.2187,
+      "step": 7140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.524885418216677,
+      "learning_rate": 9.964449008968466e-06,
+      "loss": 1.1756,
+      "step": 7145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.151302393127044,
+      "learning_rate": 9.96425363893883e-06,
+      "loss": 1.1154,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.447716616076492,
+      "learning_rate": 9.964057735478233e-06,
+      "loss": 1.2077,
+      "step": 7155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.235925532608152,
+      "learning_rate": 9.963861298607725e-06,
+      "loss": 1.1647,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.19404020149289,
+      "learning_rate": 9.963664328348411e-06,
+      "loss": 1.1872,
+      "step": 7165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.228989315369796,
+      "learning_rate": 9.96346682472146e-06,
+      "loss": 1.1619,
+      "step": 7170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.338968868833145,
+      "learning_rate": 9.963268787748092e-06,
+      "loss": 1.1989,
+      "step": 7175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.038334328549363,
+      "learning_rate": 9.963070217449591e-06,
+      "loss": 1.1683,
+      "step": 7180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.920020476923544,
+      "learning_rate": 9.962871113847288e-06,
+      "loss": 1.1962,
+      "step": 7185
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.116495517138593,
+      "learning_rate": 9.962671476962583e-06,
+      "loss": 1.1286,
+      "step": 7190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.209402600116505,
+      "learning_rate": 9.962471306816925e-06,
+      "loss": 1.1648,
+      "step": 7195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.265572653210617,
+      "learning_rate": 9.962270603431826e-06,
+      "loss": 1.1683,
+      "step": 7200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.246673041635199,
+      "learning_rate": 9.962069366828849e-06,
+      "loss": 1.1302,
+      "step": 7205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.521403920817779,
+      "learning_rate": 9.961867597029618e-06,
+      "loss": 1.1689,
+      "step": 7210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.598069555056234,
+      "learning_rate": 9.961665294055816e-06,
+      "loss": 1.1613,
+      "step": 7215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.443138575493562,
+      "learning_rate": 9.961462457929182e-06,
+      "loss": 1.1608,
+      "step": 7220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.528161260282685,
+      "learning_rate": 9.961259088671509e-06,
+      "loss": 1.1756,
+      "step": 7225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.732843000819472,
+      "learning_rate": 9.961055186304652e-06,
+      "loss": 1.1278,
+      "step": 7230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.928199552615537,
+      "learning_rate": 9.960850750850521e-06,
+      "loss": 1.1876,
+      "step": 7235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.7538052857280295,
+      "learning_rate": 9.960645782331083e-06,
+      "loss": 1.1372,
+      "step": 7240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.7527271250132,
+      "learning_rate": 9.96044028076836e-06,
+      "loss": 1.1318,
+      "step": 7245
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.569642847177645,
+      "learning_rate": 9.960234246184439e-06,
+      "loss": 1.1291,
+      "step": 7250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.466131806616952,
+      "learning_rate": 9.960027678601455e-06,
+      "loss": 1.1659,
+      "step": 7255
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.726398135012966,
+      "learning_rate": 9.95982057804161e-06,
+      "loss": 1.1753,
+      "step": 7260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.759427658793014,
+      "learning_rate": 9.959612944527152e-06,
+      "loss": 1.1702,
+      "step": 7265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.03926074299141,
+      "learning_rate": 9.959404778080395e-06,
+      "loss": 1.1924,
+      "step": 7270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.511034051194109,
+      "learning_rate": 9.959196078723708e-06,
+      "loss": 1.1421,
+      "step": 7275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.710780995117776,
+      "learning_rate": 9.958986846479515e-06,
+      "loss": 1.1799,
+      "step": 7280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 25.475026249973894,
+      "learning_rate": 9.9587770813703e-06,
+      "loss": 1.2095,
+      "step": 7285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.379179706896757,
+      "learning_rate": 9.958566783418603e-06,
+      "loss": 1.1971,
+      "step": 7290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.930843043565,
+      "learning_rate": 9.95835595264702e-06,
+      "loss": 1.1745,
+      "step": 7295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.1184329702384,
+      "learning_rate": 9.958144589078209e-06,
+      "loss": 1.1448,
+      "step": 7300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.214396997144293,
+      "learning_rate": 9.957932692734879e-06,
+      "loss": 1.1173,
+      "step": 7305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.193770064250026,
+      "learning_rate": 9.9577202636398e-06,
+      "loss": 1.1487,
+      "step": 7310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.036156338067784,
+      "learning_rate": 9.957507301815798e-06,
+      "loss": 1.1727,
+      "step": 7315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.139115079392365,
+      "learning_rate": 9.957293807285758e-06,
+      "loss": 1.1694,
+      "step": 7320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.630935230906257,
+      "learning_rate": 9.95707978007262e-06,
+      "loss": 1.1421,
+      "step": 7325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.386875452735694,
+      "learning_rate": 9.956865220199383e-06,
+      "loss": 1.182,
+      "step": 7330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.897287158181644,
+      "learning_rate": 9.956650127689101e-06,
+      "loss": 1.1908,
+      "step": 7335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.42602733433497,
+      "learning_rate": 9.956434502564888e-06,
+      "loss": 1.1796,
+      "step": 7340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.429001917172855,
+      "learning_rate": 9.956218344849912e-06,
+      "loss": 1.1795,
+      "step": 7345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.039869565325827,
+      "learning_rate": 9.956001654567404e-06,
+      "loss": 1.1654,
+      "step": 7350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.339977479867002,
+      "learning_rate": 9.955784431740643e-06,
+      "loss": 1.1704,
+      "step": 7355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.89971002783044,
+      "learning_rate": 9.955566676392975e-06,
+      "loss": 1.1557,
+      "step": 7360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.37787907433655,
+      "learning_rate": 9.955348388547797e-06,
+      "loss": 1.1279,
+      "step": 7365
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 58.86646691267968,
+      "learning_rate": 9.955129568228563e-06,
+      "loss": 1.1859,
+      "step": 7370
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 29.512956198196843,
+      "learning_rate": 9.95491021545879e-06,
+      "loss": 1.1538,
+      "step": 7375
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 35.16903316999807,
+      "learning_rate": 9.954690330262047e-06,
+      "loss": 1.1339,
+      "step": 7380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.539298402278382,
+      "learning_rate": 9.95446991266196e-06,
+      "loss": 1.1266,
+      "step": 7385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 28.489479868594366,
+      "learning_rate": 9.954248962682216e-06,
+      "loss": 1.1918,
+      "step": 7390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 27.764903782860227,
+      "learning_rate": 9.954027480346557e-06,
+      "loss": 1.1717,
+      "step": 7395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.632048537372299,
+      "learning_rate": 9.95380546567878e-06,
+      "loss": 1.1579,
+      "step": 7400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.585556934143545,
+      "learning_rate": 9.953582918702743e-06,
+      "loss": 1.1804,
+      "step": 7405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.3839843334155795,
+      "learning_rate": 9.95335983944236e-06,
+      "loss": 1.1692,
+      "step": 7410
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.115502943063046,
+      "learning_rate": 9.9531362279216e-06,
+      "loss": 1.1288,
+      "step": 7415
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.697219157530452,
+      "learning_rate": 9.952912084164494e-06,
+      "loss": 1.1831,
+      "step": 7420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.229541416735028,
+      "learning_rate": 9.952687408195126e-06,
+      "loss": 1.1768,
+      "step": 7425
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.249622746628285,
+      "learning_rate": 9.952462200037638e-06,
+      "loss": 1.1885,
+      "step": 7430
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.0974162297522865,
+      "learning_rate": 9.952236459716229e-06,
+      "loss": 1.2087,
+      "step": 7435
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.0724449592560745,
+      "learning_rate": 9.952010187255157e-06,
+      "loss": 1.1495,
+      "step": 7440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.3183452017936395,
+      "learning_rate": 9.951783382678735e-06,
+      "loss": 1.2043,
+      "step": 7445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.136328602147941,
+      "learning_rate": 9.951556046011336e-06,
+      "loss": 1.1497,
+      "step": 7450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.902056875987189,
+      "learning_rate": 9.951328177277385e-06,
+      "loss": 1.1182,
+      "step": 7455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.775983601548823,
+      "learning_rate": 9.95109977650137e-06,
+      "loss": 1.0865,
+      "step": 7460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.015998979312961,
+      "learning_rate": 9.950870843707834e-06,
+      "loss": 1.1436,
+      "step": 7465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.719815788065228,
+      "learning_rate": 9.950641378921375e-06,
+      "loss": 1.1483,
+      "step": 7470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.832901041077575,
+      "learning_rate": 9.950411382166652e-06,
+      "loss": 1.1446,
+      "step": 7475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.117160855859674,
+      "learning_rate": 9.950180853468377e-06,
+      "loss": 1.16,
+      "step": 7480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.2973352722122415,
+      "learning_rate": 9.949949792851323e-06,
+      "loss": 1.1186,
+      "step": 7485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.620669119265208,
+      "learning_rate": 9.949718200340319e-06,
+      "loss": 1.0882,
+      "step": 7490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.0780256771354,
+      "learning_rate": 9.949486075960248e-06,
+      "loss": 1.1225,
+      "step": 7495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.618159635632885,
+      "learning_rate": 9.949253419736055e-06,
+      "loss": 1.1239,
+      "step": 7500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.298895161700539,
+      "learning_rate": 9.94902023169274e-06,
+      "loss": 1.1855,
+      "step": 7505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.658594732022927,
+      "learning_rate": 9.948786511855358e-06,
+      "loss": 1.1472,
+      "step": 7510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.893464538077141,
+      "learning_rate": 9.948552260249025e-06,
+      "loss": 1.1582,
+      "step": 7515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.1407490289993705,
+      "learning_rate": 9.948317476898913e-06,
+      "loss": 1.149,
+      "step": 7520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.917156613629743,
+      "learning_rate": 9.948082161830249e-06,
+      "loss": 1.1368,
+      "step": 7525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.888994387264798,
+      "learning_rate": 9.947846315068319e-06,
+      "loss": 1.1887,
+      "step": 7530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.561917198848331,
+      "learning_rate": 9.947609936638468e-06,
+      "loss": 1.1206,
+      "step": 7535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.849109167201615,
+      "learning_rate": 9.947373026566089e-06,
+      "loss": 1.1509,
+      "step": 7540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.52580304317239,
+      "learning_rate": 9.947135584876649e-06,
+      "loss": 1.155,
+      "step": 7545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.229866299809071,
+      "learning_rate": 9.946897611595653e-06,
+      "loss": 1.201,
+      "step": 7550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.462107315609448,
+      "learning_rate": 9.94665910674868e-06,
+      "loss": 1.1297,
+      "step": 7555
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.331772324871634,
+      "learning_rate": 9.946420070361351e-06,
+      "loss": 1.1619,
+      "step": 7560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.34909614633429,
+      "learning_rate": 9.946180502459357e-06,
+      "loss": 1.0676,
+      "step": 7565
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.632008971359138,
+      "learning_rate": 9.94594040306844e-06,
+      "loss": 1.1648,
+      "step": 7570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.9404498847889675,
+      "learning_rate": 9.945699772214396e-06,
+      "loss": 1.0979,
+      "step": 7575
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.043731618251074,
+      "learning_rate": 9.945458609923087e-06,
+      "loss": 1.1777,
+      "step": 7580
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.426535997154359,
+      "learning_rate": 9.945216916220424e-06,
+      "loss": 1.1692,
+      "step": 7585
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.159336716140868,
+      "learning_rate": 9.944974691132377e-06,
+      "loss": 1.1385,
+      "step": 7590
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.590672728725128,
+      "learning_rate": 9.944731934684977e-06,
+      "loss": 1.1692,
+      "step": 7595
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.229077313744524,
+      "learning_rate": 9.944488646904307e-06,
+      "loss": 1.18,
+      "step": 7600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.825320467379788,
+      "learning_rate": 9.94424482781651e-06,
+      "loss": 1.1294,
+      "step": 7605
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.643517762309204,
+      "learning_rate": 9.944000477447786e-06,
+      "loss": 1.1281,
+      "step": 7610
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.815094794970016,
+      "learning_rate": 9.943755595824391e-06,
+      "loss": 1.1863,
+      "step": 7615
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.291662113231529,
+      "learning_rate": 9.94351018297264e-06,
+      "loss": 1.1704,
+      "step": 7620
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.313163076129426,
+      "learning_rate": 9.943264238918902e-06,
+      "loss": 1.1465,
+      "step": 7625
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.646545555699191,
+      "learning_rate": 9.943017763689604e-06,
+      "loss": 1.0835,
+      "step": 7630
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.405593689627591,
+      "learning_rate": 9.942770757311233e-06,
+      "loss": 1.1047,
+      "step": 7635
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.37489587592504,
+      "learning_rate": 9.94252321981033e-06,
+      "loss": 1.133,
+      "step": 7640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.762815622953501,
+      "learning_rate": 9.942275151213494e-06,
+      "loss": 1.1666,
+      "step": 7645
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.110801366127728,
+      "learning_rate": 9.942026551547379e-06,
+      "loss": 1.1827,
+      "step": 7650
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.930267312022218,
+      "learning_rate": 9.941777420838703e-06,
+      "loss": 1.1684,
+      "step": 7655
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.520164914312182,
+      "learning_rate": 9.941527759114233e-06,
+      "loss": 1.1296,
+      "step": 7660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.823430279904931,
+      "learning_rate": 9.941277566400796e-06,
+      "loss": 1.1771,
+      "step": 7665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.016326789857857,
+      "learning_rate": 9.941026842725276e-06,
+      "loss": 1.1328,
+      "step": 7670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.560869699096651,
+      "learning_rate": 9.940775588114618e-06,
+      "loss": 1.1103,
+      "step": 7675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.34295283052181,
+      "learning_rate": 9.940523802595815e-06,
+      "loss": 1.1335,
+      "step": 7680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.5167217444197965,
+      "learning_rate": 9.940271486195927e-06,
+      "loss": 1.1454,
+      "step": 7685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.396138579977828,
+      "learning_rate": 9.940018638942064e-06,
+      "loss": 1.1313,
+      "step": 7690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.22222680824031,
+      "learning_rate": 9.939765260861396e-06,
+      "loss": 1.15,
+      "step": 7695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.151215904216118,
+      "learning_rate": 9.93951135198115e-06,
+      "loss": 1.1359,
+      "step": 7700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.737205761509378,
+      "learning_rate": 9.93925691232861e-06,
+      "loss": 1.1412,
+      "step": 7705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.671639300887087,
+      "learning_rate": 9.939001941931117e-06,
+      "loss": 1.1733,
+      "step": 7710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.5574397761620045,
+      "learning_rate": 9.938746440816066e-06,
+      "loss": 1.1836,
+      "step": 7715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.85283106925243,
+      "learning_rate": 9.938490409010915e-06,
+      "loss": 1.1481,
+      "step": 7720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.541983204590011,
+      "learning_rate": 9.938233846543172e-06,
+      "loss": 1.1624,
+      "step": 7725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.380153215234209,
+      "learning_rate": 9.93797675344041e-06,
+      "loss": 1.1082,
+      "step": 7730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.2240106442182075,
+      "learning_rate": 9.937719129730255e-06,
+      "loss": 1.1443,
+      "step": 7735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.153292924266712,
+      "learning_rate": 9.937460975440386e-06,
+      "loss": 1.1028,
+      "step": 7740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.5515823330429805,
+      "learning_rate": 9.937202290598541e-06,
+      "loss": 1.1233,
+      "step": 7745
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.039659230667111,
+      "learning_rate": 9.936943075232525e-06,
+      "loss": 1.163,
+      "step": 7750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.662962324633087,
+      "learning_rate": 9.936683329370184e-06,
+      "loss": 1.1696,
+      "step": 7755
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.239963909223691,
+      "learning_rate": 9.936423053039434e-06,
+      "loss": 1.1282,
+      "step": 7760
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.721532366769768,
+      "learning_rate": 9.936162246268242e-06,
+      "loss": 1.1587,
+      "step": 7765
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.721408140815109,
+      "learning_rate": 9.935900909084629e-06,
+      "loss": 1.0901,
+      "step": 7770
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.135733225197569,
+      "learning_rate": 9.93563904151668e-06,
+      "loss": 1.1181,
+      "step": 7775
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.593760239533519,
+      "learning_rate": 9.935376643592534e-06,
+      "loss": 1.1923,
+      "step": 7780
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.112961542721804,
+      "learning_rate": 9.935113715340386e-06,
+      "loss": 1.1687,
+      "step": 7785
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.566380989507593,
+      "learning_rate": 9.934850256788489e-06,
+      "loss": 1.1062,
+      "step": 7790
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.984241345080738,
+      "learning_rate": 9.934586267965154e-06,
+      "loss": 1.2385,
+      "step": 7795
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.231142467968554,
+      "learning_rate": 9.934321748898746e-06,
+      "loss": 1.1334,
+      "step": 7800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.889388377549382,
+      "learning_rate": 9.934056699617686e-06,
+      "loss": 1.1132,
+      "step": 7805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.186228712007964,
+      "learning_rate": 9.933791120150461e-06,
+      "loss": 1.1427,
+      "step": 7810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.151427183839397,
+      "learning_rate": 9.933525010525607e-06,
+      "loss": 1.1325,
+      "step": 7815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.792545697946491,
+      "learning_rate": 9.933258370771714e-06,
+      "loss": 1.1537,
+      "step": 7820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.400827415408557,
+      "learning_rate": 9.932991200917438e-06,
+      "loss": 1.1571,
+      "step": 7825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.807104821525981,
+      "learning_rate": 9.932723500991488e-06,
+      "loss": 1.1225,
+      "step": 7830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.30671607188519,
+      "learning_rate": 9.932455271022627e-06,
+      "loss": 1.1676,
+      "step": 7835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.924940762504854,
+      "learning_rate": 9.93218651103968e-06,
+      "loss": 1.1827,
+      "step": 7840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.50411716805518,
+      "learning_rate": 9.931917221071522e-06,
+      "loss": 1.1229,
+      "step": 7845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.676640545210377,
+      "learning_rate": 9.931647401147093e-06,
+      "loss": 1.1518,
+      "step": 7850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.80959535965297,
+      "learning_rate": 9.931377051295389e-06,
+      "loss": 1.1297,
+      "step": 7855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.811323956525596,
+      "learning_rate": 9.931106171545453e-06,
+      "loss": 1.1339,
+      "step": 7860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.548378617609879,
+      "learning_rate": 9.930834761926398e-06,
+      "loss": 1.0706,
+      "step": 7865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.138696070070708,
+      "learning_rate": 9.930562822467386e-06,
+      "loss": 1.1457,
+      "step": 7870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.851137940988485,
+      "learning_rate": 9.93029035319764e-06,
+      "loss": 1.1849,
+      "step": 7875
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.246615709232681,
+      "learning_rate": 9.930017354146435e-06,
+      "loss": 1.2075,
+      "step": 7880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.004687896127214,
+      "learning_rate": 9.929743825343109e-06,
+      "loss": 1.1319,
+      "step": 7885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.31333613381446,
+      "learning_rate": 9.929469766817049e-06,
+      "loss": 1.1473,
+      "step": 7890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.947978349580737,
+      "learning_rate": 9.929195178597711e-06,
+      "loss": 1.1454,
+      "step": 7895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.596565430355072,
+      "learning_rate": 9.928920060714594e-06,
+      "loss": 1.1403,
+      "step": 7900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.853373502977693,
+      "learning_rate": 9.928644413197263e-06,
+      "loss": 1.2014,
+      "step": 7905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.177412393431067,
+      "learning_rate": 9.92836823607534e-06,
+      "loss": 1.1609,
+      "step": 7910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.7587644267845794,
+      "learning_rate": 9.9280915293785e-06,
+      "loss": 1.1264,
+      "step": 7915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.568438535524381,
+      "learning_rate": 9.927814293136475e-06,
+      "loss": 1.1654,
+      "step": 7920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.808904676175661,
+      "learning_rate": 9.927536527379055e-06,
+      "loss": 1.112,
+      "step": 7925
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.428016435169493,
+      "learning_rate": 9.92725823213609e-06,
+      "loss": 1.1196,
+      "step": 7930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.161113797198686,
+      "learning_rate": 9.926979407437482e-06,
+      "loss": 1.1217,
+      "step": 7935
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.182527071175802,
+      "learning_rate": 9.926700053313192e-06,
+      "loss": 1.1347,
+      "step": 7940
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.627434994800816,
+      "learning_rate": 9.926420169793238e-06,
+      "loss": 1.1128,
+      "step": 7945
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.476428644298823,
+      "learning_rate": 9.926139756907696e-06,
+      "loss": 1.1408,
+      "step": 7950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.018142300408826,
+      "learning_rate": 9.925858814686695e-06,
+      "loss": 1.1461,
+      "step": 7955
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.01193884729803,
+      "learning_rate": 9.925577343160424e-06,
+      "loss": 1.1665,
+      "step": 7960
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.862190238650837,
+      "learning_rate": 9.925295342359131e-06,
+      "loss": 1.1839,
+      "step": 7965
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.228009776446208,
+      "learning_rate": 9.925012812313114e-06,
+      "loss": 1.1357,
+      "step": 7970
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.8908933139796895,
+      "learning_rate": 9.924729753052737e-06,
+      "loss": 1.1461,
+      "step": 7975
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.978023474287168,
+      "learning_rate": 9.924446164608414e-06,
+      "loss": 1.1718,
+      "step": 7980
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.084048319902944,
+      "learning_rate": 9.924162047010617e-06,
+      "loss": 1.1096,
+      "step": 7985
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.463896890378928,
+      "learning_rate": 9.923877400289873e-06,
+      "loss": 1.1951,
+      "step": 7990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.068209068736689,
+      "learning_rate": 9.923592224476776e-06,
+      "loss": 1.1646,
+      "step": 7995
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.88910232611772,
+      "learning_rate": 9.923306519601963e-06,
+      "loss": 1.1643,
+      "step": 8000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.812356468825577,
+      "learning_rate": 9.923020285696135e-06,
+      "loss": 1.1443,
+      "step": 8005
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.527069223093425,
+      "learning_rate": 9.922733522790052e-06,
+      "loss": 1.1663,
+      "step": 8010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.085587139001899,
+      "learning_rate": 9.922446230914526e-06,
+      "loss": 1.1198,
+      "step": 8015
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.178732078404221,
+      "learning_rate": 9.922158410100429e-06,
+      "loss": 1.1256,
+      "step": 8020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.751355985385874,
+      "learning_rate": 9.921870060378687e-06,
+      "loss": 1.1237,
+      "step": 8025
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.166352451055253,
+      "learning_rate": 9.921581181780284e-06,
+      "loss": 1.1568,
+      "step": 8030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.750446785875491,
+      "learning_rate": 9.921291774336263e-06,
+      "loss": 1.1299,
+      "step": 8035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.322602400290547,
+      "learning_rate": 9.921001838077722e-06,
+      "loss": 1.1105,
+      "step": 8040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.597169205942858,
+      "learning_rate": 9.920711373035815e-06,
+      "loss": 1.1518,
+      "step": 8045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.550861867124471,
+      "learning_rate": 9.920420379241755e-06,
+      "loss": 1.1043,
+      "step": 8050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.414144737746152,
+      "learning_rate": 9.920128856726809e-06,
+      "loss": 1.1357,
+      "step": 8055
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.7879234758635,
+      "learning_rate": 9.919836805522303e-06,
+      "loss": 1.181,
+      "step": 8060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.935363874803764,
+      "learning_rate": 9.91954422565962e-06,
+      "loss": 1.1308,
+      "step": 8065
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.197737171064103,
+      "learning_rate": 9.919251117170197e-06,
+      "loss": 1.1213,
+      "step": 8070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.491328310522661,
+      "learning_rate": 9.918957480085534e-06,
+      "loss": 1.1616,
+      "step": 8075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.186871938635413,
+      "learning_rate": 9.918663314437178e-06,
+      "loss": 1.1603,
+      "step": 8080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.805034098385828,
+      "learning_rate": 9.918368620256742e-06,
+      "loss": 1.1597,
+      "step": 8085
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.897184300130437,
+      "learning_rate": 9.918073397575892e-06,
+      "loss": 1.1258,
+      "step": 8090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.925310029747384,
+      "learning_rate": 9.91777764642635e-06,
+      "loss": 1.1098,
+      "step": 8095
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.66010055733957,
+      "learning_rate": 9.917481366839896e-06,
+      "loss": 1.1646,
+      "step": 8100
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.550626296738047,
+      "learning_rate": 9.917184558848366e-06,
+      "loss": 1.1764,
+      "step": 8105
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.349916065932398,
+      "learning_rate": 9.916887222483654e-06,
+      "loss": 1.1355,
+      "step": 8110
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.41914856625254,
+      "learning_rate": 9.91658935777771e-06,
+      "loss": 1.1179,
+      "step": 8115
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.949966070018647,
+      "learning_rate": 9.916290964762542e-06,
+      "loss": 1.0995,
+      "step": 8120
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.116656904071101,
+      "learning_rate": 9.915992043470213e-06,
+      "loss": 1.1451,
+      "step": 8125
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.57665638720786,
+      "learning_rate": 9.915692593932842e-06,
+      "loss": 1.0656,
+      "step": 8130
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.121901944957988,
+      "learning_rate": 9.915392616182606e-06,
+      "loss": 1.1285,
+      "step": 8135
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.763329818073514,
+      "learning_rate": 9.915092110251742e-06,
+      "loss": 1.1415,
+      "step": 8140
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.174463104279873,
+      "learning_rate": 9.914791076172539e-06,
+      "loss": 1.1603,
+      "step": 8145
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.687202572936014,
+      "learning_rate": 9.914489513977344e-06,
+      "loss": 1.173,
+      "step": 8150
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.025317968688368,
+      "learning_rate": 9.91418742369856e-06,
+      "loss": 1.1799,
+      "step": 8155
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.580685167669746,
+      "learning_rate": 9.91388480536865e-06,
+      "loss": 1.1417,
+      "step": 8160
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.427871347485707,
+      "learning_rate": 9.913581659020131e-06,
+      "loss": 1.1431,
+      "step": 8165
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.919016707093965,
+      "learning_rate": 9.913277984685577e-06,
+      "loss": 1.1343,
+      "step": 8170
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.0728116259922,
+      "learning_rate": 9.91297378239762e-06,
+      "loss": 1.1714,
+      "step": 8175
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.784404388459732,
+      "learning_rate": 9.912669052188948e-06,
+      "loss": 1.1071,
+      "step": 8180
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.867643724095714,
+      "learning_rate": 9.912363794092306e-06,
+      "loss": 1.1165,
+      "step": 8185
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.227876275716403,
+      "learning_rate": 9.912058008140493e-06,
+      "loss": 1.1179,
+      "step": 8190
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.926626840265723,
+      "learning_rate": 9.911751694366368e-06,
+      "loss": 1.1547,
+      "step": 8195
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.251534839962707,
+      "learning_rate": 9.911444852802848e-06,
+      "loss": 1.1411,
+      "step": 8200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.8221131575983085,
+      "learning_rate": 9.911137483482902e-06,
+      "loss": 1.1618,
+      "step": 8205
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.951383048318903,
+      "learning_rate": 9.910829586439557e-06,
+      "loss": 1.1901,
+      "step": 8210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.394284096549184,
+      "learning_rate": 9.910521161705901e-06,
+      "loss": 1.1189,
+      "step": 8215
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.058500387595784,
+      "learning_rate": 9.910212209315075e-06,
+      "loss": 1.1217,
+      "step": 8220
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.475745755358913,
+      "learning_rate": 9.909902729300276e-06,
+      "loss": 1.1029,
+      "step": 8225
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.321152476451876,
+      "learning_rate": 9.90959272169476e-06,
+      "loss": 1.165,
+      "step": 8230
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.709737407683546,
+      "learning_rate": 9.909282186531838e-06,
+      "loss": 1.1298,
+      "step": 8235
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.942039054846149,
+      "learning_rate": 9.90897112384488e-06,
+      "loss": 1.1079,
+      "step": 8240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.523822588201016,
+      "learning_rate": 9.908659533667308e-06,
+      "loss": 1.0914,
+      "step": 8245
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.18599056710619,
+      "learning_rate": 9.908347416032607e-06,
+      "loss": 1.131,
+      "step": 8250
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.298366837145794,
+      "learning_rate": 9.908034770974314e-06,
+      "loss": 1.1187,
+      "step": 8255
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.266402219942813,
+      "learning_rate": 9.907721598526023e-06,
+      "loss": 1.0883,
+      "step": 8260
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.441723810002678,
+      "learning_rate": 9.907407898721388e-06,
+      "loss": 1.176,
+      "step": 8265
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.332685159253233,
+      "learning_rate": 9.907093671594116e-06,
+      "loss": 1.1322,
+      "step": 8270
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.278492210164009,
+      "learning_rate": 9.906778917177971e-06,
+      "loss": 1.1418,
+      "step": 8275
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.144336594304728,
+      "learning_rate": 9.906463635506778e-06,
+      "loss": 1.1232,
+      "step": 8280
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.062488243792572,
+      "learning_rate": 9.906147826614412e-06,
+      "loss": 1.1556,
+      "step": 8285
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.890762453851772,
+      "learning_rate": 9.90583149053481e-06,
+      "loss": 1.1198,
+      "step": 8290
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.562691160318177,
+      "learning_rate": 9.905514627301964e-06,
+      "loss": 1.1202,
+      "step": 8295
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.883518091845976,
+      "learning_rate": 9.90519723694992e-06,
+      "loss": 1.1748,
+      "step": 8300
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.3553947016629015,
+      "learning_rate": 9.904879319512786e-06,
+      "loss": 1.1136,
+      "step": 8305
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.529370012571578,
+      "learning_rate": 9.904560875024722e-06,
+      "loss": 1.1744,
+      "step": 8310
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.86046118665941,
+      "learning_rate": 9.904241903519946e-06,
+      "loss": 1.1164,
+      "step": 8315
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.259825004498953,
+      "learning_rate": 9.903922405032733e-06,
+      "loss": 1.1582,
+      "step": 8320
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.898525371631383,
+      "learning_rate": 9.903602379597417e-06,
+      "loss": 1.1757,
+      "step": 8325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.642737075991965,
+      "learning_rate": 9.90328182724838e-06,
+      "loss": 1.1861,
+      "step": 8330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.60510125662481,
+      "learning_rate": 9.902960748020075e-06,
+      "loss": 1.1455,
+      "step": 8335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.650827903465728,
+      "learning_rate": 9.902639141946998e-06,
+      "loss": 1.1599,
+      "step": 8340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.65537674984556,
+      "learning_rate": 9.902317009063708e-06,
+      "loss": 1.1387,
+      "step": 8345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.416947762342985,
+      "learning_rate": 9.901994349404819e-06,
+      "loss": 1.1795,
+      "step": 8350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.1507077578132,
+      "learning_rate": 9.901671163005005e-06,
+      "loss": 1.149,
+      "step": 8355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.715233692577106,
+      "learning_rate": 9.90134744989899e-06,
+      "loss": 1.1451,
+      "step": 8360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.030405268061639,
+      "learning_rate": 9.901023210121561e-06,
+      "loss": 1.1271,
+      "step": 8365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.825841721827905,
+      "learning_rate": 9.900698443707557e-06,
+      "loss": 1.1039,
+      "step": 8370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.09960502605246,
+      "learning_rate": 9.900373150691878e-06,
+      "loss": 1.1529,
+      "step": 8375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.932854658669767,
+      "learning_rate": 9.900047331109477e-06,
+      "loss": 1.1195,
+      "step": 8380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.611839452996128,
+      "learning_rate": 9.899720984995364e-06,
+      "loss": 1.1164,
+      "step": 8385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.143709420338695,
+      "learning_rate": 9.899394112384607e-06,
+      "loss": 1.1199,
+      "step": 8390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.93958495130495,
+      "learning_rate": 9.89906671331233e-06,
+      "loss": 1.1552,
+      "step": 8395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.113234791150843,
+      "learning_rate": 9.898738787813714e-06,
+      "loss": 1.098,
+      "step": 8400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.88229020201969,
+      "learning_rate": 9.898410335923996e-06,
+      "loss": 1.1889,
+      "step": 8405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.719444611762174,
+      "learning_rate": 9.898081357678468e-06,
+      "loss": 1.1312,
+      "step": 8410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.299351015474322,
+      "learning_rate": 9.89775185311248e-06,
+      "loss": 1.0726,
+      "step": 8415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.398252044924147,
+      "learning_rate": 9.897421822261442e-06,
+      "loss": 1.166,
+      "step": 8420
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.370785867700546,
+      "learning_rate": 9.897091265160816e-06,
+      "loss": 1.136,
+      "step": 8425
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.826027373171965,
+      "learning_rate": 9.896760181846119e-06,
+      "loss": 1.0938,
+      "step": 8430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.512961897362741,
+      "learning_rate": 9.896428572352931e-06,
+      "loss": 1.1433,
+      "step": 8435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.732017075522059,
+      "learning_rate": 9.896096436716881e-06,
+      "loss": 1.1107,
+      "step": 8440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.09926900760528,
+      "learning_rate": 9.895763774973664e-06,
+      "loss": 1.1526,
+      "step": 8445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.859586133410717,
+      "learning_rate": 9.895430587159021e-06,
+      "loss": 1.113,
+      "step": 8450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.243516666618795,
+      "learning_rate": 9.895096873308757e-06,
+      "loss": 1.1483,
+      "step": 8455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.918660122367753,
+      "learning_rate": 9.894762633458728e-06,
+      "loss": 1.124,
+      "step": 8460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.463828178244237,
+      "learning_rate": 9.894427867644854e-06,
+      "loss": 1.1644,
+      "step": 8465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.938550384195903,
+      "learning_rate": 9.894092575903102e-06,
+      "loss": 1.1308,
+      "step": 8470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 20.086290071419377,
+      "learning_rate": 9.893756758269506e-06,
+      "loss": 1.2175,
+      "step": 8475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 28.692590544152967,
+      "learning_rate": 9.893420414780148e-06,
+      "loss": 1.1635,
+      "step": 8480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.991419829713242,
+      "learning_rate": 9.893083545471172e-06,
+      "loss": 1.1577,
+      "step": 8485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.811681583399924,
+      "learning_rate": 9.892746150378772e-06,
+      "loss": 1.1738,
+      "step": 8490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.16934930585563,
+      "learning_rate": 9.892408229539206e-06,
+      "loss": 1.0848,
+      "step": 8495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.418373610077596,
+      "learning_rate": 9.892069782988783e-06,
+      "loss": 1.1275,
+      "step": 8500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.7309271487165825,
+      "learning_rate": 9.891730810763872e-06,
+      "loss": 1.1448,
+      "step": 8505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.856025700983017,
+      "learning_rate": 9.891391312900895e-06,
+      "loss": 1.1658,
+      "step": 8510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.637788056262348,
+      "learning_rate": 9.891051289436335e-06,
+      "loss": 1.1244,
+      "step": 8515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.527743712569118,
+      "learning_rate": 9.890710740406728e-06,
+      "loss": 1.1124,
+      "step": 8520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.382154647775197,
+      "learning_rate": 9.890369665848667e-06,
+      "loss": 1.1388,
+      "step": 8525
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 22.201215017275665,
+      "learning_rate": 9.890028065798805e-06,
+      "loss": 1.1463,
+      "step": 8530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.946811844953196,
+      "learning_rate": 9.889685940293843e-06,
+      "loss": 1.1496,
+      "step": 8535
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.018285658393802,
+      "learning_rate": 9.889343289370549e-06,
+      "loss": 1.1317,
+      "step": 8540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.327708999852175,
+      "learning_rate": 9.88900011306574e-06,
+      "loss": 1.1272,
+      "step": 8545
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.334283488082707,
+      "learning_rate": 9.888656411416292e-06,
+      "loss": 1.1236,
+      "step": 8550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.873331119329881,
+      "learning_rate": 9.888312184459137e-06,
+      "loss": 1.151,
+      "step": 8555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.367691852507198,
+      "learning_rate": 9.887967432231264e-06,
+      "loss": 1.0753,
+      "step": 8560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.632051776820594,
+      "learning_rate": 9.887622154769718e-06,
+      "loss": 1.0876,
+      "step": 8565
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.774701151149621,
+      "learning_rate": 9.8872763521116e-06,
+      "loss": 1.2065,
+      "step": 8570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.0813360645562,
+      "learning_rate": 9.886930024294071e-06,
+      "loss": 1.1686,
+      "step": 8575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 31.83631671917243,
+      "learning_rate": 9.88658317135434e-06,
+      "loss": 1.1813,
+      "step": 8580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.237608240177158,
+      "learning_rate": 9.886235793329684e-06,
+      "loss": 1.1262,
+      "step": 8585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.974889676681796,
+      "learning_rate": 9.885887890257426e-06,
+      "loss": 1.0847,
+      "step": 8590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.314712493839495,
+      "learning_rate": 9.885539462174951e-06,
+      "loss": 1.1294,
+      "step": 8595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.084275999170101,
+      "learning_rate": 9.8851905091197e-06,
+      "loss": 1.1343,
+      "step": 8600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.725744409281265,
+      "learning_rate": 9.884841031129168e-06,
+      "loss": 1.104,
+      "step": 8605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.49318806526292,
+      "learning_rate": 9.88449102824091e-06,
+      "loss": 1.1604,
+      "step": 8610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.260404077681832,
+      "learning_rate": 9.884140500492534e-06,
+      "loss": 1.1262,
+      "step": 8615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.042390101029518,
+      "learning_rate": 9.883789447921704e-06,
+      "loss": 1.1761,
+      "step": 8620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.651268719845782,
+      "learning_rate": 9.883437870566145e-06,
+      "loss": 1.1607,
+      "step": 8625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.810416629282749,
+      "learning_rate": 9.883085768463636e-06,
+      "loss": 1.0684,
+      "step": 8630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.49464688357209,
+      "learning_rate": 9.88273314165201e-06,
+      "loss": 1.1466,
+      "step": 8635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.831790418939747,
+      "learning_rate": 9.882379990169157e-06,
+      "loss": 1.1272,
+      "step": 8640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.087714408267669,
+      "learning_rate": 9.882026314053028e-06,
+      "loss": 1.1517,
+      "step": 8645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.38482569343233,
+      "learning_rate": 9.881672113341625e-06,
+      "loss": 1.0987,
+      "step": 8650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.978551408443714,
+      "learning_rate": 9.881317388073008e-06,
+      "loss": 1.1682,
+      "step": 8655
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.577930016224028,
+      "learning_rate": 9.880962138285296e-06,
+      "loss": 1.1361,
+      "step": 8660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.003460191777767,
+      "learning_rate": 9.88060636401666e-06,
+      "loss": 1.0991,
+      "step": 8665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.613147576585526,
+      "learning_rate": 9.880250065305333e-06,
+      "loss": 1.0562,
+      "step": 8670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.268487046062377,
+      "learning_rate": 9.879893242189596e-06,
+      "loss": 1.1133,
+      "step": 8675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.275443322415458,
+      "learning_rate": 9.879535894707793e-06,
+      "loss": 1.1094,
+      "step": 8680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.671029277425907,
+      "learning_rate": 9.879178022898325e-06,
+      "loss": 1.1468,
+      "step": 8685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.411401122550819,
+      "learning_rate": 9.878819626799644e-06,
+      "loss": 1.1176,
+      "step": 8690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.55076792576607,
+      "learning_rate": 9.878460706450261e-06,
+      "loss": 1.1391,
+      "step": 8695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.601832477972204,
+      "learning_rate": 9.878101261888747e-06,
+      "loss": 1.1068,
+      "step": 8700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.325638205613911,
+      "learning_rate": 9.877741293153721e-06,
+      "loss": 1.1427,
+      "step": 8705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.384084077786898,
+      "learning_rate": 9.877380800283869e-06,
+      "loss": 1.0938,
+      "step": 8710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.743329701469579,
+      "learning_rate": 9.877019783317924e-06,
+      "loss": 1.1259,
+      "step": 8715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.85590857224492,
+      "learning_rate": 9.876658242294679e-06,
+      "loss": 1.1482,
+      "step": 8720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.252324703638708,
+      "learning_rate": 9.876296177252982e-06,
+      "loss": 1.0955,
+      "step": 8725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.543101434848318,
+      "learning_rate": 9.87593358823174e-06,
+      "loss": 1.1265,
+      "step": 8730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.11810196707246,
+      "learning_rate": 9.875570475269918e-06,
+      "loss": 1.117,
+      "step": 8735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.870819290397085,
+      "learning_rate": 9.875206838406528e-06,
+      "loss": 1.1028,
+      "step": 8740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.127359370236454,
+      "learning_rate": 9.874842677680646e-06,
+      "loss": 1.1071,
+      "step": 8745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.461513470793973,
+      "learning_rate": 9.874477993131408e-06,
+      "loss": 1.1503,
+      "step": 8750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.90646484454837,
+      "learning_rate": 9.874112784797996e-06,
+      "loss": 1.1434,
+      "step": 8755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.401254201075805,
+      "learning_rate": 9.873747052719653e-06,
+      "loss": 1.1286,
+      "step": 8760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.199430274042413,
+      "learning_rate": 9.873380796935678e-06,
+      "loss": 1.0951,
+      "step": 8765
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.247900062417915,
+      "learning_rate": 9.873014017485432e-06,
+      "loss": 1.102,
+      "step": 8770
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.515539280285301,
+      "learning_rate": 9.87264671440832e-06,
+      "loss": 1.1216,
+      "step": 8775
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.969039162871018,
+      "learning_rate": 9.872278887743817e-06,
+      "loss": 1.106,
+      "step": 8780
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.045261243073863,
+      "learning_rate": 9.871910537531444e-06,
+      "loss": 1.1104,
+      "step": 8785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.103090105477524,
+      "learning_rate": 9.87154166381078e-06,
+      "loss": 1.1338,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.637827420064879,
+      "learning_rate": 9.871172266621467e-06,
+      "loss": 1.089,
+      "step": 8795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.192056301224735,
+      "learning_rate": 9.870802346003194e-06,
+      "loss": 1.159,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.177519571168098,
+      "learning_rate": 9.870431901995713e-06,
+      "loss": 1.1193,
+      "step": 8805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.03963120388044,
+      "learning_rate": 9.87006093463883e-06,
+      "loss": 1.1666,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.5455509691663,
+      "learning_rate": 9.869689443972407e-06,
+      "loss": 1.1137,
+      "step": 8815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.175597327505168,
+      "learning_rate": 9.869317430036361e-06,
+      "loss": 1.1661,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.13390041830503,
+      "learning_rate": 9.868944892870667e-06,
+      "loss": 1.161,
+      "step": 8825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.106778316679549,
+      "learning_rate": 9.868571832515356e-06,
+      "loss": 1.1145,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.580016108080615,
+      "learning_rate": 9.868198249010515e-06,
+      "loss": 1.1323,
+      "step": 8835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.646913338222264,
+      "learning_rate": 9.867824142396287e-06,
+      "loss": 1.1246,
+      "step": 8840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.789300711494043,
+      "learning_rate": 9.86744951271287e-06,
+      "loss": 1.1512,
+      "step": 8845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.44934645700223,
+      "learning_rate": 9.867074360000524e-06,
+      "loss": 1.1336,
+      "step": 8850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.130068306637817,
+      "learning_rate": 9.866698684299556e-06,
+      "loss": 1.1355,
+      "step": 8855
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.38322882222291,
+      "learning_rate": 9.866322485650338e-06,
+      "loss": 1.1629,
+      "step": 8860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.17977244139066,
+      "learning_rate": 9.86594576409329e-06,
+      "loss": 1.1563,
+      "step": 8865
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 36.22213876931674,
+      "learning_rate": 9.865568519668896e-06,
+      "loss": 1.0806,
+      "step": 8870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.914144236726561,
+      "learning_rate": 9.865190752417691e-06,
+      "loss": 1.1389,
+      "step": 8875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 18.950702674673007,
+      "learning_rate": 9.864812462380267e-06,
+      "loss": 1.1194,
+      "step": 8880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 21.554579307891707,
+      "learning_rate": 9.864433649597274e-06,
+      "loss": 1.0886,
+      "step": 8885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.952643715967486,
+      "learning_rate": 9.864054314109417e-06,
+      "loss": 1.1137,
+      "step": 8890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.188043647866964,
+      "learning_rate": 9.863674455957457e-06,
+      "loss": 1.1232,
+      "step": 8895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.059623519784218,
+      "learning_rate": 9.863294075182212e-06,
+      "loss": 1.0934,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.50108779868795,
+      "learning_rate": 9.862913171824557e-06,
+      "loss": 1.0984,
+      "step": 8905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.470460801032964,
+      "learning_rate": 9.862531745925419e-06,
+      "loss": 1.1431,
+      "step": 8910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.635266053974252,
+      "learning_rate": 9.862149797525785e-06,
+      "loss": 1.1121,
+      "step": 8915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.74285422981206,
+      "learning_rate": 9.861767326666695e-06,
+      "loss": 1.0956,
+      "step": 8920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.988992153452946,
+      "learning_rate": 9.86138433338925e-06,
+      "loss": 1.1517,
+      "step": 8925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.902158338876868,
+      "learning_rate": 9.861000817734605e-06,
+      "loss": 1.1701,
+      "step": 8930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.715489638432372,
+      "learning_rate": 9.860616779743968e-06,
+      "loss": 1.1776,
+      "step": 8935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.804661819538458,
+      "learning_rate": 9.860232219458607e-06,
+      "loss": 1.1341,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 43.304049769680255,
+      "learning_rate": 9.859847136919842e-06,
+      "loss": 1.1533,
+      "step": 8945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 26.25033258227799,
+      "learning_rate": 9.859461532169057e-06,
+      "loss": 1.1089,
+      "step": 8950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.19941763309832,
+      "learning_rate": 9.859075405247682e-06,
+      "loss": 1.1551,
+      "step": 8955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 32.981446791884494,
+      "learning_rate": 9.85868875619721e-06,
+      "loss": 1.1759,
+      "step": 8960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 21.893948142009414,
+      "learning_rate": 9.858301585059188e-06,
+      "loss": 1.1261,
+      "step": 8965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.915641489935265,
+      "learning_rate": 9.857913891875222e-06,
+      "loss": 1.1577,
+      "step": 8970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.042933842324215,
+      "learning_rate": 9.857525676686966e-06,
+      "loss": 1.1192,
+      "step": 8975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 27.99212123543072,
+      "learning_rate": 9.857136939536137e-06,
+      "loss": 1.117,
+      "step": 8980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 18.57975891297702,
+      "learning_rate": 9.856747680464511e-06,
+      "loss": 1.1142,
+      "step": 8985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 44.46051231204827,
+      "learning_rate": 9.856357899513911e-06,
+      "loss": 1.1014,
+      "step": 8990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.635132820820676,
+      "learning_rate": 9.85596759672622e-06,
+      "loss": 1.0509,
+      "step": 8995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 22.916868614206592,
+      "learning_rate": 9.855576772143383e-06,
+      "loss": 1.1111,
+      "step": 9000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.594514080170606,
+      "learning_rate": 9.85518542580739e-06,
+      "loss": 1.1419,
+      "step": 9005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 20.252084074455954,
+      "learning_rate": 9.854793557760298e-06,
+      "loss": 1.0845,
+      "step": 9010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 25.571367513281952,
+      "learning_rate": 9.85440116804421e-06,
+      "loss": 1.1459,
+      "step": 9015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.735541352000926,
+      "learning_rate": 9.854008256701293e-06,
+      "loss": 1.1173,
+      "step": 9020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.966493016531478,
+      "learning_rate": 9.853614823773768e-06,
+      "loss": 1.1297,
+      "step": 9025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.44739383925974,
+      "learning_rate": 9.853220869303907e-06,
+      "loss": 1.1311,
+      "step": 9030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.261661916692985,
+      "learning_rate": 9.852826393334045e-06,
+      "loss": 1.1272,
+      "step": 9035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.115260549378783,
+      "learning_rate": 9.852431395906571e-06,
+      "loss": 1.1183,
+      "step": 9040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.911338588302018,
+      "learning_rate": 9.852035877063928e-06,
+      "loss": 1.1403,
+      "step": 9045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.65319648960198,
+      "learning_rate": 9.851639836848617e-06,
+      "loss": 1.1183,
+      "step": 9050
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.598607649715242,
+      "learning_rate": 9.851243275303193e-06,
+      "loss": 1.1406,
+      "step": 9055
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.302947523515613,
+      "learning_rate": 9.850846192470268e-06,
+      "loss": 1.1586,
+      "step": 9060
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.521566317114554,
+      "learning_rate": 9.850448588392512e-06,
+      "loss": 1.0928,
+      "step": 9065
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.635274741612458,
+      "learning_rate": 9.850050463112649e-06,
+      "loss": 1.1316,
+      "step": 9070
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 23.599549137314003,
+      "learning_rate": 9.849651816673458e-06,
+      "loss": 1.1373,
+      "step": 9075
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.554228488313795,
+      "learning_rate": 9.849252649117776e-06,
+      "loss": 1.1979,
+      "step": 9080
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.777262147948352,
+      "learning_rate": 9.848852960488496e-06,
+      "loss": 1.1417,
+      "step": 9085
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.990552729003635,
+      "learning_rate": 9.848452750828566e-06,
+      "loss": 1.1574,
+      "step": 9090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.988614664868656,
+      "learning_rate": 9.848052020180989e-06,
+      "loss": 1.1198,
+      "step": 9095
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.631348102925648,
+      "learning_rate": 9.847650768588828e-06,
+      "loss": 1.1256,
+      "step": 9100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.998195523071896,
+      "learning_rate": 9.847248996095195e-06,
+      "loss": 1.1529,
+      "step": 9105
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.971478396242029,
+      "learning_rate": 9.846846702743266e-06,
+      "loss": 1.1416,
+      "step": 9110
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.8668717586872745,
+      "learning_rate": 9.84644388857627e-06,
+      "loss": 1.1337,
+      "step": 9115
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.713458049355822,
+      "learning_rate": 9.846040553637488e-06,
+      "loss": 1.0957,
+      "step": 9120
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.565561738164757,
+      "learning_rate": 9.84563669797026e-06,
+      "loss": 1.1209,
+      "step": 9125
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.321301054180532,
+      "learning_rate": 9.845232321617984e-06,
+      "loss": 1.1173,
+      "step": 9130
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.966828282551532,
+      "learning_rate": 9.844827424624113e-06,
+      "loss": 1.1384,
+      "step": 9135
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.342561458763036,
+      "learning_rate": 9.844422007032151e-06,
+      "loss": 1.1627,
+      "step": 9140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.1135760028307615,
+      "learning_rate": 9.844016068885667e-06,
+      "loss": 1.1183,
+      "step": 9145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.567234212670838,
+      "learning_rate": 9.843609610228275e-06,
+      "loss": 1.0926,
+      "step": 9150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.844832749172317,
+      "learning_rate": 9.843202631103657e-06,
+      "loss": 1.101,
+      "step": 9155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.7398542540443,
+      "learning_rate": 9.84279513155554e-06,
+      "loss": 1.0889,
+      "step": 9160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.016156765500929,
+      "learning_rate": 9.842387111627713e-06,
+      "loss": 1.1259,
+      "step": 9165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.424652351047815,
+      "learning_rate": 9.84197857136402e-06,
+      "loss": 1.1377,
+      "step": 9170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.964914639726848,
+      "learning_rate": 9.84156951080836e-06,
+      "loss": 1.1055,
+      "step": 9175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.336882081168872,
+      "learning_rate": 9.84115993000469e-06,
+      "loss": 1.1356,
+      "step": 9180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.574342064183927,
+      "learning_rate": 9.840749828997017e-06,
+      "loss": 1.1032,
+      "step": 9185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.71087525020578,
+      "learning_rate": 9.840339207829413e-06,
+      "loss": 1.1026,
+      "step": 9190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.347434640965776,
+      "learning_rate": 9.839928066545999e-06,
+      "loss": 1.1359,
+      "step": 9195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.793520773613837,
+      "learning_rate": 9.839516405190955e-06,
+      "loss": 1.1373,
+      "step": 9200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.815179199001182,
+      "learning_rate": 9.839104223808512e-06,
+      "loss": 1.135,
+      "step": 9205
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.861131066719821,
+      "learning_rate": 9.838691522442964e-06,
+      "loss": 1.1211,
+      "step": 9210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.0619283350116895,
+      "learning_rate": 9.83827830113866e-06,
+      "loss": 1.1362,
+      "step": 9215
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.937969609258512,
+      "learning_rate": 9.837864559939998e-06,
+      "loss": 1.1229,
+      "step": 9220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.942605201843459,
+      "learning_rate": 9.837450298891437e-06,
+      "loss": 1.0998,
+      "step": 9225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.948647943576111,
+      "learning_rate": 9.837035518037495e-06,
+      "loss": 1.1823,
+      "step": 9230
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.964748069626795,
+      "learning_rate": 9.836620217422735e-06,
+      "loss": 1.0749,
+      "step": 9235
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.750341701906731,
+      "learning_rate": 9.83620439709179e-06,
+      "loss": 1.1699,
+      "step": 9240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.70908142575643,
+      "learning_rate": 9.835788057089337e-06,
+      "loss": 1.1368,
+      "step": 9245
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 30.98794907794137,
+      "learning_rate": 9.835371197460117e-06,
+      "loss": 1.11,
+      "step": 9250
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.88380382654531,
+      "learning_rate": 9.83495381824892e-06,
+      "loss": 1.1427,
+      "step": 9255
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.790206155729912,
+      "learning_rate": 9.834535919500598e-06,
+      "loss": 1.1571,
+      "step": 9260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.644767137946909,
+      "learning_rate": 9.834117501260054e-06,
+      "loss": 1.1426,
+      "step": 9265
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.418037283719267,
+      "learning_rate": 9.833698563572252e-06,
+      "loss": 1.1347,
+      "step": 9270
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 20.399167830787746,
+      "learning_rate": 9.833279106482205e-06,
+      "loss": 1.0875,
+      "step": 9275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.52690583505267,
+      "learning_rate": 9.832859130034987e-06,
+      "loss": 1.1515,
+      "step": 9280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.841054786410181,
+      "learning_rate": 9.832438634275726e-06,
+      "loss": 1.0755,
+      "step": 9285
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.689457248869871,
+      "learning_rate": 9.83201761924961e-06,
+      "loss": 1.1238,
+      "step": 9290
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.5688380761419305,
+      "learning_rate": 9.83159608500187e-06,
+      "loss": 1.0593,
+      "step": 9295
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.710014392324094,
+      "learning_rate": 9.831174031577812e-06,
+      "loss": 1.1344,
+      "step": 9300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.699248344298323,
+      "learning_rate": 9.83075145902278e-06,
+      "loss": 1.119,
+      "step": 9305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.59237732027422,
+      "learning_rate": 9.830328367382185e-06,
+      "loss": 1.0859,
+      "step": 9310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 18.493955902375863,
+      "learning_rate": 9.82990475670149e-06,
+      "loss": 1.1008,
+      "step": 9315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.571250308510237,
+      "learning_rate": 9.82948062702621e-06,
+      "loss": 1.1307,
+      "step": 9320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.61127247389848,
+      "learning_rate": 9.829055978401925e-06,
+      "loss": 1.1027,
+      "step": 9325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.779161856424208,
+      "learning_rate": 9.828630810874262e-06,
+      "loss": 1.0758,
+      "step": 9330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.012146361952567,
+      "learning_rate": 9.828205124488909e-06,
+      "loss": 1.1016,
+      "step": 9335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.38028933904781,
+      "learning_rate": 9.827778919291606e-06,
+      "loss": 1.181,
+      "step": 9340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.999416557246386,
+      "learning_rate": 9.827352195328151e-06,
+      "loss": 1.1356,
+      "step": 9345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.390646370920223,
+      "learning_rate": 9.8269249526444e-06,
+      "loss": 1.1025,
+      "step": 9350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.002946454465762,
+      "learning_rate": 9.826497191286257e-06,
+      "loss": 1.1567,
+      "step": 9355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.933892786820671,
+      "learning_rate": 9.826068911299692e-06,
+      "loss": 1.1228,
+      "step": 9360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.402605896074404,
+      "learning_rate": 9.825640112730723e-06,
+      "loss": 1.1722,
+      "step": 9365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.073746902114985,
+      "learning_rate": 9.825210795625426e-06,
+      "loss": 1.1494,
+      "step": 9370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.505770539857927,
+      "learning_rate": 9.824780960029934e-06,
+      "loss": 1.1371,
+      "step": 9375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.13806849938393,
+      "learning_rate": 9.824350605990434e-06,
+      "loss": 1.1668,
+      "step": 9380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.390959115620975,
+      "learning_rate": 9.823919733553171e-06,
+      "loss": 1.0758,
+      "step": 9385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.114862692148765,
+      "learning_rate": 9.823488342764442e-06,
+      "loss": 1.0858,
+      "step": 9390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.306169585772543,
+      "learning_rate": 9.823056433670603e-06,
+      "loss": 1.1119,
+      "step": 9395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.935150769462741,
+      "learning_rate": 9.822624006318065e-06,
+      "loss": 1.0986,
+      "step": 9400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.6496393432298495,
+      "learning_rate": 9.822191060753295e-06,
+      "loss": 1.1087,
+      "step": 9405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.943038279068391,
+      "learning_rate": 9.821757597022811e-06,
+      "loss": 1.121,
+      "step": 9410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.326836780816448,
+      "learning_rate": 9.821323615173195e-06,
+      "loss": 1.101,
+      "step": 9415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.551718225947079,
+      "learning_rate": 9.820889115251078e-06,
+      "loss": 1.0919,
+      "step": 9420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.127512681130131,
+      "learning_rate": 9.820454097303148e-06,
+      "loss": 1.132,
+      "step": 9425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.432004802350892,
+      "learning_rate": 9.820018561376153e-06,
+      "loss": 1.1897,
+      "step": 9430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.049844572848882,
+      "learning_rate": 9.81958250751689e-06,
+      "loss": 1.0832,
+      "step": 9435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.318701917149032,
+      "learning_rate": 9.819145935772217e-06,
+      "loss": 1.1318,
+      "step": 9440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.985454154216166,
+      "learning_rate": 9.818708846189041e-06,
+      "loss": 1.098,
+      "step": 9445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.19872846134024,
+      "learning_rate": 9.818271238814339e-06,
+      "loss": 1.1313,
+      "step": 9450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.521411137478754,
+      "learning_rate": 9.817833113695123e-06,
+      "loss": 1.132,
+      "step": 9455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.820409261480718,
+      "learning_rate": 9.817394470878479e-06,
+      "loss": 1.1095,
+      "step": 9460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.5678130983731124,
+      "learning_rate": 9.816955310411537e-06,
+      "loss": 1.0875,
+      "step": 9465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.667916390733254,
+      "learning_rate": 9.816515632341487e-06,
+      "loss": 1.1544,
+      "step": 9470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.102140788856543,
+      "learning_rate": 9.816075436715577e-06,
+      "loss": 1.1282,
+      "step": 9475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.432976660373403,
+      "learning_rate": 9.815634723581106e-06,
+      "loss": 1.0735,
+      "step": 9480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.8805646616392675,
+      "learning_rate": 9.815193492985432e-06,
+      "loss": 1.1147,
+      "step": 9485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.43628359747152,
+      "learning_rate": 9.814751744975964e-06,
+      "loss": 1.1482,
+      "step": 9490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.433577785768476,
+      "learning_rate": 9.814309479600173e-06,
+      "loss": 1.1606,
+      "step": 9495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.866263071903223,
+      "learning_rate": 9.813866696905581e-06,
+      "loss": 1.1712,
+      "step": 9500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 49.551941694609866,
+      "learning_rate": 9.813423396939767e-06,
+      "loss": 1.1682,
+      "step": 9505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.95562468305858,
+      "learning_rate": 9.812979579750367e-06,
+      "loss": 1.1726,
+      "step": 9510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.918281155827465,
+      "learning_rate": 9.812535245385069e-06,
+      "loss": 1.1331,
+      "step": 9515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.239698737640707,
+      "learning_rate": 9.81209039389162e-06,
+      "loss": 1.1467,
+      "step": 9520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.542450852206898,
+      "learning_rate": 9.811645025317819e-06,
+      "loss": 1.0984,
+      "step": 9525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.839824606202814,
+      "learning_rate": 9.811199139711528e-06,
+      "loss": 1.1459,
+      "step": 9530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.062707119434517,
+      "learning_rate": 9.810752737120652e-06,
+      "loss": 1.1725,
+      "step": 9535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.920835764199387,
+      "learning_rate": 9.810305817593167e-06,
+      "loss": 1.146,
+      "step": 9540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.823441097598744,
+      "learning_rate": 9.809858381177089e-06,
+      "loss": 1.1867,
+      "step": 9545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 18.483353058092547,
+      "learning_rate": 9.809410427920501e-06,
+      "loss": 1.1255,
+      "step": 9550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.792922139008056,
+      "learning_rate": 9.808961957871539e-06,
+      "loss": 1.0821,
+      "step": 9555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.5466522897197414,
+      "learning_rate": 9.80851297107839e-06,
+      "loss": 1.153,
+      "step": 9560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.5036408297158,
+      "learning_rate": 9.8080634675893e-06,
+      "loss": 1.1436,
+      "step": 9565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 30.775175095239852,
+      "learning_rate": 9.80761344745257e-06,
+      "loss": 1.1672,
+      "step": 9570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.275504767645358,
+      "learning_rate": 9.80716291071656e-06,
+      "loss": 1.1187,
+      "step": 9575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 18.957991468277854,
+      "learning_rate": 9.806711857429678e-06,
+      "loss": 1.1397,
+      "step": 9580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.726706035900417,
+      "learning_rate": 9.806260287640393e-06,
+      "loss": 1.0869,
+      "step": 9585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 23.00409452812483,
+      "learning_rate": 9.805808201397229e-06,
+      "loss": 1.0672,
+      "step": 9590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.166416366509448,
+      "learning_rate": 9.805355598748766e-06,
+      "loss": 1.1041,
+      "step": 9595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 22.53928607518761,
+      "learning_rate": 9.804902479743632e-06,
+      "loss": 1.1162,
+      "step": 9600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 34.38349793200656,
+      "learning_rate": 9.804448844430523e-06,
+      "loss": 1.129,
+      "step": 9605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.854031830090808,
+      "learning_rate": 9.803994692858184e-06,
+      "loss": 1.1163,
+      "step": 9610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.911815034854689,
+      "learning_rate": 9.80354002507541e-06,
+      "loss": 1.1344,
+      "step": 9615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.943951719386559,
+      "learning_rate": 9.803084841131063e-06,
+      "loss": 1.0573,
+      "step": 9620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.655449294062125,
+      "learning_rate": 9.802629141074053e-06,
+      "loss": 1.1198,
+      "step": 9625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.731686298770386,
+      "learning_rate": 9.802172924953344e-06,
+      "loss": 1.1135,
+      "step": 9630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.784653923093681,
+      "learning_rate": 9.801716192817962e-06,
+      "loss": 1.1518,
+      "step": 9635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.075680462367813,
+      "learning_rate": 9.801258944716982e-06,
+      "loss": 1.1319,
+      "step": 9640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.699910823794717,
+      "learning_rate": 9.800801180699542e-06,
+      "loss": 1.1177,
+      "step": 9645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.696372845431933,
+      "learning_rate": 9.800342900814825e-06,
+      "loss": 1.1271,
+      "step": 9650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.95683377412942,
+      "learning_rate": 9.799884105112078e-06,
+      "loss": 1.199,
+      "step": 9655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.02638592169939,
+      "learning_rate": 9.7994247936406e-06,
+      "loss": 1.1163,
+      "step": 9660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.014399656225942,
+      "learning_rate": 9.798964966449746e-06,
+      "loss": 1.1542,
+      "step": 9665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.85741257340364,
+      "learning_rate": 9.798504623588929e-06,
+      "loss": 1.1081,
+      "step": 9670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 28.312484526129953,
+      "learning_rate": 9.798043765107612e-06,
+      "loss": 1.1111,
+      "step": 9675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 16.087873458378628,
+      "learning_rate": 9.797582391055317e-06,
+      "loss": 1.1902,
+      "step": 9680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 16.109718740528173,
+      "learning_rate": 9.79712050148162e-06,
+      "loss": 1.1585,
+      "step": 9685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.241416079713798,
+      "learning_rate": 9.796658096436155e-06,
+      "loss": 1.1262,
+      "step": 9690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.099802194732407,
+      "learning_rate": 9.796195175968609e-06,
+      "loss": 1.1314,
+      "step": 9695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.806323165690559,
+      "learning_rate": 9.79573174012872e-06,
+      "loss": 1.1419,
+      "step": 9700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.84510943304211,
+      "learning_rate": 9.795267788966293e-06,
+      "loss": 1.079,
+      "step": 9705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.244004409262631,
+      "learning_rate": 9.79480332253118e-06,
+      "loss": 1.1117,
+      "step": 9710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 25.94214805244186,
+      "learning_rate": 9.794338340873288e-06,
+      "loss": 1.1416,
+      "step": 9715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.56855739260311,
+      "learning_rate": 9.793872844042582e-06,
+      "loss": 1.1155,
+      "step": 9720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.1216852701226,
+      "learning_rate": 9.793406832089084e-06,
+      "loss": 1.0934,
+      "step": 9725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 21.617542471609994,
+      "learning_rate": 9.792940305062863e-06,
+      "loss": 1.1141,
+      "step": 9730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.095356050481866,
+      "learning_rate": 9.792473263014056e-06,
+      "loss": 1.1001,
+      "step": 9735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.271670284567916,
+      "learning_rate": 9.792005705992847e-06,
+      "loss": 1.0696,
+      "step": 9740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.60935982925622,
+      "learning_rate": 9.791537634049473e-06,
+      "loss": 1.1412,
+      "step": 9745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.4281297841404506,
+      "learning_rate": 9.791069047234237e-06,
+      "loss": 1.126,
+      "step": 9750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 16.9394370256431,
+      "learning_rate": 9.790599945597484e-06,
+      "loss": 1.1089,
+      "step": 9755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.121878911508034,
+      "learning_rate": 9.790130329189626e-06,
+      "loss": 1.1086,
+      "step": 9760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.022861117918076,
+      "learning_rate": 9.789660198061124e-06,
+      "loss": 1.1442,
+      "step": 9765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.53097471180735,
+      "learning_rate": 9.789189552262494e-06,
+      "loss": 1.1686,
+      "step": 9770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.115874036538647,
+      "learning_rate": 9.788718391844312e-06,
+      "loss": 1.1327,
+      "step": 9775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.10906533014789,
+      "learning_rate": 9.788246716857203e-06,
+      "loss": 1.1365,
+      "step": 9780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.257897698293505,
+      "learning_rate": 9.787774527351853e-06,
+      "loss": 1.1272,
+      "step": 9785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.987710480713961,
+      "learning_rate": 9.787301823379e-06,
+      "loss": 1.1448,
+      "step": 9790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.36636842861665,
+      "learning_rate": 9.786828604989438e-06,
+      "loss": 1.1113,
+      "step": 9795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.169325055729368,
+      "learning_rate": 9.786354872234016e-06,
+      "loss": 1.1119,
+      "step": 9800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.567259582207006,
+      "learning_rate": 9.78588062516364e-06,
+      "loss": 1.0768,
+      "step": 9805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.523502957031439,
+      "learning_rate": 9.78540586382927e-06,
+      "loss": 1.0812,
+      "step": 9810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.818214437373952,
+      "learning_rate": 9.784930588281918e-06,
+      "loss": 1.1467,
+      "step": 9815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.5736218420296995,
+      "learning_rate": 9.784454798572658e-06,
+      "loss": 1.1653,
+      "step": 9820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.68106601324899,
+      "learning_rate": 9.783978494752613e-06,
+      "loss": 1.0969,
+      "step": 9825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.622379031185475,
+      "learning_rate": 9.783501676872968e-06,
+      "loss": 1.1296,
+      "step": 9830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.75996294680562,
+      "learning_rate": 9.783024344984956e-06,
+      "loss": 1.0826,
+      "step": 9835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.334303857518385,
+      "learning_rate": 9.782546499139867e-06,
+      "loss": 1.1397,
+      "step": 9840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.214488195724635,
+      "learning_rate": 9.782068139389051e-06,
+      "loss": 1.1237,
+      "step": 9845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.796337377225278,
+      "learning_rate": 9.781589265783909e-06,
+      "loss": 1.0843,
+      "step": 9850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.332913308661327,
+      "learning_rate": 9.781109878375897e-06,
+      "loss": 1.089,
+      "step": 9855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.006643164952569,
+      "learning_rate": 9.78062997721653e-06,
+      "loss": 1.1162,
+      "step": 9860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.827538006070318,
+      "learning_rate": 9.78014956235737e-06,
+      "loss": 1.0892,
+      "step": 9865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.829102172200375,
+      "learning_rate": 9.779668633850046e-06,
+      "loss": 1.0721,
+      "step": 9870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.863495499204848,
+      "learning_rate": 9.779187191746232e-06,
+      "loss": 1.0963,
+      "step": 9875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.755519011212401,
+      "learning_rate": 9.778705236097661e-06,
+      "loss": 1.1033,
+      "step": 9880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.485676461703726,
+      "learning_rate": 9.778222766956122e-06,
+      "loss": 1.1395,
+      "step": 9885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.514552674263018,
+      "learning_rate": 9.777739784373461e-06,
+      "loss": 1.1078,
+      "step": 9890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.024209150490186,
+      "learning_rate": 9.777256288401573e-06,
+      "loss": 1.1041,
+      "step": 9895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.417370515958933,
+      "learning_rate": 9.776772279092414e-06,
+      "loss": 1.1283,
+      "step": 9900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 14.681527551512538,
+      "learning_rate": 9.776287756497991e-06,
+      "loss": 1.0713,
+      "step": 9905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.332222561837927,
+      "learning_rate": 9.77580272067037e-06,
+      "loss": 1.1034,
+      "step": 9910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.079960594509294,
+      "learning_rate": 9.775317171661668e-06,
+      "loss": 1.0918,
+      "step": 9915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 18.95619231372345,
+      "learning_rate": 9.774831109524064e-06,
+      "loss": 1.1158,
+      "step": 9920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.754390005346306,
+      "learning_rate": 9.774344534309782e-06,
+      "loss": 1.0926,
+      "step": 9925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.839537890368287,
+      "learning_rate": 9.773857446071108e-06,
+      "loss": 1.1181,
+      "step": 9930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.620360785136541,
+      "learning_rate": 9.773369844860384e-06,
+      "loss": 1.127,
+      "step": 9935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.252806013880718,
+      "learning_rate": 9.772881730730007e-06,
+      "loss": 1.1544,
+      "step": 9940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.308373503259316,
+      "learning_rate": 9.772393103732417e-06,
+      "loss": 1.1641,
+      "step": 9945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.50809360555174,
+      "learning_rate": 9.771903963920132e-06,
+      "loss": 1.1238,
+      "step": 9950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.510544230636716,
+      "learning_rate": 9.771414311345702e-06,
+      "loss": 1.1382,
+      "step": 9955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.4853102905021505,
+      "learning_rate": 9.770924146061749e-06,
+      "loss": 1.15,
+      "step": 9960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.086961941226198,
+      "learning_rate": 9.77043346812094e-06,
+      "loss": 1.1442,
+      "step": 9965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.514056476883807,
+      "learning_rate": 9.769942277576001e-06,
+      "loss": 1.1134,
+      "step": 9970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.974031982611894,
+      "learning_rate": 9.769450574479714e-06,
+      "loss": 1.0835,
+      "step": 9975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 30.281983181732873,
+      "learning_rate": 9.76895835888491e-06,
+      "loss": 1.0753,
+      "step": 9980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.572762073580233,
+      "learning_rate": 9.768465630844487e-06,
+      "loss": 1.1356,
+      "step": 9985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 15.469596086217209,
+      "learning_rate": 9.767972390411386e-06,
+      "loss": 1.1499,
+      "step": 9990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.324454244157174,
+      "learning_rate": 9.767478637638609e-06,
+      "loss": 1.1274,
+      "step": 9995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 15.751386696752718,
+      "learning_rate": 9.766984372579212e-06,
+      "loss": 1.1674,
+      "step": 10000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.757330861218081,
+      "learning_rate": 9.766489595286306e-06,
+      "loss": 1.1062,
+      "step": 10005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.577967008442595,
+      "learning_rate": 9.765994305813056e-06,
+      "loss": 1.1584,
+      "step": 10010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.715385504908574,
+      "learning_rate": 9.765498504212685e-06,
+      "loss": 1.1571,
+      "step": 10015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.188995733588726,
+      "learning_rate": 9.765002190538467e-06,
+      "loss": 1.1157,
+      "step": 10020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.937873724368163,
+      "learning_rate": 9.764505364843734e-06,
+      "loss": 1.1288,
+      "step": 10025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 19.581915732852572,
+      "learning_rate": 9.764008027181872e-06,
+      "loss": 1.1096,
+      "step": 10030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 35.12020185934868,
+      "learning_rate": 9.763510177606324e-06,
+      "loss": 1.1504,
+      "step": 10035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 27.79462408084681,
+      "learning_rate": 9.763011816170583e-06,
+      "loss": 1.147,
+      "step": 10040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.135063625409902,
+      "learning_rate": 9.762512942928201e-06,
+      "loss": 1.1711,
+      "step": 10045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.7427326896889035,
+      "learning_rate": 9.762013557932787e-06,
+      "loss": 1.1437,
+      "step": 10050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 62.042302136295355,
+      "learning_rate": 9.761513661237998e-06,
+      "loss": 1.1346,
+      "step": 10055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.482157923920138,
+      "learning_rate": 9.761013252897554e-06,
+      "loss": 1.1313,
+      "step": 10060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.77522261793372,
+      "learning_rate": 9.760512332965221e-06,
+      "loss": 1.0506,
+      "step": 10065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.82416686579001,
+      "learning_rate": 9.76001090149483e-06,
+      "loss": 1.0977,
+      "step": 10070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.98131908318692,
+      "learning_rate": 9.75950895854026e-06,
+      "loss": 1.1557,
+      "step": 10075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 16.822062970124897,
+      "learning_rate": 9.759006504155446e-06,
+      "loss": 1.1455,
+      "step": 10080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.864234312537614,
+      "learning_rate": 9.758503538394382e-06,
+      "loss": 1.1413,
+      "step": 10085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.079842809205594,
+      "learning_rate": 9.75800006131111e-06,
+      "loss": 1.1507,
+      "step": 10090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.107264063664683,
+      "learning_rate": 9.757496072959734e-06,
+      "loss": 1.089,
+      "step": 10095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.195801576453392,
+      "learning_rate": 9.756991573394407e-06,
+      "loss": 1.1802,
+      "step": 10100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.779377619654823,
+      "learning_rate": 9.756486562669342e-06,
+      "loss": 1.1259,
+      "step": 10105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.386319012783584,
+      "learning_rate": 9.755981040838805e-06,
+      "loss": 1.0957,
+      "step": 10110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.811523355585202,
+      "learning_rate": 9.755475007957114e-06,
+      "loss": 1.0775,
+      "step": 10115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.369045677901603,
+      "learning_rate": 9.754968464078647e-06,
+      "loss": 1.0964,
+      "step": 10120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.597199366258872,
+      "learning_rate": 9.754461409257832e-06,
+      "loss": 1.1203,
+      "step": 10125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.471153307129219,
+      "learning_rate": 9.753953843549157e-06,
+      "loss": 1.0989,
+      "step": 10130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.875140466143439,
+      "learning_rate": 9.753445767007158e-06,
+      "loss": 1.1246,
+      "step": 10135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 20.75088919991107,
+      "learning_rate": 9.752937179686438e-06,
+      "loss": 1.1419,
+      "step": 10140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 17.543916367697925,
+      "learning_rate": 9.752428081641636e-06,
+      "loss": 1.1393,
+      "step": 10145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.194304111540609,
+      "learning_rate": 9.751918472927466e-06,
+      "loss": 1.1308,
+      "step": 10150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.916008625280372,
+      "learning_rate": 9.751408353598684e-06,
+      "loss": 1.0633,
+      "step": 10155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.8423195861966875,
+      "learning_rate": 9.750897723710104e-06,
+      "loss": 1.1195,
+      "step": 10160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.902355377355023,
+      "learning_rate": 9.750386583316595e-06,
+      "loss": 1.0883,
+      "step": 10165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.081120340072049,
+      "learning_rate": 9.749874932473085e-06,
+      "loss": 1.1056,
+      "step": 10170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.1463943403080865,
+      "learning_rate": 9.74936277123455e-06,
+      "loss": 1.1277,
+      "step": 10175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.747013479713555,
+      "learning_rate": 9.748850099656024e-06,
+      "loss": 1.1424,
+      "step": 10180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.854675542442633,
+      "learning_rate": 9.748336917792595e-06,
+      "loss": 1.091,
+      "step": 10185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.580645624406868,
+      "learning_rate": 9.747823225699411e-06,
+      "loss": 1.1058,
+      "step": 10190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.97250071910222,
+      "learning_rate": 9.747309023431668e-06,
+      "loss": 1.0889,
+      "step": 10195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 25.806132412860915,
+      "learning_rate": 9.746794311044616e-06,
+      "loss": 1.1003,
+      "step": 10200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.945519407440568,
+      "learning_rate": 9.746279088593569e-06,
+      "loss": 1.147,
+      "step": 10205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.64393437502863,
+      "learning_rate": 9.745763356133884e-06,
+      "loss": 1.0804,
+      "step": 10210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.292462338414721,
+      "learning_rate": 9.745247113720985e-06,
+      "loss": 1.1553,
+      "step": 10215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.827026435628405,
+      "learning_rate": 9.74473036141034e-06,
+      "loss": 1.1515,
+      "step": 10220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.04797847257774,
+      "learning_rate": 9.744213099257477e-06,
+      "loss": 1.1039,
+      "step": 10225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 21.339149586266636,
+      "learning_rate": 9.743695327317979e-06,
+      "loss": 1.1227,
+      "step": 10230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.996841412197717,
+      "learning_rate": 9.743177045647482e-06,
+      "loss": 1.1313,
+      "step": 10235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.945502506850865,
+      "learning_rate": 9.742658254301679e-06,
+      "loss": 1.1399,
+      "step": 10240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.075472790928725,
+      "learning_rate": 9.742138953336316e-06,
+      "loss": 1.1114,
+      "step": 10245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 24.094149374045454,
+      "learning_rate": 9.741619142807192e-06,
+      "loss": 1.1586,
+      "step": 10250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.250546938046734,
+      "learning_rate": 9.741098822770166e-06,
+      "loss": 1.0988,
+      "step": 10255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.897431171273826,
+      "learning_rate": 9.740577993281148e-06,
+      "loss": 1.0643,
+      "step": 10260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.013004346619281,
+      "learning_rate": 9.740056654396101e-06,
+      "loss": 1.1582,
+      "step": 10265
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.888002368756538,
+      "learning_rate": 9.73953480617105e-06,
+      "loss": 1.1335,
+      "step": 10270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.43670125601399,
+      "learning_rate": 9.739012448662064e-06,
+      "loss": 1.1315,
+      "step": 10275
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.826848129313802,
+      "learning_rate": 9.738489581925277e-06,
+      "loss": 1.1661,
+      "step": 10280
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 24.349128088776393,
+      "learning_rate": 9.73796620601687e-06,
+      "loss": 1.1004,
+      "step": 10285
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.10533008169947,
+      "learning_rate": 9.737442320993087e-06,
+      "loss": 1.1286,
+      "step": 10290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.222686842711923,
+      "learning_rate": 9.736917926910217e-06,
+      "loss": 1.1114,
+      "step": 10295
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.196280282450498,
+      "learning_rate": 9.73639302382461e-06,
+      "loss": 1.0839,
+      "step": 10300
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.100321322337622,
+      "learning_rate": 9.73586761179267e-06,
+      "loss": 1.1142,
+      "step": 10305
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.209475139679098,
+      "learning_rate": 9.735341690870855e-06,
+      "loss": 1.0887,
+      "step": 10310
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.444327449895846,
+      "learning_rate": 9.734815261115675e-06,
+      "loss": 1.1318,
+      "step": 10315
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.827873819975245,
+      "learning_rate": 9.734288322583699e-06,
+      "loss": 1.1374,
+      "step": 10320
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 14.783411255982069,
+      "learning_rate": 9.733760875331548e-06,
+      "loss": 1.1242,
+      "step": 10325
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.493691481525475,
+      "learning_rate": 9.7332329194159e-06,
+      "loss": 1.1307,
+      "step": 10330
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.581919740015162,
+      "learning_rate": 9.732704454893486e-06,
+      "loss": 1.1099,
+      "step": 10335
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.715058082145536,
+      "learning_rate": 9.732175481821092e-06,
+      "loss": 1.1331,
+      "step": 10340
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.192343668844734,
+      "learning_rate": 9.731646000255558e-06,
+      "loss": 1.1448,
+      "step": 10345
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.679577190945963,
+      "learning_rate": 9.731116010253775e-06,
+      "loss": 1.1377,
+      "step": 10350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.573944570129123,
+      "learning_rate": 9.730585511872702e-06,
+      "loss": 1.1072,
+      "step": 10355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 40.535612120264155,
+      "learning_rate": 9.730054505169334e-06,
+      "loss": 1.1553,
+      "step": 10360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 28.689454298470146,
+      "learning_rate": 9.729522990200735e-06,
+      "loss": 1.1225,
+      "step": 10365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.917519588416386,
+      "learning_rate": 9.72899096702402e-06,
+      "loss": 1.1283,
+      "step": 10370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 24.286237193416884,
+      "learning_rate": 9.728458435696352e-06,
+      "loss": 1.164,
+      "step": 10375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 55.07682352740546,
+      "learning_rate": 9.727925396274958e-06,
+      "loss": 1.1373,
+      "step": 10380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 20.71368053216417,
+      "learning_rate": 9.727391848817114e-06,
+      "loss": 1.1441,
+      "step": 10385
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.705263989766612,
+      "learning_rate": 9.726857793380153e-06,
+      "loss": 1.1164,
+      "step": 10390
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.05921596663414,
+      "learning_rate": 9.72632323002146e-06,
+      "loss": 1.1428,
+      "step": 10395
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.16323932261975,
+      "learning_rate": 9.725788158798477e-06,
+      "loss": 1.0978,
+      "step": 10400
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.777814903551782,
+      "learning_rate": 9.725252579768701e-06,
+      "loss": 1.1299,
+      "step": 10405
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.197498292246134,
+      "learning_rate": 9.724716492989681e-06,
+      "loss": 1.144,
+      "step": 10410
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 19.40569514490288,
+      "learning_rate": 9.724179898519022e-06,
+      "loss": 1.1844,
+      "step": 10415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.509725936530929,
+      "learning_rate": 9.723642796414383e-06,
+      "loss": 1.0931,
+      "step": 10420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.606846928753214,
+      "learning_rate": 9.723105186733479e-06,
+      "loss": 1.11,
+      "step": 10425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.221554586179171,
+      "learning_rate": 9.722567069534077e-06,
+      "loss": 1.1516,
+      "step": 10430
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.220405910858158,
+      "learning_rate": 9.722028444874002e-06,
+      "loss": 1.0947,
+      "step": 10435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.691082528659585,
+      "learning_rate": 9.721489312811129e-06,
+      "loss": 1.1555,
+      "step": 10440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.400540137386916,
+      "learning_rate": 9.720949673403395e-06,
+      "loss": 1.0979,
+      "step": 10445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.724417889986143,
+      "learning_rate": 9.720409526708782e-06,
+      "loss": 1.0846,
+      "step": 10450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.604818406860277,
+      "learning_rate": 9.719868872785331e-06,
+      "loss": 1.1206,
+      "step": 10455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.495729299942727,
+      "learning_rate": 9.719327711691142e-06,
+      "loss": 1.0591,
+      "step": 10460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.494537501359319,
+      "learning_rate": 9.71878604348436e-06,
+      "loss": 1.0871,
+      "step": 10465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.284017470531094,
+      "learning_rate": 9.718243868223193e-06,
+      "loss": 1.1264,
+      "step": 10470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.005211207587475,
+      "learning_rate": 9.7177011859659e-06,
+      "loss": 1.0903,
+      "step": 10475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.188547912156826,
+      "learning_rate": 9.717157996770793e-06,
+      "loss": 1.1578,
+      "step": 10480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 24.351181207494125,
+      "learning_rate": 9.716614300696242e-06,
+      "loss": 1.1269,
+      "step": 10485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.444515779435607,
+      "learning_rate": 9.716070097800665e-06,
+      "loss": 1.1311,
+      "step": 10490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 25.85354651689487,
+      "learning_rate": 9.715525388142545e-06,
+      "loss": 1.109,
+      "step": 10495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.276499819770264,
+      "learning_rate": 9.71498017178041e-06,
+      "loss": 1.1172,
+      "step": 10500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.75130989314831,
+      "learning_rate": 9.714434448772847e-06,
+      "loss": 1.107,
+      "step": 10505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.5014359906003,
+      "learning_rate": 9.713888219178497e-06,
+      "loss": 1.1076,
+      "step": 10510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 19.842848512386926,
+      "learning_rate": 9.713341483056054e-06,
+      "loss": 1.1029,
+      "step": 10515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.135521788293767,
+      "learning_rate": 9.712794240464265e-06,
+      "loss": 1.1221,
+      "step": 10520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.428356454329931,
+      "learning_rate": 9.712246491461937e-06,
+      "loss": 1.15,
+      "step": 10525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 18.754908177493192,
+      "learning_rate": 9.711698236107927e-06,
+      "loss": 1.1815,
+      "step": 10530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 30.93622654744006,
+      "learning_rate": 9.711149474461148e-06,
+      "loss": 1.151,
+      "step": 10535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.04329072521145,
+      "learning_rate": 9.710600206580565e-06,
+      "loss": 1.1517,
+      "step": 10540
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 29.8315475145888,
+      "learning_rate": 9.7100504325252e-06,
+      "loss": 1.1409,
+      "step": 10545
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.004271179187006,
+      "learning_rate": 9.709500152354128e-06,
+      "loss": 1.1159,
+      "step": 10550
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 20.74101271516686,
+      "learning_rate": 9.708949366126482e-06,
+      "loss": 1.1402,
+      "step": 10555
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.55534773566369,
+      "learning_rate": 9.708398073901444e-06,
+      "loss": 1.1007,
+      "step": 10560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 17.54004758170369,
+      "learning_rate": 9.707846275738253e-06,
+      "loss": 1.1298,
+      "step": 10565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.400850159105255,
+      "learning_rate": 9.707293971696202e-06,
+      "loss": 1.0853,
+      "step": 10570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 32.876751163781954,
+      "learning_rate": 9.70674116183464e-06,
+      "loss": 1.1536,
+      "step": 10575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 22.089008325041274,
+      "learning_rate": 9.706187846212967e-06,
+      "loss": 1.0851,
+      "step": 10580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 24.455632146508183,
+      "learning_rate": 9.70563402489064e-06,
+      "loss": 1.1009,
+      "step": 10585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 20.348149765166458,
+      "learning_rate": 9.705079697927168e-06,
+      "loss": 1.1786,
+      "step": 10590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.071164665487567,
+      "learning_rate": 9.70452486538212e-06,
+      "loss": 1.1392,
+      "step": 10595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.021139837990567,
+      "learning_rate": 9.70396952731511e-06,
+      "loss": 1.1434,
+      "step": 10600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 25.640077434092976,
+      "learning_rate": 9.703413683785817e-06,
+      "loss": 1.174,
+      "step": 10605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 17.417474433355128,
+      "learning_rate": 9.702857334853963e-06,
+      "loss": 1.1219,
+      "step": 10610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 22.32809964297896,
+      "learning_rate": 9.702300480579334e-06,
+      "loss": 1.1181,
+      "step": 10615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.940608868765343,
+      "learning_rate": 9.701743121021767e-06,
+      "loss": 1.1203,
+      "step": 10620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.463064479304906,
+      "learning_rate": 9.70118525624115e-06,
+      "loss": 1.1328,
+      "step": 10625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.441182793738268,
+      "learning_rate": 9.70062688629743e-06,
+      "loss": 1.1172,
+      "step": 10630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.042403549492438,
+      "learning_rate": 9.700068011250605e-06,
+      "loss": 1.1447,
+      "step": 10635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.664854623155703,
+      "learning_rate": 9.699508631160728e-06,
+      "loss": 1.0961,
+      "step": 10640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.71255557018499,
+      "learning_rate": 9.69894874608791e-06,
+      "loss": 1.0962,
+      "step": 10645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.352924314417328,
+      "learning_rate": 9.698388356092311e-06,
+      "loss": 1.1042,
+      "step": 10650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.5339482523967956,
+      "learning_rate": 9.697827461234147e-06,
+      "loss": 1.1122,
+      "step": 10655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.970137140377447,
+      "learning_rate": 9.69726606157369e-06,
+      "loss": 1.0622,
+      "step": 10660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.677722904681502,
+      "learning_rate": 9.696704157171262e-06,
+      "loss": 1.1195,
+      "step": 10665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.48068164376845,
+      "learning_rate": 9.696141748087247e-06,
+      "loss": 1.0789,
+      "step": 10670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.094365418738656,
+      "learning_rate": 9.695578834382074e-06,
+      "loss": 1.1123,
+      "step": 10675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.725765410505641,
+      "learning_rate": 9.695015416116232e-06,
+      "loss": 1.0964,
+      "step": 10680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.571844782783497,
+      "learning_rate": 9.694451493350264e-06,
+      "loss": 1.1276,
+      "step": 10685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.570916415682218,
+      "learning_rate": 9.693887066144762e-06,
+      "loss": 1.0941,
+      "step": 10690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.616549306214913,
+      "learning_rate": 9.693322134560382e-06,
+      "loss": 1.1304,
+      "step": 10695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.2412219935147695,
+      "learning_rate": 9.692756698657824e-06,
+      "loss": 1.1628,
+      "step": 10700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.592995785583279,
+      "learning_rate": 9.692190758497848e-06,
+      "loss": 1.1511,
+      "step": 10705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.956380223647882,
+      "learning_rate": 9.691624314141269e-06,
+      "loss": 1.1217,
+      "step": 10710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.601176942316527,
+      "learning_rate": 9.691057365648948e-06,
+      "loss": 1.0889,
+      "step": 10715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.289383284440312,
+      "learning_rate": 9.690489913081811e-06,
+      "loss": 1.0935,
+      "step": 10720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.769763701506643,
+      "learning_rate": 9.689921956500833e-06,
+      "loss": 1.141,
+      "step": 10725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.03707990084338,
+      "learning_rate": 9.689353495967043e-06,
+      "loss": 1.1382,
+      "step": 10730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.154239808856511,
+      "learning_rate": 9.688784531541524e-06,
+      "loss": 1.0971,
+      "step": 10735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.68709279393053,
+      "learning_rate": 9.688215063285414e-06,
+      "loss": 1.0991,
+      "step": 10740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.410367886881257,
+      "learning_rate": 9.687645091259902e-06,
+      "loss": 1.1552,
+      "step": 10745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.982096164450052,
+      "learning_rate": 9.68707461552624e-06,
+      "loss": 1.093,
+      "step": 10750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.726619292438254,
+      "learning_rate": 9.686503636145724e-06,
+      "loss": 1.1491,
+      "step": 10755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.841557450606214,
+      "learning_rate": 9.685932153179709e-06,
+      "loss": 1.0842,
+      "step": 10760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.599943842601997,
+      "learning_rate": 9.685360166689604e-06,
+      "loss": 1.1019,
+      "step": 10765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.1858095951090615,
+      "learning_rate": 9.68478767673687e-06,
+      "loss": 1.1489,
+      "step": 10770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.360875052792865,
+      "learning_rate": 9.684214683383027e-06,
+      "loss": 1.1493,
+      "step": 10775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.082698170941965,
+      "learning_rate": 9.683641186689642e-06,
+      "loss": 1.1402,
+      "step": 10780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.030738278295825,
+      "learning_rate": 9.68306718671834e-06,
+      "loss": 1.1297,
+      "step": 10785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.431781035672577,
+      "learning_rate": 9.682492683530802e-06,
+      "loss": 1.1744,
+      "step": 10790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.512006836021577,
+      "learning_rate": 9.681917677188758e-06,
+      "loss": 1.1445,
+      "step": 10795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.940268972104937,
+      "learning_rate": 9.681342167753998e-06,
+      "loss": 1.0945,
+      "step": 10800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 29.290167157837768,
+      "learning_rate": 9.680766155288362e-06,
+      "loss": 1.1571,
+      "step": 10805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.498490605308364,
+      "learning_rate": 9.680189639853743e-06,
+      "loss": 1.1203,
+      "step": 10810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.251821035868401,
+      "learning_rate": 9.679612621512093e-06,
+      "loss": 1.1003,
+      "step": 10815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.45276028054799,
+      "learning_rate": 9.679035100325416e-06,
+      "loss": 1.0907,
+      "step": 10820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.69058800986446,
+      "learning_rate": 9.678457076355764e-06,
+      "loss": 1.1457,
+      "step": 10825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.27444012265364,
+      "learning_rate": 9.677878549665254e-06,
+      "loss": 1.1632,
+      "step": 10830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.006361526713334,
+      "learning_rate": 9.677299520316048e-06,
+      "loss": 1.1473,
+      "step": 10835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.301868180611034,
+      "learning_rate": 9.676719988370366e-06,
+      "loss": 1.1146,
+      "step": 10840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 21.612832280318436,
+      "learning_rate": 9.676139953890482e-06,
+      "loss": 1.1083,
+      "step": 10845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.914221632107225,
+      "learning_rate": 9.675559416938723e-06,
+      "loss": 1.1085,
+      "step": 10850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.31008654260704,
+      "learning_rate": 9.674978377577468e-06,
+      "loss": 1.1384,
+      "step": 10855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.975980953182093,
+      "learning_rate": 9.674396835869155e-06,
+      "loss": 1.1205,
+      "step": 10860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 19.910273772561663,
+      "learning_rate": 9.673814791876273e-06,
+      "loss": 1.1497,
+      "step": 10865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 27.86541853928646,
+      "learning_rate": 9.673232245661364e-06,
+      "loss": 1.148,
+      "step": 10870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 37.35132208073142,
+      "learning_rate": 9.672649197287025e-06,
+      "loss": 1.1147,
+      "step": 10875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 38.39023156929308,
+      "learning_rate": 9.672065646815906e-06,
+      "loss": 1.1558,
+      "step": 10880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.555144351333318,
+      "learning_rate": 9.671481594310718e-06,
+      "loss": 1.1641,
+      "step": 10885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.056419692430639,
+      "learning_rate": 9.670897039834212e-06,
+      "loss": 1.1314,
+      "step": 10890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.318068102424403,
+      "learning_rate": 9.670311983449207e-06,
+      "loss": 1.1023,
+      "step": 10895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.5630145979294685,
+      "learning_rate": 9.669726425218569e-06,
+      "loss": 1.0887,
+      "step": 10900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.587877654307853,
+      "learning_rate": 9.669140365205216e-06,
+      "loss": 1.1014,
+      "step": 10905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.1165568028333945,
+      "learning_rate": 9.668553803472124e-06,
+      "loss": 1.1145,
+      "step": 10910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.5629788196879115,
+      "learning_rate": 9.667966740082322e-06,
+      "loss": 1.1576,
+      "step": 10915
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.812033807852437,
+      "learning_rate": 9.667379175098892e-06,
+      "loss": 1.1285,
+      "step": 10920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.718265197899088,
+      "learning_rate": 9.666791108584973e-06,
+      "loss": 1.1424,
+      "step": 10925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 44.69221850490623,
+      "learning_rate": 9.66620254060375e-06,
+      "loss": 1.1466,
+      "step": 10930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.271379033231685,
+      "learning_rate": 9.665613471218474e-06,
+      "loss": 1.1064,
+      "step": 10935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.796541858818676,
+      "learning_rate": 9.665023900492437e-06,
+      "loss": 1.1728,
+      "step": 10940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.316819695961973,
+      "learning_rate": 9.664433828488994e-06,
+      "loss": 1.1235,
+      "step": 10945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 20.668594591122304,
+      "learning_rate": 9.66384325527155e-06,
+      "loss": 1.082,
+      "step": 10950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.085360159693657,
+      "learning_rate": 9.663252180903567e-06,
+      "loss": 1.1437,
+      "step": 10955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.637507941505123,
+      "learning_rate": 9.662660605448556e-06,
+      "loss": 1.168,
+      "step": 10960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.885822912023088,
+      "learning_rate": 9.662068528970084e-06,
+      "loss": 1.093,
+      "step": 10965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.247551153290033,
+      "learning_rate": 9.661475951531775e-06,
+      "loss": 1.097,
+      "step": 10970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.014622813957114,
+      "learning_rate": 9.660882873197302e-06,
+      "loss": 1.1774,
+      "step": 10975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.453537591758799,
+      "learning_rate": 9.660289294030392e-06,
+      "loss": 1.1138,
+      "step": 10980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.838324746880735,
+      "learning_rate": 9.659695214094831e-06,
+      "loss": 1.134,
+      "step": 10985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.156741371796533,
+      "learning_rate": 9.659100633454456e-06,
+      "loss": 1.1393,
+      "step": 10990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.96563855251562,
+      "learning_rate": 9.658505552173154e-06,
+      "loss": 1.1404,
+      "step": 10995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 32.05125425748238,
+      "learning_rate": 9.657909970314872e-06,
+      "loss": 1.106,
+      "step": 11000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.088904851396622,
+      "learning_rate": 9.657313887943607e-06,
+      "loss": 1.0857,
+      "step": 11005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.298157623219927,
+      "learning_rate": 9.656717305123409e-06,
+      "loss": 1.1081,
+      "step": 11010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.974985408266656,
+      "learning_rate": 9.656120221918385e-06,
+      "loss": 1.1928,
+      "step": 11015
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.08615653618582,
+      "learning_rate": 9.655522638392696e-06,
+      "loss": 1.1115,
+      "step": 11020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.919446621579743,
+      "learning_rate": 9.654924554610552e-06,
+      "loss": 1.0954,
+      "step": 11025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.256431814444242,
+      "learning_rate": 9.65432597063622e-06,
+      "loss": 1.1222,
+      "step": 11030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.140945493593927,
+      "learning_rate": 9.653726886534022e-06,
+      "loss": 1.1338,
+      "step": 11035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.731735490306578,
+      "learning_rate": 9.653127302368332e-06,
+      "loss": 1.1436,
+      "step": 11040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 18.113124315147232,
+      "learning_rate": 9.652527218203578e-06,
+      "loss": 1.088,
+      "step": 11045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.334135425166144,
+      "learning_rate": 9.65192663410424e-06,
+      "loss": 1.1179,
+      "step": 11050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.709769850492832,
+      "learning_rate": 9.651325550134855e-06,
+      "loss": 1.128,
+      "step": 11055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.286184891474969,
+      "learning_rate": 9.650723966360012e-06,
+      "loss": 1.0974,
+      "step": 11060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.57027027045137,
+      "learning_rate": 9.650121882844353e-06,
+      "loss": 1.0875,
+      "step": 11065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.518062256318835,
+      "learning_rate": 9.649519299652577e-06,
+      "loss": 1.1408,
+      "step": 11070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.44312304087487,
+      "learning_rate": 9.64891621684943e-06,
+      "loss": 1.0707,
+      "step": 11075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 19.777767208451827,
+      "learning_rate": 9.64831263449972e-06,
+      "loss": 1.1336,
+      "step": 11080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.74015049633733,
+      "learning_rate": 9.647708552668301e-06,
+      "loss": 1.1165,
+      "step": 11085
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.864708070004711,
+      "learning_rate": 9.647103971420087e-06,
+      "loss": 1.1266,
+      "step": 11090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.11088254375396,
+      "learning_rate": 9.646498890820044e-06,
+      "loss": 1.1073,
+      "step": 11095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.35672591968441,
+      "learning_rate": 9.645893310933185e-06,
+      "loss": 1.1284,
+      "step": 11100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.738344826519528,
+      "learning_rate": 9.64528723182459e-06,
+      "loss": 1.0926,
+      "step": 11105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.000288653758377,
+      "learning_rate": 9.644680653559377e-06,
+      "loss": 1.0761,
+      "step": 11110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.472794927018789,
+      "learning_rate": 9.64407357620273e-06,
+      "loss": 1.1454,
+      "step": 11115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.352227962786161,
+      "learning_rate": 9.643465999819884e-06,
+      "loss": 1.1623,
+      "step": 11120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.433874733615224,
+      "learning_rate": 9.64285792447612e-06,
+      "loss": 1.0855,
+      "step": 11125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.968813243175576,
+      "learning_rate": 9.642249350236783e-06,
+      "loss": 1.0918,
+      "step": 11130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 15.476869741505018,
+      "learning_rate": 9.641640277167265e-06,
+      "loss": 1.1205,
+      "step": 11135
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.141844055717403,
+      "learning_rate": 9.641030705333014e-06,
+      "loss": 1.1208,
+      "step": 11140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 19.984238053045534,
+      "learning_rate": 9.640420634799531e-06,
+      "loss": 1.1526,
+      "step": 11145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 22.91811477675895,
+      "learning_rate": 9.639810065632373e-06,
+      "loss": 1.0788,
+      "step": 11150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 29.178932705364485,
+      "learning_rate": 9.639198997897144e-06,
+      "loss": 1.1161,
+      "step": 11155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 15.426420048454668,
+      "learning_rate": 9.63858743165951e-06,
+      "loss": 1.1295,
+      "step": 11160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.088387589648,
+      "learning_rate": 9.637975366985185e-06,
+      "loss": 1.1297,
+      "step": 11165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.9223722868315205,
+      "learning_rate": 9.637362803939936e-06,
+      "loss": 1.1254,
+      "step": 11170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.37856404730091,
+      "learning_rate": 9.636749742589588e-06,
+      "loss": 1.1098,
+      "step": 11175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.068630448717673,
+      "learning_rate": 9.636136183000017e-06,
+      "loss": 1.1188,
+      "step": 11180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.227362861670856,
+      "learning_rate": 9.635522125237153e-06,
+      "loss": 1.1407,
+      "step": 11185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 25.22172729921892,
+      "learning_rate": 9.634907569366978e-06,
+      "loss": 1.1145,
+      "step": 11190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.242894970507499,
+      "learning_rate": 9.63429251545553e-06,
+      "loss": 1.0853,
+      "step": 11195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.959942429038014,
+      "learning_rate": 9.633676963568897e-06,
+      "loss": 1.1038,
+      "step": 11200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.917994257142175,
+      "learning_rate": 9.633060913773226e-06,
+      "loss": 1.1111,
+      "step": 11205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.581888544269615,
+      "learning_rate": 9.632444366134713e-06,
+      "loss": 1.1097,
+      "step": 11210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.729514803292108,
+      "learning_rate": 9.631827320719605e-06,
+      "loss": 1.1022,
+      "step": 11215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.545240669926472,
+      "learning_rate": 9.631209777594213e-06,
+      "loss": 1.1128,
+      "step": 11220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.390526908087484,
+      "learning_rate": 9.630591736824888e-06,
+      "loss": 1.1747,
+      "step": 11225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.924949816061919,
+      "learning_rate": 9.629973198478048e-06,
+      "loss": 1.1005,
+      "step": 11230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.8916345427556,
+      "learning_rate": 9.629354162620152e-06,
+      "loss": 1.1376,
+      "step": 11235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.155518159629471,
+      "learning_rate": 9.62873462931772e-06,
+      "loss": 1.101,
+      "step": 11240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.6167227681067695,
+      "learning_rate": 9.628114598637325e-06,
+      "loss": 1.1598,
+      "step": 11245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.280942433837521,
+      "learning_rate": 9.627494070645591e-06,
+      "loss": 1.1055,
+      "step": 11250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.5000890376880385,
+      "learning_rate": 9.626873045409198e-06,
+      "loss": 1.144,
+      "step": 11255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.010584592440216,
+      "learning_rate": 9.626251522994875e-06,
+      "loss": 1.0922,
+      "step": 11260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 22.03434250940769,
+      "learning_rate": 9.625629503469407e-06,
+      "loss": 1.0946,
+      "step": 11265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 20.686737338818922,
+      "learning_rate": 9.625006986899634e-06,
+      "loss": 1.1348,
+      "step": 11270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.696667859947064,
+      "learning_rate": 9.624383973352452e-06,
+      "loss": 1.1247,
+      "step": 11275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.455077034763566,
+      "learning_rate": 9.6237604628948e-06,
+      "loss": 1.1122,
+      "step": 11280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 26.755922699137464,
+      "learning_rate": 9.623136455593682e-06,
+      "loss": 1.1146,
+      "step": 11285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.402977119405639,
+      "learning_rate": 9.622511951516147e-06,
+      "loss": 1.1356,
+      "step": 11290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 15.582304160611613,
+      "learning_rate": 9.621886950729303e-06,
+      "loss": 1.1585,
+      "step": 11295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.307559358077564,
+      "learning_rate": 9.621261453300307e-06,
+      "loss": 1.1062,
+      "step": 11300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 25.541829751747773,
+      "learning_rate": 9.620635459296375e-06,
+      "loss": 1.1223,
+      "step": 11305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.158057177799695,
+      "learning_rate": 9.620008968784768e-06,
+      "loss": 1.1469,
+      "step": 11310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.343254958460044,
+      "learning_rate": 9.619381981832809e-06,
+      "loss": 1.1296,
+      "step": 11315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 22.229399202952123,
+      "learning_rate": 9.618754498507869e-06,
+      "loss": 1.0891,
+      "step": 11320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.81759293368276,
+      "learning_rate": 9.618126518877374e-06,
+      "loss": 1.0831,
+      "step": 11325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.725847092971947,
+      "learning_rate": 9.617498043008805e-06,
+      "loss": 1.0687,
+      "step": 11330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.418814423887673,
+      "learning_rate": 9.616869070969688e-06,
+      "loss": 1.1319,
+      "step": 11335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 84.34338677835348,
+      "learning_rate": 9.616239602827617e-06,
+      "loss": 1.1773,
+      "step": 11340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 30.81992728108807,
+      "learning_rate": 9.615609638650228e-06,
+      "loss": 1.1321,
+      "step": 11345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.365825982658668,
+      "learning_rate": 9.614979178505212e-06,
+      "loss": 1.1478,
+      "step": 11350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.84936705469163,
+      "learning_rate": 9.614348222460315e-06,
+      "loss": 1.0954,
+      "step": 11355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.93458593210332,
+      "learning_rate": 9.613716770583338e-06,
+      "loss": 1.1145,
+      "step": 11360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.001447461288683,
+      "learning_rate": 9.613084822942132e-06,
+      "loss": 1.1395,
+      "step": 11365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.025687520103381,
+      "learning_rate": 9.612452379604602e-06,
+      "loss": 1.1146,
+      "step": 11370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.09606954537574,
+      "learning_rate": 9.611819440638708e-06,
+      "loss": 1.1094,
+      "step": 11375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 15.237321502220459,
+      "learning_rate": 9.611186006112463e-06,
+      "loss": 1.1907,
+      "step": 11380
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.133020630062902,
+      "learning_rate": 9.61055207609393e-06,
+      "loss": 1.1237,
+      "step": 11385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 24.470807860632377,
+      "learning_rate": 9.609917650651227e-06,
+      "loss": 1.1231,
+      "step": 11390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.721493486936613,
+      "learning_rate": 9.609282729852528e-06,
+      "loss": 1.1749,
+      "step": 11395
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.29379204980772,
+      "learning_rate": 9.608647313766057e-06,
+      "loss": 1.1255,
+      "step": 11400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.066054564641332,
+      "learning_rate": 9.608011402460095e-06,
+      "loss": 1.1613,
+      "step": 11405
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.663241784772767,
+      "learning_rate": 9.607374996002969e-06,
+      "loss": 1.0887,
+      "step": 11410
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 25.35957482725709,
+      "learning_rate": 9.606738094463065e-06,
+      "loss": 1.1547,
+      "step": 11415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 34.80978633832712,
+      "learning_rate": 9.606100697908824e-06,
+      "loss": 1.2203,
+      "step": 11420
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.828161132685219,
+      "learning_rate": 9.605462806408735e-06,
+      "loss": 1.1553,
+      "step": 11425
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 32.93371904842403,
+      "learning_rate": 9.60482442003134e-06,
+      "loss": 1.1089,
+      "step": 11430
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 31.082856630775527,
+      "learning_rate": 9.604185538845238e-06,
+      "loss": 1.1546,
+      "step": 11435
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 32.24987579908004,
+      "learning_rate": 9.603546162919082e-06,
+      "loss": 1.1026,
+      "step": 11440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 20.54133506282557,
+      "learning_rate": 9.602906292321572e-06,
+      "loss": 1.155,
+      "step": 11445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.975735053920653,
+      "learning_rate": 9.602265927121468e-06,
+      "loss": 1.148,
+      "step": 11450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.164774929887983,
+      "learning_rate": 9.60162506738758e-06,
+      "loss": 1.1451,
+      "step": 11455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.773471807523148,
+      "learning_rate": 9.600983713188767e-06,
+      "loss": 1.0728,
+      "step": 11460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 38.009791961725284,
+      "learning_rate": 9.60034186459395e-06,
+      "loss": 1.1672,
+      "step": 11465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 33.001981166921155,
+      "learning_rate": 9.599699521672096e-06,
+      "loss": 1.1464,
+      "step": 11470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 21.661184898745827,
+      "learning_rate": 9.59905668449223e-06,
+      "loss": 1.1131,
+      "step": 11475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.865634158608945,
+      "learning_rate": 9.598413353123423e-06,
+      "loss": 1.1506,
+      "step": 11480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.847809446167107,
+      "learning_rate": 9.59776952763481e-06,
+      "loss": 1.1242,
+      "step": 11485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.802102621857443,
+      "learning_rate": 9.597125208095568e-06,
+      "loss": 1.1191,
+      "step": 11490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.835095796388343,
+      "learning_rate": 9.596480394574934e-06,
+      "loss": 1.1456,
+      "step": 11495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.391157972821688,
+      "learning_rate": 9.595835087142197e-06,
+      "loss": 1.125,
+      "step": 11500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.646892000756408,
+      "learning_rate": 9.595189285866695e-06,
+      "loss": 1.1115,
+      "step": 11505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.334311064510455,
+      "learning_rate": 9.594542990817826e-06,
+      "loss": 1.1285,
+      "step": 11510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.386206355177405,
+      "learning_rate": 9.593896202065034e-06,
+      "loss": 1.105,
+      "step": 11515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 18.89044425074765,
+      "learning_rate": 9.593248919677821e-06,
+      "loss": 1.1278,
+      "step": 11520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.383003304725248,
+      "learning_rate": 9.592601143725741e-06,
+      "loss": 1.1239,
+      "step": 11525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.073223723033902,
+      "learning_rate": 9.591952874278398e-06,
+      "loss": 1.1152,
+      "step": 11530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.953832256371221,
+      "learning_rate": 9.591304111405453e-06,
+      "loss": 1.1307,
+      "step": 11535
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.423248294487648,
+      "learning_rate": 9.59065485517662e-06,
+      "loss": 1.1665,
+      "step": 11540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 17.34771202769372,
+      "learning_rate": 9.590005105661661e-06,
+      "loss": 1.0943,
+      "step": 11545
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.829346074591298,
+      "learning_rate": 9.589354862930396e-06,
+      "loss": 1.0949,
+      "step": 11550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 43.08719051302611,
+      "learning_rate": 9.588704127052697e-06,
+      "loss": 1.1418,
+      "step": 11555
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 61.34795040610486,
+      "learning_rate": 9.58805289809849e-06,
+      "loss": 1.1212,
+      "step": 11560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.887902416872716,
+      "learning_rate": 9.587401176137747e-06,
+      "loss": 1.1878,
+      "step": 11565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 31.760457368731775,
+      "learning_rate": 9.586748961240504e-06,
+      "loss": 1.1356,
+      "step": 11570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 22.836393516635077,
+      "learning_rate": 9.586096253476843e-06,
+      "loss": 1.1151,
+      "step": 11575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.1359009032005165,
+      "learning_rate": 9.585443052916897e-06,
+      "loss": 1.1344,
+      "step": 11580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.691199733776354,
+      "learning_rate": 9.58478935963086e-06,
+      "loss": 1.117,
+      "step": 11585
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.466776266088328,
+      "learning_rate": 9.58413517368897e-06,
+      "loss": 1.093,
+      "step": 11590
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.753417729649698,
+      "learning_rate": 9.583480495161525e-06,
+      "loss": 1.1043,
+      "step": 11595
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.974458243917208,
+      "learning_rate": 9.582825324118874e-06,
+      "loss": 1.0903,
+      "step": 11600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.199533039789745,
+      "learning_rate": 9.582169660631416e-06,
+      "loss": 1.113,
+      "step": 11605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.3870764961647035,
+      "learning_rate": 9.581513504769603e-06,
+      "loss": 1.0672,
+      "step": 11610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.472341001370613,
+      "learning_rate": 9.580856856603947e-06,
+      "loss": 1.1478,
+      "step": 11615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.022746532359818,
+      "learning_rate": 9.580199716205003e-06,
+      "loss": 1.1633,
+      "step": 11620
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.709660529258535,
+      "learning_rate": 9.579542083643387e-06,
+      "loss": 1.1064,
+      "step": 11625
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.06231994676766,
+      "learning_rate": 9.578883958989762e-06,
+      "loss": 1.093,
+      "step": 11630
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.955678856201514,
+      "learning_rate": 9.578225342314848e-06,
+      "loss": 1.1366,
+      "step": 11635
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.496948620728515,
+      "learning_rate": 9.577566233689416e-06,
+      "loss": 1.15,
+      "step": 11640
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.55004295467598,
+      "learning_rate": 9.576906633184289e-06,
+      "loss": 1.1449,
+      "step": 11645
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.379818781430924,
+      "learning_rate": 9.576246540870345e-06,
+      "loss": 1.1382,
+      "step": 11650
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.247858126577775,
+      "learning_rate": 9.575585956818513e-06,
+      "loss": 1.1147,
+      "step": 11655
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.38251281562076,
+      "learning_rate": 9.574924881099778e-06,
+      "loss": 1.0541,
+      "step": 11660
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.237389167544144,
+      "learning_rate": 9.574263313785173e-06,
+      "loss": 1.1484,
+      "step": 11665
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.060677713515952,
+      "learning_rate": 9.573601254945787e-06,
+      "loss": 1.2049,
+      "step": 11670
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.757179417318337,
+      "learning_rate": 9.572938704652762e-06,
+      "loss": 1.1077,
+      "step": 11675
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.052695169299957,
+      "learning_rate": 9.57227566297729e-06,
+      "loss": 1.1462,
+      "step": 11680
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.8877390017597495,
+      "learning_rate": 9.571612129990619e-06,
+      "loss": 1.0599,
+      "step": 11685
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.081068887101823,
+      "learning_rate": 9.57094810576405e-06,
+      "loss": 1.1275,
+      "step": 11690
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.668535300593515,
+      "learning_rate": 9.570283590368932e-06,
+      "loss": 1.1313,
+      "step": 11695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.314851210452614,
+      "learning_rate": 9.569618583876675e-06,
+      "loss": 1.1157,
+      "step": 11700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.089547607492213,
+      "learning_rate": 9.568953086358732e-06,
+      "loss": 1.0894,
+      "step": 11705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.295051141884013,
+      "learning_rate": 9.568287097886614e-06,
+      "loss": 1.0726,
+      "step": 11710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.4161503464096405,
+      "learning_rate": 9.567620618531889e-06,
+      "loss": 1.1214,
+      "step": 11715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.554322530448706,
+      "learning_rate": 9.566953648366169e-06,
+      "loss": 1.107,
+      "step": 11720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.651018714519556,
+      "learning_rate": 9.566286187461122e-06,
+      "loss": 1.0729,
+      "step": 11725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.22959428725363,
+      "learning_rate": 9.565618235888474e-06,
+      "loss": 1.1256,
+      "step": 11730
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.850813398519219,
+      "learning_rate": 9.564949793719997e-06,
+      "loss": 1.1022,
+      "step": 11735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.90267362929282,
+      "learning_rate": 9.564280861027517e-06,
+      "loss": 1.116,
+      "step": 11740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.369881594861055,
+      "learning_rate": 9.563611437882914e-06,
+      "loss": 1.0746,
+      "step": 11745
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.54724950131561,
+      "learning_rate": 9.562941524358123e-06,
+      "loss": 1.0923,
+      "step": 11750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.606386548664831,
+      "learning_rate": 9.562271120525127e-06,
+      "loss": 1.1049,
+      "step": 11755
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.328639732082031,
+      "learning_rate": 9.561600226455964e-06,
+      "loss": 1.1159,
+      "step": 11760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.54670197631698,
+      "learning_rate": 9.560928842222724e-06,
+      "loss": 1.0662,
+      "step": 11765
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.447986882313948,
+      "learning_rate": 9.560256967897553e-06,
+      "loss": 1.1336,
+      "step": 11770
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.7565950850095975,
+      "learning_rate": 9.559584603552642e-06,
+      "loss": 1.1301,
+      "step": 11775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.995963174413299,
+      "learning_rate": 9.558911749260245e-06,
+      "loss": 1.0952,
+      "step": 11780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.126469703774925,
+      "learning_rate": 9.558238405092658e-06,
+      "loss": 1.0694,
+      "step": 11785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.828798065105701,
+      "learning_rate": 9.55756457112224e-06,
+      "loss": 1.0996,
+      "step": 11790
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.771380834874629,
+      "learning_rate": 9.556890247421393e-06,
+      "loss": 1.09,
+      "step": 11795
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.759844219172296,
+      "learning_rate": 9.556215434062579e-06,
+      "loss": 1.1084,
+      "step": 11800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.8439149752138055,
+      "learning_rate": 9.555540131118308e-06,
+      "loss": 1.1204,
+      "step": 11805
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.6160793086283505,
+      "learning_rate": 9.554864338661144e-06,
+      "loss": 1.133,
+      "step": 11810
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.417168157300613,
+      "learning_rate": 9.554188056763706e-06,
+      "loss": 1.0639,
+      "step": 11815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.622621161220346,
+      "learning_rate": 9.553511285498661e-06,
+      "loss": 1.1193,
+      "step": 11820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.965269087906013,
+      "learning_rate": 9.552834024938733e-06,
+      "loss": 1.1219,
+      "step": 11825
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.255698255179819,
+      "learning_rate": 9.552156275156696e-06,
+      "loss": 1.1086,
+      "step": 11830
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.945998298437943,
+      "learning_rate": 9.551478036225376e-06,
+      "loss": 1.1316,
+      "step": 11835
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.717388145642351,
+      "learning_rate": 9.550799308217655e-06,
+      "loss": 1.1139,
+      "step": 11840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.814865182414534,
+      "learning_rate": 9.550120091206463e-06,
+      "loss": 1.1251,
+      "step": 11845
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.874893174733337,
+      "learning_rate": 9.549440385264786e-06,
+      "loss": 1.1257,
+      "step": 11850
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.089310520514331,
+      "learning_rate": 9.548760190465663e-06,
+      "loss": 1.1265,
+      "step": 11855
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.172026895489223,
+      "learning_rate": 9.54807950688218e-06,
+      "loss": 1.074,
+      "step": 11860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.840493183637697,
+      "learning_rate": 9.547398334587483e-06,
+      "loss": 1.125,
+      "step": 11865
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.200894965676413,
+      "learning_rate": 9.546716673654766e-06,
+      "loss": 1.1552,
+      "step": 11870
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.473078826019139,
+      "learning_rate": 9.546034524157275e-06,
+      "loss": 1.1332,
+      "step": 11875
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.216539156849562,
+      "learning_rate": 9.545351886168313e-06,
+      "loss": 1.1079,
+      "step": 11880
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.450721010310948,
+      "learning_rate": 9.544668759761228e-06,
+      "loss": 1.1098,
+      "step": 11885
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.544296233518932,
+      "learning_rate": 9.543985145009431e-06,
+      "loss": 1.1314,
+      "step": 11890
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.429425340593898,
+      "learning_rate": 9.543301041986375e-06,
+      "loss": 1.1266,
+      "step": 11895
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.186894712391446,
+      "learning_rate": 9.542616450765569e-06,
+      "loss": 1.1712,
+      "step": 11900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.400204099652655,
+      "learning_rate": 9.541931371420579e-06,
+      "loss": 1.0878,
+      "step": 11905
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 20.827900543882034,
+      "learning_rate": 9.541245804025018e-06,
+      "loss": 1.1044,
+      "step": 11910
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.128200251301965,
+      "learning_rate": 9.540559748652553e-06,
+      "loss": 1.122,
+      "step": 11915
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.20857238719293,
+      "learning_rate": 9.539873205376903e-06,
+      "loss": 1.122,
+      "step": 11920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 27.086352044189386,
+      "learning_rate": 9.539186174271843e-06,
+      "loss": 1.1286,
+      "step": 11925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.218026239759777,
+      "learning_rate": 9.538498655411194e-06,
+      "loss": 1.1099,
+      "step": 11930
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.664747237291802,
+      "learning_rate": 9.537810648868835e-06,
+      "loss": 1.1122,
+      "step": 11935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.146816761770653,
+      "learning_rate": 9.537122154718697e-06,
+      "loss": 1.0863,
+      "step": 11940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.066789291300878,
+      "learning_rate": 9.536433173034756e-06,
+      "loss": 1.1175,
+      "step": 11945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.20282720723064,
+      "learning_rate": 9.535743703891052e-06,
+      "loss": 1.0828,
+      "step": 11950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.36317320547394,
+      "learning_rate": 9.535053747361668e-06,
+      "loss": 1.2184,
+      "step": 11955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.969134634017165,
+      "learning_rate": 9.534363303520747e-06,
+      "loss": 1.1383,
+      "step": 11960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.74126901801494,
+      "learning_rate": 9.533672372442472e-06,
+      "loss": 1.0959,
+      "step": 11965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.145065327359825,
+      "learning_rate": 9.532980954201096e-06,
+      "loss": 1.1254,
+      "step": 11970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.073135816460807,
+      "learning_rate": 9.532289048870911e-06,
+      "loss": 1.1273,
+      "step": 11975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.701958034273098,
+      "learning_rate": 9.531596656526263e-06,
+      "loss": 1.165,
+      "step": 11980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.511324696376247,
+      "learning_rate": 9.530903777241558e-06,
+      "loss": 1.0995,
+      "step": 11985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.987164352630714,
+      "learning_rate": 9.530210411091244e-06,
+      "loss": 1.1514,
+      "step": 11990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 25.501563990598655,
+      "learning_rate": 9.529516558149828e-06,
+      "loss": 1.1004,
+      "step": 11995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.545471466725064,
+      "learning_rate": 9.528822218491867e-06,
+      "loss": 1.109,
+      "step": 12000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.807814285801163,
+      "learning_rate": 9.52812739219197e-06,
+      "loss": 1.1089,
+      "step": 12005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.0473320061145674,
+      "learning_rate": 9.527432079324804e-06,
+      "loss": 1.0957,
+      "step": 12010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.348008458634869,
+      "learning_rate": 9.526736279965077e-06,
+      "loss": 1.0909,
+      "step": 12015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.184962795913268,
+      "learning_rate": 9.526039994187561e-06,
+      "loss": 1.1278,
+      "step": 12020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.793071498515623,
+      "learning_rate": 9.525343222067071e-06,
+      "loss": 1.1198,
+      "step": 12025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 26.328754883464295,
+      "learning_rate": 9.524645963678481e-06,
+      "loss": 1.1322,
+      "step": 12030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.585776506978878,
+      "learning_rate": 9.523948219096712e-06,
+      "loss": 1.1266,
+      "step": 12035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.381052915621866,
+      "learning_rate": 9.523249988396744e-06,
+      "loss": 1.1359,
+      "step": 12040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 28.246476797003368,
+      "learning_rate": 9.5225512716536e-06,
+      "loss": 1.1124,
+      "step": 12045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.77813837924769,
+      "learning_rate": 9.521852068942363e-06,
+      "loss": 1.0634,
+      "step": 12050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.407532549369275,
+      "learning_rate": 9.521152380338163e-06,
+      "loss": 1.1502,
+      "step": 12055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.363183334664006,
+      "learning_rate": 9.520452205916188e-06,
+      "loss": 1.1825,
+      "step": 12060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.088172289393352,
+      "learning_rate": 9.519751545751674e-06,
+      "loss": 1.1215,
+      "step": 12065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.068443371883795,
+      "learning_rate": 9.519050399919908e-06,
+      "loss": 1.1404,
+      "step": 12070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.890841705730148,
+      "learning_rate": 9.518348768496233e-06,
+      "loss": 1.0997,
+      "step": 12075
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 34.471194434444435,
+      "learning_rate": 9.51764665155604e-06,
+      "loss": 1.1898,
+      "step": 12080
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.6024386929285575,
+      "learning_rate": 9.51694404917478e-06,
+      "loss": 1.1087,
+      "step": 12085
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 19.07403481880703,
+      "learning_rate": 9.516240961427943e-06,
+      "loss": 1.0759,
+      "step": 12090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.854328530035872,
+      "learning_rate": 9.515537388391085e-06,
+      "loss": 1.1387,
+      "step": 12095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.475056391091014,
+      "learning_rate": 9.514833330139805e-06,
+      "loss": 1.1455,
+      "step": 12100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.198787155357344,
+      "learning_rate": 9.51412878674976e-06,
+      "loss": 1.1423,
+      "step": 12105
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.039969911802466,
+      "learning_rate": 9.513423758296653e-06,
+      "loss": 1.1566,
+      "step": 12110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.913201267676782,
+      "learning_rate": 9.512718244856245e-06,
+      "loss": 1.0734,
+      "step": 12115
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.850794747649243,
+      "learning_rate": 9.512012246504346e-06,
+      "loss": 1.1062,
+      "step": 12120
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.115240551426224,
+      "learning_rate": 9.511305763316817e-06,
+      "loss": 1.1188,
+      "step": 12125
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.138770327836259,
+      "learning_rate": 9.510598795369575e-06,
+      "loss": 1.0632,
+      "step": 12130
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.139595677373027,
+      "learning_rate": 9.509891342738583e-06,
+      "loss": 1.1162,
+      "step": 12135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.661063797525696,
+      "learning_rate": 9.509183405499865e-06,
+      "loss": 1.0781,
+      "step": 12140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 24.926770999531197,
+      "learning_rate": 9.508474983729487e-06,
+      "loss": 1.1565,
+      "step": 12145
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.598977988434706,
+      "learning_rate": 9.507766077503577e-06,
+      "loss": 1.1564,
+      "step": 12150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.651586997734471,
+      "learning_rate": 9.507056686898308e-06,
+      "loss": 1.0918,
+      "step": 12155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 22.631146266550914,
+      "learning_rate": 9.506346811989905e-06,
+      "loss": 1.134,
+      "step": 12160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.340879746240121,
+      "learning_rate": 9.50563645285465e-06,
+      "loss": 1.1409,
+      "step": 12165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.252479325858253,
+      "learning_rate": 9.504925609568874e-06,
+      "loss": 1.1211,
+      "step": 12170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.2896935670529395,
+      "learning_rate": 9.504214282208957e-06,
+      "loss": 1.147,
+      "step": 12175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.6237948896331735,
+      "learning_rate": 9.503502470851338e-06,
+      "loss": 1.2065,
+      "step": 12180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.299142755373992,
+      "learning_rate": 9.502790175572506e-06,
+      "loss": 1.1081,
+      "step": 12185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.382570891103345,
+      "learning_rate": 9.502077396448992e-06,
+      "loss": 1.0987,
+      "step": 12190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.755366928410425,
+      "learning_rate": 9.501364133557396e-06,
+      "loss": 1.0734,
+      "step": 12195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.063385333619541,
+      "learning_rate": 9.500650386974358e-06,
+      "loss": 1.1213,
+      "step": 12200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.343002531403815,
+      "learning_rate": 9.499936156776571e-06,
+      "loss": 1.1006,
+      "step": 12205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.078932580486301,
+      "learning_rate": 9.499221443040784e-06,
+      "loss": 1.0581,
+      "step": 12210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.873417008498132,
+      "learning_rate": 9.4985062458438e-06,
+      "loss": 1.1499,
+      "step": 12215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.043719682987097,
+      "learning_rate": 9.497790565262464e-06,
+      "loss": 1.0777,
+      "step": 12220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.991103333781751,
+      "learning_rate": 9.49707440137368e-06,
+      "loss": 1.1144,
+      "step": 12225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.454431628315685,
+      "learning_rate": 9.496357754254406e-06,
+      "loss": 1.1191,
+      "step": 12230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.117265450369398,
+      "learning_rate": 9.495640623981648e-06,
+      "loss": 1.1407,
+      "step": 12235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.431191236012026,
+      "learning_rate": 9.494923010632465e-06,
+      "loss": 1.1575,
+      "step": 12240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 23.020638829126447,
+      "learning_rate": 9.494204914283964e-06,
+      "loss": 1.1306,
+      "step": 12245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.48881759716046,
+      "learning_rate": 9.493486335013313e-06,
+      "loss": 1.1143,
+      "step": 12250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.75435397065997,
+      "learning_rate": 9.492767272897723e-06,
+      "loss": 1.1055,
+      "step": 12255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.7700584546842,
+      "learning_rate": 9.492047728014463e-06,
+      "loss": 1.1206,
+      "step": 12260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.537033957318727,
+      "learning_rate": 9.491327700440847e-06,
+      "loss": 1.1481,
+      "step": 12265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.04442380933225,
+      "learning_rate": 9.490607190254248e-06,
+      "loss": 1.1422,
+      "step": 12270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.898531664093339,
+      "learning_rate": 9.48988619753209e-06,
+      "loss": 1.108,
+      "step": 12275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.662052853577302,
+      "learning_rate": 9.489164722351843e-06,
+      "loss": 1.1449,
+      "step": 12280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.7918430639108,
+      "learning_rate": 9.488442764791034e-06,
+      "loss": 1.1292,
+      "step": 12285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.608927898417105,
+      "learning_rate": 9.487720324927243e-06,
+      "loss": 1.0711,
+      "step": 12290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.709125647527724,
+      "learning_rate": 9.486997402838096e-06,
+      "loss": 1.1668,
+      "step": 12295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.814286830550152,
+      "learning_rate": 9.486273998601276e-06,
+      "loss": 1.1231,
+      "step": 12300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.057419899841028,
+      "learning_rate": 9.485550112294515e-06,
+      "loss": 1.1414,
+      "step": 12305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.894706702379888,
+      "learning_rate": 9.484825743995598e-06,
+      "loss": 1.0623,
+      "step": 12310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.476779439178674,
+      "learning_rate": 9.484100893782362e-06,
+      "loss": 1.1581,
+      "step": 12315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.564348801849615,
+      "learning_rate": 9.483375561732696e-06,
+      "loss": 1.1211,
+      "step": 12320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.079720238698098,
+      "learning_rate": 9.482649747924538e-06,
+      "loss": 1.1678,
+      "step": 12325
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.497766593119248,
+      "learning_rate": 9.481923452435882e-06,
+      "loss": 1.1616,
+      "step": 12330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.077638479398647,
+      "learning_rate": 9.481196675344769e-06,
+      "loss": 1.1073,
+      "step": 12335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.133263841220293,
+      "learning_rate": 9.480469416729297e-06,
+      "loss": 1.1109,
+      "step": 12340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.066355834477365,
+      "learning_rate": 9.479741676667614e-06,
+      "loss": 1.1155,
+      "step": 12345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 34.65322650311844,
+      "learning_rate": 9.479013455237916e-06,
+      "loss": 1.1326,
+      "step": 12350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 25.697613214100894,
+      "learning_rate": 9.478284752518455e-06,
+      "loss": 1.1244,
+      "step": 12355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.598786341599224,
+      "learning_rate": 9.477555568587533e-06,
+      "loss": 1.1553,
+      "step": 12360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.598780582515386,
+      "learning_rate": 9.476825903523503e-06,
+      "loss": 1.1745,
+      "step": 12365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 30.328799584656238,
+      "learning_rate": 9.476095757404775e-06,
+      "loss": 1.1028,
+      "step": 12370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 27.474643890991686,
+      "learning_rate": 9.4753651303098e-06,
+      "loss": 1.1761,
+      "step": 12375
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.396515626595093,
+      "learning_rate": 9.474634022317093e-06,
+      "loss": 1.121,
+      "step": 12380
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.16064964195702,
+      "learning_rate": 9.473902433505211e-06,
+      "loss": 1.1631,
+      "step": 12385
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.314931852071062,
+      "learning_rate": 9.473170363952768e-06,
+      "loss": 1.1405,
+      "step": 12390
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.304196512597075,
+      "learning_rate": 9.472437813738427e-06,
+      "loss": 1.1715,
+      "step": 12395
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.657950395685992,
+      "learning_rate": 9.471704782940905e-06,
+      "loss": 1.1413,
+      "step": 12400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.089581989436484,
+      "learning_rate": 9.47097127163897e-06,
+      "loss": 1.1055,
+      "step": 12405
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.120573640959808,
+      "learning_rate": 9.470237279911439e-06,
+      "loss": 1.1003,
+      "step": 12410
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.316663572348183,
+      "learning_rate": 9.469502807837184e-06,
+      "loss": 1.115,
+      "step": 12415
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.674143620666513,
+      "learning_rate": 9.468767855495126e-06,
+      "loss": 1.0986,
+      "step": 12420
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.409574716827853,
+      "learning_rate": 9.468032422964241e-06,
+      "loss": 1.0582,
+      "step": 12425
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.799152871815183,
+      "learning_rate": 9.467296510323553e-06,
+      "loss": 1.1067,
+      "step": 12430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 38.93868324170623,
+      "learning_rate": 9.466560117652138e-06,
+      "loss": 1.156,
+      "step": 12435
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.158157085498642,
+      "learning_rate": 9.465823245029128e-06,
+      "loss": 1.1023,
+      "step": 12440
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.20520513141992,
+      "learning_rate": 9.465085892533698e-06,
+      "loss": 1.1292,
+      "step": 12445
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.32160737032039,
+      "learning_rate": 9.464348060245086e-06,
+      "loss": 1.0815,
+      "step": 12450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.380708385902926,
+      "learning_rate": 9.463609748242571e-06,
+      "loss": 1.1177,
+      "step": 12455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.72739984623385,
+      "learning_rate": 9.46287095660549e-06,
+      "loss": 1.1198,
+      "step": 12460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.125505495264395,
+      "learning_rate": 9.462131685413227e-06,
+      "loss": 1.116,
+      "step": 12465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.8834113702680035,
+      "learning_rate": 9.461391934745222e-06,
+      "loss": 1.1438,
+      "step": 12470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.843331232735709,
+      "learning_rate": 9.460651704680965e-06,
+      "loss": 1.1385,
+      "step": 12475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.1626993983384155,
+      "learning_rate": 9.459910995299995e-06,
+      "loss": 1.1198,
+      "step": 12480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 29.643221699883394,
+      "learning_rate": 9.459169806681906e-06,
+      "loss": 1.1186,
+      "step": 12485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.42517040378149,
+      "learning_rate": 9.458428138906342e-06,
+      "loss": 1.1036,
+      "step": 12490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.556379813323755,
+      "learning_rate": 9.457685992052996e-06,
+      "loss": 1.1198,
+      "step": 12495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.551313425469406,
+      "learning_rate": 9.456943366201618e-06,
+      "loss": 1.095,
+      "step": 12500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.183185951237004,
+      "learning_rate": 9.456200261432006e-06,
+      "loss": 1.1424,
+      "step": 12505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.03201152976163,
+      "learning_rate": 9.455456677824009e-06,
+      "loss": 1.1254,
+      "step": 12510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.551929579653649,
+      "learning_rate": 9.454712615457528e-06,
+      "loss": 1.1135,
+      "step": 12515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.795386521634065,
+      "learning_rate": 9.453968074412516e-06,
+      "loss": 1.0912,
+      "step": 12520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.142409810564299,
+      "learning_rate": 9.453223054768978e-06,
+      "loss": 1.1362,
+      "step": 12525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.122040193018915,
+      "learning_rate": 9.45247755660697e-06,
+      "loss": 1.1672,
+      "step": 12530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 18.03597120205186,
+      "learning_rate": 9.451731580006596e-06,
+      "loss": 1.1714,
+      "step": 12535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.957972181590634,
+      "learning_rate": 9.450985125048018e-06,
+      "loss": 1.1231,
+      "step": 12540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.068861456533899,
+      "learning_rate": 9.450238191811445e-06,
+      "loss": 1.1487,
+      "step": 12545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.735441167765313,
+      "learning_rate": 9.449490780377139e-06,
+      "loss": 1.1021,
+      "step": 12550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.475548072152762,
+      "learning_rate": 9.44874289082541e-06,
+      "loss": 1.1374,
+      "step": 12555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.076666324686386,
+      "learning_rate": 9.447994523236623e-06,
+      "loss": 1.1247,
+      "step": 12560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.846464886781207,
+      "learning_rate": 9.447245677691194e-06,
+      "loss": 1.1001,
+      "step": 12565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.457530808912342,
+      "learning_rate": 9.446496354269591e-06,
+      "loss": 1.1332,
+      "step": 12570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.20611072601647,
+      "learning_rate": 9.44574655305233e-06,
+      "loss": 1.1267,
+      "step": 12575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.860558295108078,
+      "learning_rate": 9.444996274119981e-06,
+      "loss": 1.1038,
+      "step": 12580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.990814196429046,
+      "learning_rate": 9.444245517553168e-06,
+      "loss": 1.115,
+      "step": 12585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.581261097929349,
+      "learning_rate": 9.443494283432558e-06,
+      "loss": 1.1787,
+      "step": 12590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.616809454658205,
+      "learning_rate": 9.442742571838877e-06,
+      "loss": 1.0425,
+      "step": 12595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 16.261423757744502,
+      "learning_rate": 9.441990382852898e-06,
+      "loss": 1.1253,
+      "step": 12600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.263933207187424,
+      "learning_rate": 9.44123771655545e-06,
+      "loss": 1.1094,
+      "step": 12605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.228412573657085,
+      "learning_rate": 9.440484573027406e-06,
+      "loss": 1.1011,
+      "step": 12610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.28916239685707,
+      "learning_rate": 9.4397309523497e-06,
+      "loss": 1.0897,
+      "step": 12615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 31.018750337265722,
+      "learning_rate": 9.438976854603308e-06,
+      "loss": 1.1461,
+      "step": 12620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 26.742957868697705,
+      "learning_rate": 9.438222279869263e-06,
+      "loss": 1.1322,
+      "step": 12625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 35.31543257261958,
+      "learning_rate": 9.437467228228647e-06,
+      "loss": 1.1515,
+      "step": 12630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.414277899110981,
+      "learning_rate": 9.436711699762592e-06,
+      "loss": 1.1076,
+      "step": 12635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.928930974362663,
+      "learning_rate": 9.435955694552284e-06,
+      "loss": 1.1389,
+      "step": 12640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.980139924511439,
+      "learning_rate": 9.435199212678961e-06,
+      "loss": 1.0712,
+      "step": 12645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.085234525025728,
+      "learning_rate": 9.434442254223907e-06,
+      "loss": 1.0766,
+      "step": 12650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.039348080058898,
+      "learning_rate": 9.433684819268463e-06,
+      "loss": 1.1415,
+      "step": 12655
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.361505124190733,
+      "learning_rate": 9.432926907894018e-06,
+      "loss": 1.1213,
+      "step": 12660
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.373513634400794,
+      "learning_rate": 9.432168520182012e-06,
+      "loss": 1.1114,
+      "step": 12665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.208057103046562,
+      "learning_rate": 9.431409656213938e-06,
+      "loss": 1.1266,
+      "step": 12670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.63737716382715,
+      "learning_rate": 9.43065031607134e-06,
+      "loss": 1.1667,
+      "step": 12675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.35561817323881,
+      "learning_rate": 9.42989049983581e-06,
+      "loss": 1.1265,
+      "step": 12680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.766013721858562,
+      "learning_rate": 9.429130207588996e-06,
+      "loss": 1.1129,
+      "step": 12685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.69039586276047,
+      "learning_rate": 9.428369439412593e-06,
+      "loss": 1.1336,
+      "step": 12690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.336678591045394,
+      "learning_rate": 9.42760819538835e-06,
+      "loss": 1.1616,
+      "step": 12695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.158976172458836,
+      "learning_rate": 9.426846475598068e-06,
+      "loss": 1.0746,
+      "step": 12700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.24143874107018,
+      "learning_rate": 9.426084280123593e-06,
+      "loss": 1.1158,
+      "step": 12705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.022196927358343,
+      "learning_rate": 9.425321609046828e-06,
+      "loss": 1.125,
+      "step": 12710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.518529232070218,
+      "learning_rate": 9.42455846244973e-06,
+      "loss": 1.1057,
+      "step": 12715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.04460834141142,
+      "learning_rate": 9.423794840414293e-06,
+      "loss": 1.1004,
+      "step": 12720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.614250069444788,
+      "learning_rate": 9.423030743022578e-06,
+      "loss": 1.1326,
+      "step": 12725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.289685085298933,
+      "learning_rate": 9.422266170356691e-06,
+      "loss": 1.1302,
+      "step": 12730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.474578407116928,
+      "learning_rate": 9.421501122498786e-06,
+      "loss": 1.1302,
+      "step": 12735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.38616676280056,
+      "learning_rate": 9.420735599531072e-06,
+      "loss": 1.0935,
+      "step": 12740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.511450693752472,
+      "learning_rate": 9.419969601535807e-06,
+      "loss": 1.182,
+      "step": 12745
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 19.66843139983669,
+      "learning_rate": 9.419203128595303e-06,
+      "loss": 1.1721,
+      "step": 12750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.951035396013895,
+      "learning_rate": 9.418436180791916e-06,
+      "loss": 1.0998,
+      "step": 12755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 42.28858037071631,
+      "learning_rate": 9.417668758208065e-06,
+      "loss": 1.1737,
+      "step": 12760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.417100947954943,
+      "learning_rate": 9.416900860926208e-06,
+      "loss": 1.0882,
+      "step": 12765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.57613525666104,
+      "learning_rate": 9.41613248902886e-06,
+      "loss": 1.0914,
+      "step": 12770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.205950045129088,
+      "learning_rate": 9.415363642598584e-06,
+      "loss": 1.1423,
+      "step": 12775
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 41.362035494699974,
+      "learning_rate": 9.414594321718002e-06,
+      "loss": 1.0901,
+      "step": 12780
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 39.278881813521956,
+      "learning_rate": 9.413824526469776e-06,
+      "loss": 1.1721,
+      "step": 12785
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 25.838193458336175,
+      "learning_rate": 9.413054256936625e-06,
+      "loss": 1.1703,
+      "step": 12790
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.044351025135112,
+      "learning_rate": 9.412283513201316e-06,
+      "loss": 1.1498,
+      "step": 12795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 30.754042227083023,
+      "learning_rate": 9.411512295346673e-06,
+      "loss": 1.1617,
+      "step": 12800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 33.277278107276686,
+      "learning_rate": 9.410740603455561e-06,
+      "loss": 1.1287,
+      "step": 12805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.448915195955019,
+      "learning_rate": 9.409968437610908e-06,
+      "loss": 1.106,
+      "step": 12810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.749745902002564,
+      "learning_rate": 9.409195797895683e-06,
+      "loss": 1.1242,
+      "step": 12815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.928985641085489,
+      "learning_rate": 9.40842268439291e-06,
+      "loss": 1.1539,
+      "step": 12820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.05039753422052,
+      "learning_rate": 9.407649097185665e-06,
+      "loss": 1.1024,
+      "step": 12825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.689759133065871,
+      "learning_rate": 9.406875036357071e-06,
+      "loss": 1.1873,
+      "step": 12830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.944921197427781,
+      "learning_rate": 9.406100501990305e-06,
+      "loss": 1.1143,
+      "step": 12835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.128427185251159,
+      "learning_rate": 9.405325494168594e-06,
+      "loss": 1.1226,
+      "step": 12840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.106836168194338,
+      "learning_rate": 9.404550012975218e-06,
+      "loss": 1.1855,
+      "step": 12845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.895518718252985,
+      "learning_rate": 9.403774058493502e-06,
+      "loss": 1.0902,
+      "step": 12850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.105264985860612,
+      "learning_rate": 9.40299763080683e-06,
+      "loss": 1.1333,
+      "step": 12855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.406074166906579,
+      "learning_rate": 9.402220729998631e-06,
+      "loss": 1.1542,
+      "step": 12860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.42796449588289,
+      "learning_rate": 9.401443356152384e-06,
+      "loss": 1.0913,
+      "step": 12865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 29.87094411953661,
+      "learning_rate": 9.400665509351627e-06,
+      "loss": 1.1323,
+      "step": 12870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 16.188931817073026,
+      "learning_rate": 9.399887189679936e-06,
+      "loss": 1.1057,
+      "step": 12875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.538725457071587,
+      "learning_rate": 9.399108397220949e-06,
+      "loss": 1.0832,
+      "step": 12880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 17.01460622218668,
+      "learning_rate": 9.398329132058351e-06,
+      "loss": 1.1324,
+      "step": 12885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.268106924225547,
+      "learning_rate": 9.397549394275875e-06,
+      "loss": 1.1151,
+      "step": 12890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.245982486740813,
+      "learning_rate": 9.39676918395731e-06,
+      "loss": 1.1472,
+      "step": 12895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.940761226310704,
+      "learning_rate": 9.395988501186491e-06,
+      "loss": 1.1105,
+      "step": 12900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.575413387910367,
+      "learning_rate": 9.395207346047308e-06,
+      "loss": 1.1148,
+      "step": 12905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.495914888733515,
+      "learning_rate": 9.394425718623697e-06,
+      "loss": 1.1721,
+      "step": 12910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.274415686603875,
+      "learning_rate": 9.393643618999649e-06,
+      "loss": 1.0729,
+      "step": 12915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.731290248345159,
+      "learning_rate": 9.392861047259202e-06,
+      "loss": 1.1506,
+      "step": 12920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.641337298477747,
+      "learning_rate": 9.39207800348645e-06,
+      "loss": 1.1063,
+      "step": 12925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.427365041800352,
+      "learning_rate": 9.391294487765532e-06,
+      "loss": 1.1098,
+      "step": 12930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.378995705528802,
+      "learning_rate": 9.390510500180641e-06,
+      "loss": 1.0928,
+      "step": 12935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.170561431325469,
+      "learning_rate": 9.389726040816018e-06,
+      "loss": 1.1684,
+      "step": 12940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.561798633145985,
+      "learning_rate": 9.388941109755962e-06,
+      "loss": 1.1274,
+      "step": 12945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.0767443314504,
+      "learning_rate": 9.388155707084811e-06,
+      "loss": 1.1512,
+      "step": 12950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.904459162577236,
+      "learning_rate": 9.387369832886966e-06,
+      "loss": 1.128,
+      "step": 12955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.6162427506300805,
+      "learning_rate": 9.386583487246868e-06,
+      "loss": 1.1184,
+      "step": 12960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.286757855866375,
+      "learning_rate": 9.385796670249014e-06,
+      "loss": 1.1162,
+      "step": 12965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.1933303579712,
+      "learning_rate": 9.385009381977953e-06,
+      "loss": 1.1486,
+      "step": 12970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.081005075327386,
+      "learning_rate": 9.38422162251828e-06,
+      "loss": 1.1298,
+      "step": 12975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.3677400467950065,
+      "learning_rate": 9.383433391954646e-06,
+      "loss": 1.0608,
+      "step": 12980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.733173453317965,
+      "learning_rate": 9.382644690371748e-06,
+      "loss": 1.1192,
+      "step": 12985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.855145146612309,
+      "learning_rate": 9.381855517854337e-06,
+      "loss": 1.1356,
+      "step": 12990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.39183657773669,
+      "learning_rate": 9.38106587448721e-06,
+      "loss": 1.113,
+      "step": 12995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.329719563088435,
+      "learning_rate": 9.380275760355221e-06,
+      "loss": 1.1151,
+      "step": 13000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.092401493419849,
+      "learning_rate": 9.37948517554327e-06,
+      "loss": 1.0717,
+      "step": 13005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.658258926016858,
+      "learning_rate": 9.378694120136307e-06,
+      "loss": 1.1207,
+      "step": 13010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 33.77583955578754,
+      "learning_rate": 9.377902594219338e-06,
+      "loss": 1.1206,
+      "step": 13015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.825310853171759,
+      "learning_rate": 9.377110597877412e-06,
+      "loss": 1.153,
+      "step": 13020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 24.86660956764498,
+      "learning_rate": 9.376318131195637e-06,
+      "loss": 1.1212,
+      "step": 13025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.584119360644792,
+      "learning_rate": 9.375525194259163e-06,
+      "loss": 1.098,
+      "step": 13030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 36.207602275449005,
+      "learning_rate": 9.374731787153197e-06,
+      "loss": 1.119,
+      "step": 13035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 26.33804591732181,
+      "learning_rate": 9.373937909962994e-06,
+      "loss": 1.1115,
+      "step": 13040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.184503587946828,
+      "learning_rate": 9.373143562773858e-06,
+      "loss": 1.1246,
+      "step": 13045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.241634386912347,
+      "learning_rate": 9.372348745671147e-06,
+      "loss": 1.1627,
+      "step": 13050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.767982749391692,
+      "learning_rate": 9.371553458740266e-06,
+      "loss": 1.1461,
+      "step": 13055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.61474824336374,
+      "learning_rate": 9.370757702066671e-06,
+      "loss": 1.1405,
+      "step": 13060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 22.444813504062605,
+      "learning_rate": 9.369961475735874e-06,
+      "loss": 1.0994,
+      "step": 13065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 25.129565243638375,
+      "learning_rate": 9.369164779833429e-06,
+      "loss": 1.1301,
+      "step": 13070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 22.436779647983265,
+      "learning_rate": 9.368367614444947e-06,
+      "loss": 1.155,
+      "step": 13075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.435391066135868,
+      "learning_rate": 9.367569979656084e-06,
+      "loss": 1.0868,
+      "step": 13080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.260614251676454,
+      "learning_rate": 9.366771875552552e-06,
+      "loss": 1.1021,
+      "step": 13085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.837475874790362,
+      "learning_rate": 9.365973302220108e-06,
+      "loss": 1.1082,
+      "step": 13090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.314431233857963,
+      "learning_rate": 9.365174259744568e-06,
+      "loss": 1.0692,
+      "step": 13095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.575194003628091,
+      "learning_rate": 9.364374748211787e-06,
+      "loss": 1.1054,
+      "step": 13100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.383191716857047,
+      "learning_rate": 9.363574767707677e-06,
+      "loss": 1.1237,
+      "step": 13105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.693730546301416,
+      "learning_rate": 9.3627743183182e-06,
+      "loss": 1.1613,
+      "step": 13110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.90265164994131,
+      "learning_rate": 9.36197340012937e-06,
+      "loss": 1.1803,
+      "step": 13115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.667282459197571,
+      "learning_rate": 9.361172013227246e-06,
+      "loss": 1.1383,
+      "step": 13120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.654847477922495,
+      "learning_rate": 9.360370157697942e-06,
+      "loss": 1.1111,
+      "step": 13125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.672817964407618,
+      "learning_rate": 9.359567833627622e-06,
+      "loss": 1.2001,
+      "step": 13130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.34903538280646,
+      "learning_rate": 9.358765041102496e-06,
+      "loss": 1.0858,
+      "step": 13135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.155886734362207,
+      "learning_rate": 9.35796178020883e-06,
+      "loss": 1.1416,
+      "step": 13140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.623987602243092,
+      "learning_rate": 9.357158051032938e-06,
+      "loss": 1.1261,
+      "step": 13145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.447376695700821,
+      "learning_rate": 9.356353853661184e-06,
+      "loss": 1.0647,
+      "step": 13150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.756070117139443,
+      "learning_rate": 9.355549188179982e-06,
+      "loss": 1.1354,
+      "step": 13155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.57267547815171,
+      "learning_rate": 9.354744054675798e-06,
+      "loss": 1.1301,
+      "step": 13160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.313112073726026,
+      "learning_rate": 9.353938453235146e-06,
+      "loss": 1.074,
+      "step": 13165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.403610122509042,
+      "learning_rate": 9.353132383944591e-06,
+      "loss": 1.0868,
+      "step": 13170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.199064665877822,
+      "learning_rate": 9.352325846890751e-06,
+      "loss": 1.1155,
+      "step": 13175
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.66609846236514,
+      "learning_rate": 9.35151884216029e-06,
+      "loss": 1.093,
+      "step": 13180
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.704564594035563,
+      "learning_rate": 9.350711369839925e-06,
+      "loss": 1.1022,
+      "step": 13185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.32147343235014,
+      "learning_rate": 9.34990343001642e-06,
+      "loss": 1.091,
+      "step": 13190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.5295983347581,
+      "learning_rate": 9.349095022776595e-06,
+      "loss": 1.1109,
+      "step": 13195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.339973559241184,
+      "learning_rate": 9.348286148207316e-06,
+      "loss": 1.0998,
+      "step": 13200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.329558581848236,
+      "learning_rate": 9.3474768063955e-06,
+      "loss": 1.0831,
+      "step": 13205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.166574014716085,
+      "learning_rate": 9.346666997428116e-06,
+      "loss": 1.138,
+      "step": 13210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.372868758085875,
+      "learning_rate": 9.345856721392179e-06,
+      "loss": 1.1125,
+      "step": 13215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.684416253674428,
+      "learning_rate": 9.345045978374756e-06,
+      "loss": 1.1334,
+      "step": 13220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 32.350750811462724,
+      "learning_rate": 9.344234768462968e-06,
+      "loss": 1.0502,
+      "step": 13225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 43.16607866911975,
+      "learning_rate": 9.343423091743979e-06,
+      "loss": 1.1193,
+      "step": 13230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.913472380848074,
+      "learning_rate": 9.342610948305013e-06,
+      "loss": 1.1577,
+      "step": 13235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.6992172035777,
+      "learning_rate": 9.341798338233333e-06,
+      "loss": 1.0966,
+      "step": 13240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.881311632675496,
+      "learning_rate": 9.34098526161626e-06,
+      "loss": 1.0964,
+      "step": 13245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.650245010063143,
+      "learning_rate": 9.340171718541163e-06,
+      "loss": 1.1569,
+      "step": 13250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 25.103789602155135,
+      "learning_rate": 9.33935770909546e-06,
+      "loss": 1.1086,
+      "step": 13255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 28.328221175906137,
+      "learning_rate": 9.338543233366621e-06,
+      "loss": 1.1614,
+      "step": 13260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.4071256350902,
+      "learning_rate": 9.337728291442164e-06,
+      "loss": 1.0926,
+      "step": 13265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 26.494290703300035,
+      "learning_rate": 9.336912883409656e-06,
+      "loss": 1.1193,
+      "step": 13270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 47.0797278162397,
+      "learning_rate": 9.336097009356721e-06,
+      "loss": 1.1742,
+      "step": 13275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 30.657108277570504,
+      "learning_rate": 9.335280669371024e-06,
+      "loss": 1.1142,
+      "step": 13280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 29.358195727306434,
+      "learning_rate": 9.334463863540288e-06,
+      "loss": 1.1423,
+      "step": 13285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.693673650046073,
+      "learning_rate": 9.33364659195228e-06,
+      "loss": 1.1196,
+      "step": 13290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.287072127314556,
+      "learning_rate": 9.332828854694817e-06,
+      "loss": 1.1281,
+      "step": 13295
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 40.28895374694088,
+      "learning_rate": 9.332010651855774e-06,
+      "loss": 1.103,
+      "step": 13300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 15.866896750845198,
+      "learning_rate": 9.331191983523067e-06,
+      "loss": 1.1423,
+      "step": 13305
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.218007796622238,
+      "learning_rate": 9.330372849784667e-06,
+      "loss": 1.1288,
+      "step": 13310
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 30.68812414437333,
+      "learning_rate": 9.329553250728592e-06,
+      "loss": 1.1598,
+      "step": 13315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 23.147217760840743,
+      "learning_rate": 9.328733186442914e-06,
+      "loss": 1.1207,
+      "step": 13320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.447397617180021,
+      "learning_rate": 9.32791265701575e-06,
+      "loss": 1.1665,
+      "step": 13325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 21.35296338325948,
+      "learning_rate": 9.32709166253527e-06,
+      "loss": 1.1476,
+      "step": 13330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.52903095366896,
+      "learning_rate": 9.326270203089694e-06,
+      "loss": 1.118,
+      "step": 13335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.953362206083295,
+      "learning_rate": 9.325448278767291e-06,
+      "loss": 1.1535,
+      "step": 13340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.077985539100943,
+      "learning_rate": 9.324625889656383e-06,
+      "loss": 1.1524,
+      "step": 13345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 21.737339439347682,
+      "learning_rate": 9.323803035845333e-06,
+      "loss": 1.0792,
+      "step": 13350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.055998988920857,
+      "learning_rate": 9.322979717422569e-06,
+      "loss": 1.1441,
+      "step": 13355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.448433146731257,
+      "learning_rate": 9.322155934476553e-06,
+      "loss": 1.1197,
+      "step": 13360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.665088564842439,
+      "learning_rate": 9.321331687095806e-06,
+      "loss": 1.066,
+      "step": 13365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 14.559809008757659,
+      "learning_rate": 9.320506975368901e-06,
+      "loss": 1.1375,
+      "step": 13370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 23.296918360592887,
+      "learning_rate": 9.31968179938445e-06,
+      "loss": 1.1388,
+      "step": 13375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.334157795281683,
+      "learning_rate": 9.318856159231128e-06,
+      "loss": 1.155,
+      "step": 13380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.041357722066866,
+      "learning_rate": 9.31803005499765e-06,
+      "loss": 1.1328,
+      "step": 13385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.144806421591245,
+      "learning_rate": 9.317203486772785e-06,
+      "loss": 1.0926,
+      "step": 13390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.373692989715362,
+      "learning_rate": 9.316376454645355e-06,
+      "loss": 1.1577,
+      "step": 13395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.656030314059016,
+      "learning_rate": 9.315548958704223e-06,
+      "loss": 1.0983,
+      "step": 13400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.320948089972093,
+      "learning_rate": 9.31472099903831e-06,
+      "loss": 1.1007,
+      "step": 13405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.862662990973707,
+      "learning_rate": 9.313892575736582e-06,
+      "loss": 1.1155,
+      "step": 13410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.643863414437558,
+      "learning_rate": 9.31306368888806e-06,
+      "loss": 1.1028,
+      "step": 13415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.911997035071135,
+      "learning_rate": 9.312234338581809e-06,
+      "loss": 1.0794,
+      "step": 13420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.640759175515936,
+      "learning_rate": 9.311404524906946e-06,
+      "loss": 1.1046,
+      "step": 13425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.267081615979173,
+      "learning_rate": 9.31057424795264e-06,
+      "loss": 1.1123,
+      "step": 13430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.893566893941383,
+      "learning_rate": 9.309743507808106e-06,
+      "loss": 1.1426,
+      "step": 13435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.313233252311439,
+      "learning_rate": 9.308912304562613e-06,
+      "loss": 1.1425,
+      "step": 13440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.09712988402112,
+      "learning_rate": 9.308080638305478e-06,
+      "loss": 1.1283,
+      "step": 13445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.443032462770006,
+      "learning_rate": 9.30724850912606e-06,
+      "loss": 1.136,
+      "step": 13450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.565472317043675,
+      "learning_rate": 9.306415917113784e-06,
+      "loss": 1.1052,
+      "step": 13455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.141922557312448,
+      "learning_rate": 9.305582862358113e-06,
+      "loss": 1.1289,
+      "step": 13460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.73748140898653,
+      "learning_rate": 9.304749344948559e-06,
+      "loss": 1.1245,
+      "step": 13465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.13130494238641,
+      "learning_rate": 9.303915364974691e-06,
+      "loss": 1.1061,
+      "step": 13470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.860927237913024,
+      "learning_rate": 9.303080922526122e-06,
+      "loss": 1.127,
+      "step": 13475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.545780182973548,
+      "learning_rate": 9.302246017692517e-06,
+      "loss": 1.0708,
+      "step": 13480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.800562981262573,
+      "learning_rate": 9.30141065056359e-06,
+      "loss": 1.1194,
+      "step": 13485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.6279791804018515,
+      "learning_rate": 9.300574821229106e-06,
+      "loss": 1.124,
+      "step": 13490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.023758701531525,
+      "learning_rate": 9.299738529778877e-06,
+      "loss": 1.1083,
+      "step": 13495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.248043305272939,
+      "learning_rate": 9.298901776302765e-06,
+      "loss": 1.1152,
+      "step": 13500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.22039750648539,
+      "learning_rate": 9.298064560890687e-06,
+      "loss": 1.1321,
+      "step": 13505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.509992613535459,
+      "learning_rate": 9.297226883632601e-06,
+      "loss": 1.1747,
+      "step": 13510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.280557450220159,
+      "learning_rate": 9.296388744618524e-06,
+      "loss": 1.1179,
+      "step": 13515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.72380571235983,
+      "learning_rate": 9.295550143938514e-06,
+      "loss": 1.1317,
+      "step": 13520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.9820958892304,
+      "learning_rate": 9.294711081682683e-06,
+      "loss": 1.1744,
+      "step": 13525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.3255339936694694,
+      "learning_rate": 9.293871557941192e-06,
+      "loss": 1.1813,
+      "step": 13530
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.08886677511008,
+      "learning_rate": 9.293031572804252e-06,
+      "loss": 1.0773,
+      "step": 13535
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.925705538683506,
+      "learning_rate": 9.292191126362123e-06,
+      "loss": 1.1472,
+      "step": 13540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.745251977035863,
+      "learning_rate": 9.291350218705114e-06,
+      "loss": 1.1422,
+      "step": 13545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 34.30816442876599,
+      "learning_rate": 9.290508849923585e-06,
+      "loss": 1.1174,
+      "step": 13550
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.887642679729329,
+      "learning_rate": 9.289667020107944e-06,
+      "loss": 1.1387,
+      "step": 13555
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.320860936183017,
+      "learning_rate": 9.288824729348652e-06,
+      "loss": 1.1281,
+      "step": 13560
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.581776526817508,
+      "learning_rate": 9.287981977736214e-06,
+      "loss": 1.1098,
+      "step": 13565
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.520306085948311,
+      "learning_rate": 9.287138765361187e-06,
+      "loss": 1.1389,
+      "step": 13570
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.437870663057927,
+      "learning_rate": 9.28629509231418e-06,
+      "loss": 1.1511,
+      "step": 13575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.003099252983906,
+      "learning_rate": 9.285450958685848e-06,
+      "loss": 1.1098,
+      "step": 13580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.05737331972076,
+      "learning_rate": 9.284606364566898e-06,
+      "loss": 1.1591,
+      "step": 13585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.608738856247651,
+      "learning_rate": 9.283761310048083e-06,
+      "loss": 1.124,
+      "step": 13590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.875720873844369,
+      "learning_rate": 9.282915795220213e-06,
+      "loss": 1.146,
+      "step": 13595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.6993046757266335,
+      "learning_rate": 9.282069820174136e-06,
+      "loss": 1.1302,
+      "step": 13600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.337634709434207,
+      "learning_rate": 9.281223385000759e-06,
+      "loss": 1.1196,
+      "step": 13605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.875538727178148,
+      "learning_rate": 9.280376489791035e-06,
+      "loss": 1.1057,
+      "step": 13610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.116674798519531,
+      "learning_rate": 9.279529134635967e-06,
+      "loss": 1.1533,
+      "step": 13615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.46769866569297,
+      "learning_rate": 9.278681319626606e-06,
+      "loss": 1.1566,
+      "step": 13620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.477793272182669,
+      "learning_rate": 9.277833044854054e-06,
+      "loss": 1.1103,
+      "step": 13625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.816367511849197,
+      "learning_rate": 9.276984310409463e-06,
+      "loss": 1.094,
+      "step": 13630
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.254042338449794,
+      "learning_rate": 9.276135116384031e-06,
+      "loss": 1.1292,
+      "step": 13635
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.687436225032867,
+      "learning_rate": 9.275285462869007e-06,
+      "loss": 1.1342,
+      "step": 13640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.561032557100821,
+      "learning_rate": 9.274435349955694e-06,
+      "loss": 1.1221,
+      "step": 13645
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.062241926944752,
+      "learning_rate": 9.273584777735439e-06,
+      "loss": 1.1429,
+      "step": 13650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.078878765385249,
+      "learning_rate": 9.272733746299636e-06,
+      "loss": 1.1171,
+      "step": 13655
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.379962126254871,
+      "learning_rate": 9.271882255739737e-06,
+      "loss": 1.1258,
+      "step": 13660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.731563499094715,
+      "learning_rate": 9.271030306147235e-06,
+      "loss": 1.1159,
+      "step": 13665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.3783218277511065,
+      "learning_rate": 9.270177897613677e-06,
+      "loss": 1.0826,
+      "step": 13670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.502104811157233,
+      "learning_rate": 9.26932503023066e-06,
+      "loss": 1.1116,
+      "step": 13675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 23.797859875034057,
+      "learning_rate": 9.268471704089825e-06,
+      "loss": 1.1085,
+      "step": 13680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.413733104854044,
+      "learning_rate": 9.267617919282868e-06,
+      "loss": 1.1752,
+      "step": 13685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 27.47484187655042,
+      "learning_rate": 9.26676367590153e-06,
+      "loss": 1.1258,
+      "step": 13690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 20.007395662097164,
+      "learning_rate": 9.265908974037607e-06,
+      "loss": 1.1228,
+      "step": 13695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.950419800963811,
+      "learning_rate": 9.265053813782937e-06,
+      "loss": 1.1377,
+      "step": 13700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.659025159131205,
+      "learning_rate": 9.264198195229408e-06,
+      "loss": 1.0952,
+      "step": 13705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.445367188085125,
+      "learning_rate": 9.263342118468967e-06,
+      "loss": 1.1694,
+      "step": 13710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.967010541997837,
+      "learning_rate": 9.262485583593599e-06,
+      "loss": 1.1389,
+      "step": 13715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 19.276295985618987,
+      "learning_rate": 9.261628590695344e-06,
+      "loss": 1.1292,
+      "step": 13720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.718851162447113,
+      "learning_rate": 9.260771139866287e-06,
+      "loss": 1.1319,
+      "step": 13725
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.207826272962391,
+      "learning_rate": 9.25991323119857e-06,
+      "loss": 1.0926,
+      "step": 13730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.459454509945742,
+      "learning_rate": 9.259054864784373e-06,
+      "loss": 1.128,
+      "step": 13735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.71762436735676,
+      "learning_rate": 9.258196040715935e-06,
+      "loss": 1.1374,
+      "step": 13740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.436272299996114,
+      "learning_rate": 9.25733675908554e-06,
+      "loss": 1.1349,
+      "step": 13745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.991719482585694,
+      "learning_rate": 9.256477019985521e-06,
+      "loss": 1.093,
+      "step": 13750
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.862701413478378,
+      "learning_rate": 9.25561682350826e-06,
+      "loss": 1.0909,
+      "step": 13755
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.538549779012483,
+      "learning_rate": 9.25475616974619e-06,
+      "loss": 1.0974,
+      "step": 13760
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 22.799611630950533,
+      "learning_rate": 9.253895058791794e-06,
+      "loss": 1.1136,
+      "step": 13765
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 17.952476271317316,
+      "learning_rate": 9.253033490737597e-06,
+      "loss": 1.1979,
+      "step": 13770
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.261136565668522,
+      "learning_rate": 9.252171465676183e-06,
+      "loss": 1.1116,
+      "step": 13775
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.966480746423545,
+      "learning_rate": 9.25130898370018e-06,
+      "loss": 1.1217,
+      "step": 13780
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.65468608724836,
+      "learning_rate": 9.250446044902261e-06,
+      "loss": 1.0864,
+      "step": 13785
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.9709406379466605,
+      "learning_rate": 9.249582649375159e-06,
+      "loss": 1.1271,
+      "step": 13790
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.500872689576212,
+      "learning_rate": 9.248718797211644e-06,
+      "loss": 1.143,
+      "step": 13795
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.169700697059949,
+      "learning_rate": 9.247854488504545e-06,
+      "loss": 1.1306,
+      "step": 13800
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.698725865695183,
+      "learning_rate": 9.246989723346733e-06,
+      "loss": 1.0959,
+      "step": 13805
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.749293720715379,
+      "learning_rate": 9.246124501831133e-06,
+      "loss": 1.1566,
+      "step": 13810
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.177725593555664,
+      "learning_rate": 9.245258824050715e-06,
+      "loss": 1.1323,
+      "step": 13815
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.441274968038554,
+      "learning_rate": 9.2443926900985e-06,
+      "loss": 1.1197,
+      "step": 13820
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.58083618966104,
+      "learning_rate": 9.24352610006756e-06,
+      "loss": 1.15,
+      "step": 13825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.292663141819418,
+      "learning_rate": 9.24265905405101e-06,
+      "loss": 1.1437,
+      "step": 13830
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.595715610980198,
+      "learning_rate": 9.241791552142024e-06,
+      "loss": 1.1708,
+      "step": 13835
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.905953138287774,
+      "learning_rate": 9.240923594433812e-06,
+      "loss": 1.097,
+      "step": 13840
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.481455008815603,
+      "learning_rate": 9.240055181019645e-06,
+      "loss": 1.1193,
+      "step": 13845
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.429698145079362,
+      "learning_rate": 9.239186311992834e-06,
+      "loss": 1.1881,
+      "step": 13850
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.162421162324343,
+      "learning_rate": 9.238316987446747e-06,
+      "loss": 1.1377,
+      "step": 13855
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.89095074995895,
+      "learning_rate": 9.237447207474794e-06,
+      "loss": 1.1155,
+      "step": 13860
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.822738067200005,
+      "learning_rate": 9.236576972170437e-06,
+      "loss": 1.145,
+      "step": 13865
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.081231536891007,
+      "learning_rate": 9.235706281627186e-06,
+      "loss": 1.1025,
+      "step": 13870
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.619280728725981,
+      "learning_rate": 9.234835135938602e-06,
+      "loss": 1.157,
+      "step": 13875
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 20.579302852048286,
+      "learning_rate": 9.233963535198293e-06,
+      "loss": 1.1212,
+      "step": 13880
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.05429958005681,
+      "learning_rate": 9.233091479499917e-06,
+      "loss": 1.1027,
+      "step": 13885
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.796214002590123,
+      "learning_rate": 9.232218968937179e-06,
+      "loss": 1.0815,
+      "step": 13890
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.165377012286247,
+      "learning_rate": 9.231346003603834e-06,
+      "loss": 1.0771,
+      "step": 13895
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.4699465575682815,
+      "learning_rate": 9.230472583593689e-06,
+      "loss": 1.1094,
+      "step": 13900
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.452581346147237,
+      "learning_rate": 9.22959870900059e-06,
+      "loss": 1.1625,
+      "step": 13905
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.055197124904744,
+      "learning_rate": 9.228724379918445e-06,
+      "loss": 1.1085,
+      "step": 13910
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.76551117845389,
+      "learning_rate": 9.227849596441205e-06,
+      "loss": 1.1528,
+      "step": 13915
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.930873175564468,
+      "learning_rate": 9.226974358662865e-06,
+      "loss": 1.1737,
+      "step": 13920
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.320446860384759,
+      "learning_rate": 9.226098666677474e-06,
+      "loss": 1.1629,
+      "step": 13925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.781621538328085,
+      "learning_rate": 9.22522252057913e-06,
+      "loss": 1.1138,
+      "step": 13930
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 19.338828347921968,
+      "learning_rate": 9.224345920461982e-06,
+      "loss": 1.1157,
+      "step": 13935
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.016020945129645,
+      "learning_rate": 9.22346886642022e-06,
+      "loss": 1.1395,
+      "step": 13940
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 20.16111275895161,
+      "learning_rate": 9.222591358548087e-06,
+      "loss": 1.167,
+      "step": 13945
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.864772680736904,
+      "learning_rate": 9.221713396939879e-06,
+      "loss": 1.1201,
+      "step": 13950
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 20.342063790210815,
+      "learning_rate": 9.220834981689932e-06,
+      "loss": 1.1418,
+      "step": 13955
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 36.95048013404268,
+      "learning_rate": 9.21995611289264e-06,
+      "loss": 1.1484,
+      "step": 13960
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 26.491214139685393,
+      "learning_rate": 9.219076790642441e-06,
+      "loss": 1.2125,
+      "step": 13965
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.563687511853851,
+      "learning_rate": 9.218197015033818e-06,
+      "loss": 1.1196,
+      "step": 13970
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.26285170276772,
+      "learning_rate": 9.217316786161312e-06,
+      "loss": 1.1208,
+      "step": 13975
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.454411330595025,
+      "learning_rate": 9.216436104119501e-06,
+      "loss": 1.1488,
+      "step": 13980
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.161895232722665,
+      "learning_rate": 9.215554969003026e-06,
+      "loss": 1.1275,
+      "step": 13985
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.091020849494228,
+      "learning_rate": 9.214673380906562e-06,
+      "loss": 1.1399,
+      "step": 13990
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.11861779810707,
+      "learning_rate": 9.213791339924845e-06,
+      "loss": 1.085,
+      "step": 13995
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.965318133568058,
+      "learning_rate": 9.212908846152651e-06,
+      "loss": 1.0959,
+      "step": 14000
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.660287254545757,
+      "learning_rate": 9.212025899684806e-06,
+      "loss": 1.1838,
+      "step": 14005
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.122441198003585,
+      "learning_rate": 9.211142500616193e-06,
+      "loss": 1.1873,
+      "step": 14010
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.188803797134169,
+      "learning_rate": 9.21025864904173e-06,
+      "loss": 1.2069,
+      "step": 14015
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.980304627612265,
+      "learning_rate": 9.209374345056395e-06,
+      "loss": 1.1159,
+      "step": 14020
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 35.89189235415076,
+      "learning_rate": 9.20848958875521e-06,
+      "loss": 1.166,
+      "step": 14025
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.5039723213255,
+      "learning_rate": 9.207604380233243e-06,
+      "loss": 1.1617,
+      "step": 14030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 21.597566500556304,
+      "learning_rate": 9.206718719585618e-06,
+      "loss": 1.1739,
+      "step": 14035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.964009069751182,
+      "learning_rate": 9.205832606907501e-06,
+      "loss": 1.1549,
+      "step": 14040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.820664262938266,
+      "learning_rate": 9.204946042294106e-06,
+      "loss": 1.1297,
+      "step": 14045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.643955951170904,
+      "learning_rate": 9.204059025840704e-06,
+      "loss": 1.1171,
+      "step": 14050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.945312211858417,
+      "learning_rate": 9.203171557642604e-06,
+      "loss": 1.1169,
+      "step": 14055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.930336859532545,
+      "learning_rate": 9.20228363779517e-06,
+      "loss": 1.1336,
+      "step": 14060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.606237171809015,
+      "learning_rate": 9.201395266393813e-06,
+      "loss": 1.1167,
+      "step": 14065
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.94701859813007,
+      "learning_rate": 9.200506443533994e-06,
+      "loss": 1.1178,
+      "step": 14070
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.831832557644484,
+      "learning_rate": 9.199617169311218e-06,
+      "loss": 1.093,
+      "step": 14075
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.041097715615692,
+      "learning_rate": 9.198727443821042e-06,
+      "loss": 1.1481,
+      "step": 14080
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.335247721699924,
+      "learning_rate": 9.197837267159074e-06,
+      "loss": 1.0983,
+      "step": 14085
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 27.95839710913172,
+      "learning_rate": 9.196946639420965e-06,
+      "loss": 1.1165,
+      "step": 14090
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 22.61718909579245,
+      "learning_rate": 9.196055560702414e-06,
+      "loss": 1.107,
+      "step": 14095
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 21.087475547267324,
+      "learning_rate": 9.195164031099177e-06,
+      "loss": 1.1095,
+      "step": 14100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.919048826713127,
+      "learning_rate": 9.19427205070705e-06,
+      "loss": 1.0918,
+      "step": 14105
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 38.24959780739236,
+      "learning_rate": 9.19337961962188e-06,
+      "loss": 1.1554,
+      "step": 14110
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 30.465789912767008,
+      "learning_rate": 9.192486737939565e-06,
+      "loss": 1.144,
+      "step": 14115
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.561028925518176,
+      "learning_rate": 9.191593405756045e-06,
+      "loss": 1.1474,
+      "step": 14120
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.82457678039254,
+      "learning_rate": 9.190699623167315e-06,
+      "loss": 1.1571,
+      "step": 14125
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.485736071645771,
+      "learning_rate": 9.189805390269416e-06,
+      "loss": 1.1314,
+      "step": 14130
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.880076748288326,
+      "learning_rate": 9.188910707158436e-06,
+      "loss": 1.1214,
+      "step": 14135
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.747732571231177,
+      "learning_rate": 9.188015573930513e-06,
+      "loss": 1.1376,
+      "step": 14140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.10280136658978,
+      "learning_rate": 9.187119990681835e-06,
+      "loss": 1.1462,
+      "step": 14145
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.331362335970542,
+      "learning_rate": 9.186223957508634e-06,
+      "loss": 1.106,
+      "step": 14150
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.191655740486233,
+      "learning_rate": 9.185327474507193e-06,
+      "loss": 1.0988,
+      "step": 14155
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.897305841846903,
+      "learning_rate": 9.184430541773845e-06,
+      "loss": 1.1619,
+      "step": 14160
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.539959522238624,
+      "learning_rate": 9.183533159404966e-06,
+      "loss": 1.0901,
+      "step": 14165
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.96237218812107,
+      "learning_rate": 9.182635327496988e-06,
+      "loss": 1.1275,
+      "step": 14170
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 23.810293716473804,
+      "learning_rate": 9.181737046146384e-06,
+      "loss": 1.1698,
+      "step": 14175
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.373524810798113,
+      "learning_rate": 9.180838315449678e-06,
+      "loss": 1.1174,
+      "step": 14180
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.079215045233922,
+      "learning_rate": 9.179939135503444e-06,
+      "loss": 1.1378,
+      "step": 14185
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.353180060990551,
+      "learning_rate": 9.179039506404304e-06,
+      "loss": 1.1267,
+      "step": 14190
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.091622773051276,
+      "learning_rate": 9.178139428248924e-06,
+      "loss": 1.1655,
+      "step": 14195
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.32336553810114,
+      "learning_rate": 9.177238901134022e-06,
+      "loss": 1.0768,
+      "step": 14200
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.730446886045902,
+      "learning_rate": 9.176337925156365e-06,
+      "loss": 1.0966,
+      "step": 14205
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.616189975073862,
+      "learning_rate": 9.175436500412768e-06,
+      "loss": 1.1357,
+      "step": 14210
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.225365663349692,
+      "learning_rate": 9.17453462700009e-06,
+      "loss": 1.1409,
+      "step": 14215
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.40414752329751,
+      "learning_rate": 9.173632305015244e-06,
+      "loss": 1.0925,
+      "step": 14220
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.10721750390549,
+      "learning_rate": 9.172729534555185e-06,
+      "loss": 1.0974,
+      "step": 14225
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.690636915096559,
+      "learning_rate": 9.171826315716923e-06,
+      "loss": 1.1386,
+      "step": 14230
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.255810219626824,
+      "learning_rate": 9.170922648597511e-06,
+      "loss": 1.1288,
+      "step": 14235
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.806463877990772,
+      "learning_rate": 9.170018533294054e-06,
+      "loss": 1.0899,
+      "step": 14240
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.6608942379184795,
+      "learning_rate": 9.169113969903701e-06,
+      "loss": 1.1235,
+      "step": 14245
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.4079556215100535,
+      "learning_rate": 9.168208958523652e-06,
+      "loss": 1.1528,
+      "step": 14250
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.863205442316616,
+      "learning_rate": 9.167303499251158e-06,
+      "loss": 1.1222,
+      "step": 14255
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.518350535844348,
+      "learning_rate": 9.166397592183506e-06,
+      "loss": 1.1012,
+      "step": 14260
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.880158890445731,
+      "learning_rate": 9.165491237418049e-06,
+      "loss": 1.1033,
+      "step": 14265
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.999510079626809,
+      "learning_rate": 9.164584435052173e-06,
+      "loss": 1.1345,
+      "step": 14270
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.9060481237708,
+      "learning_rate": 9.163677185183319e-06,
+      "loss": 1.0951,
+      "step": 14275
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.5116064946818,
+      "learning_rate": 9.162769487908979e-06,
+      "loss": 1.1321,
+      "step": 14280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.889670907161606,
+      "learning_rate": 9.161861343326682e-06,
+      "loss": 1.1357,
+      "step": 14285
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.560013535585883,
+      "learning_rate": 9.160952751534017e-06,
+      "loss": 1.1131,
+      "step": 14290
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.136247411587014,
+      "learning_rate": 9.160043712628617e-06,
+      "loss": 1.1223,
+      "step": 14295
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.823130733668364,
+      "learning_rate": 9.15913422670816e-06,
+      "loss": 1.096,
+      "step": 14300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.246879066563004,
+      "learning_rate": 9.158224293870372e-06,
+      "loss": 1.1279,
+      "step": 14305
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.241964967977525,
+      "learning_rate": 9.157313914213035e-06,
+      "loss": 1.0837,
+      "step": 14310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.153056223550367,
+      "learning_rate": 9.15640308783397e-06,
+      "loss": 1.1464,
+      "step": 14315
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.180470132255973,
+      "learning_rate": 9.155491814831048e-06,
+      "loss": 1.0901,
+      "step": 14320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.214125432403987,
+      "learning_rate": 9.154580095302192e-06,
+      "loss": 1.1685,
+      "step": 14325
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.747395491772629,
+      "learning_rate": 9.15366792934537e-06,
+      "loss": 1.1498,
+      "step": 14330
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.060230650050356,
+      "learning_rate": 9.152755317058595e-06,
+      "loss": 1.0823,
+      "step": 14335
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.522138693026545,
+      "learning_rate": 9.151842258539935e-06,
+      "loss": 1.1511,
+      "step": 14340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.049128238439922,
+      "learning_rate": 9.1509287538875e-06,
+      "loss": 1.1553,
+      "step": 14345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.059989275654193,
+      "learning_rate": 9.150014803199452e-06,
+      "loss": 1.1317,
+      "step": 14350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.197991414325541,
+      "learning_rate": 9.149100406573999e-06,
+      "loss": 1.1415,
+      "step": 14355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.55678018802807,
+      "learning_rate": 9.148185564109395e-06,
+      "loss": 1.1033,
+      "step": 14360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.394148323407906,
+      "learning_rate": 9.147270275903943e-06,
+      "loss": 1.126,
+      "step": 14365
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.931639924562274,
+      "learning_rate": 9.146354542055999e-06,
+      "loss": 1.1332,
+      "step": 14370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.946614747006146,
+      "learning_rate": 9.14543836266396e-06,
+      "loss": 1.1761,
+      "step": 14375
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.851780316902592,
+      "learning_rate": 9.144521737826273e-06,
+      "loss": 1.0815,
+      "step": 14380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.36034929892392,
+      "learning_rate": 9.143604667641434e-06,
+      "loss": 1.1592,
+      "step": 14385
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.50869854422606,
+      "learning_rate": 9.142687152207984e-06,
+      "loss": 1.1145,
+      "step": 14390
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 25.04190733959511,
+      "learning_rate": 9.14176919162452e-06,
+      "loss": 1.1217,
+      "step": 14395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.962800932883541,
+      "learning_rate": 9.140850785989677e-06,
+      "loss": 1.1872,
+      "step": 14400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.229314658841174,
+      "learning_rate": 9.139931935402141e-06,
+      "loss": 1.1138,
+      "step": 14405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.744552444135527,
+      "learning_rate": 9.139012639960649e-06,
+      "loss": 1.1389,
+      "step": 14410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.328030037489963,
+      "learning_rate": 9.13809289976398e-06,
+      "loss": 1.1035,
+      "step": 14415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.766340861757486,
+      "learning_rate": 9.137172714910967e-06,
+      "loss": 1.1192,
+      "step": 14420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.20875420425887,
+      "learning_rate": 9.136252085500488e-06,
+      "loss": 1.1332,
+      "step": 14425
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.330742111571248,
+      "learning_rate": 9.135331011631465e-06,
+      "loss": 1.1166,
+      "step": 14430
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.432026828331053,
+      "learning_rate": 9.134409493402877e-06,
+      "loss": 1.0995,
+      "step": 14435
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.2613884077564546,
+      "learning_rate": 9.13348753091374e-06,
+      "loss": 1.1127,
+      "step": 14440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.055163315915612,
+      "learning_rate": 9.132565124263129e-06,
+      "loss": 1.1122,
+      "step": 14445
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.96554909998541,
+      "learning_rate": 9.131642273550155e-06,
+      "loss": 1.1098,
+      "step": 14450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.056357541722985,
+      "learning_rate": 9.130718978873983e-06,
+      "loss": 1.096,
+      "step": 14455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.244625753913043,
+      "learning_rate": 9.129795240333828e-06,
+      "loss": 1.07,
+      "step": 14460
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.237572939386469,
+      "learning_rate": 9.128871058028947e-06,
+      "loss": 1.0604,
+      "step": 14465
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.58110063273889,
+      "learning_rate": 9.12794643205865e-06,
+      "loss": 1.1535,
+      "step": 14470
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 26.196714509958802,
+      "learning_rate": 9.12702136252229e-06,
+      "loss": 1.1287,
+      "step": 14475
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.609119541234472,
+      "learning_rate": 9.126095849519272e-06,
+      "loss": 1.163,
+      "step": 14480
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.672169768287137,
+      "learning_rate": 9.125169893149043e-06,
+      "loss": 1.0958,
+      "step": 14485
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.348789296891935,
+      "learning_rate": 9.124243493511105e-06,
+      "loss": 1.062,
+      "step": 14490
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.278743464736229,
+      "learning_rate": 9.123316650705001e-06,
+      "loss": 1.1346,
+      "step": 14495
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.196639817099412,
+      "learning_rate": 9.122389364830327e-06,
+      "loss": 1.0966,
+      "step": 14500
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.767105478740581,
+      "learning_rate": 9.121461635986719e-06,
+      "loss": 1.0693,
+      "step": 14505
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 16.26256420492267,
+      "learning_rate": 9.120533464273872e-06,
+      "loss": 1.117,
+      "step": 14510
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.309338429855874,
+      "learning_rate": 9.119604849791518e-06,
+      "loss": 1.1151,
+      "step": 14515
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.115462869171166,
+      "learning_rate": 9.11867579263944e-06,
+      "loss": 1.1576,
+      "step": 14520
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.463568967606593,
+      "learning_rate": 9.117746292917473e-06,
+      "loss": 1.0641,
+      "step": 14525
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.1024608003889,
+      "learning_rate": 9.116816350725492e-06,
+      "loss": 1.0755,
+      "step": 14530
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.427344289970103,
+      "learning_rate": 9.115885966163426e-06,
+      "loss": 1.1344,
+      "step": 14535
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.83017031327818,
+      "learning_rate": 9.11495513933125e-06,
+      "loss": 1.1203,
+      "step": 14540
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 19.718512738635773,
+      "learning_rate": 9.114023870328982e-06,
+      "loss": 1.1189,
+      "step": 14545
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.840400431357369,
+      "learning_rate": 9.113092159256694e-06,
+      "loss": 1.1485,
+      "step": 14550
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.694466414197182,
+      "learning_rate": 9.112160006214499e-06,
+      "loss": 1.1119,
+      "step": 14555
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.36702082978927,
+      "learning_rate": 9.111227411302564e-06,
+      "loss": 1.1151,
+      "step": 14560
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.742012490451101,
+      "learning_rate": 9.110294374621099e-06,
+      "loss": 1.1402,
+      "step": 14565
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.793883486862604,
+      "learning_rate": 9.109360896270363e-06,
+      "loss": 1.1191,
+      "step": 14570
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.332386305768537,
+      "learning_rate": 9.108426976350664e-06,
+      "loss": 1.1635,
+      "step": 14575
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.582900024472447,
+      "learning_rate": 9.107492614962352e-06,
+      "loss": 1.0885,
+      "step": 14580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.90860873158703,
+      "learning_rate": 9.106557812205833e-06,
+      "loss": 1.1498,
+      "step": 14585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.122243577764641,
+      "learning_rate": 9.105622568181554e-06,
+      "loss": 1.1213,
+      "step": 14590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.163872220241011,
+      "learning_rate": 9.104686882990009e-06,
+      "loss": 1.0777,
+      "step": 14595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.198994160118266,
+      "learning_rate": 9.103750756731743e-06,
+      "loss": 1.0947,
+      "step": 14600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.267489683648388,
+      "learning_rate": 9.102814189507347e-06,
+      "loss": 1.1456,
+      "step": 14605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.229897114218584,
+      "learning_rate": 9.10187718141746e-06,
+      "loss": 1.0682,
+      "step": 14610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.801560161867796,
+      "learning_rate": 9.100939732562766e-06,
+      "loss": 1.1285,
+      "step": 14615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.02997697852096,
+      "learning_rate": 9.100001843044e-06,
+      "loss": 1.1209,
+      "step": 14620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.329984777855054,
+      "learning_rate": 9.099063512961941e-06,
+      "loss": 1.1269,
+      "step": 14625
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.538026072721141,
+      "learning_rate": 9.098124742417417e-06,
+      "loss": 1.1004,
+      "step": 14630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.725961038889572,
+      "learning_rate": 9.097185531511303e-06,
+      "loss": 1.1349,
+      "step": 14635
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.842410136833655,
+      "learning_rate": 9.096245880344521e-06,
+      "loss": 1.11,
+      "step": 14640
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.904407469255224,
+      "learning_rate": 9.09530578901804e-06,
+      "loss": 1.1297,
+      "step": 14645
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.895722514745561,
+      "learning_rate": 9.09436525763288e-06,
+      "loss": 1.1291,
+      "step": 14650
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.609008054231528,
+      "learning_rate": 9.093424286290103e-06,
+      "loss": 1.0738,
+      "step": 14655
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 28.32829868657577,
+      "learning_rate": 9.09248287509082e-06,
+      "loss": 1.1297,
+      "step": 14660
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.456125814199051,
+      "learning_rate": 9.091541024136192e-06,
+      "loss": 1.1334,
+      "step": 14665
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.342499478758848,
+      "learning_rate": 9.090598733527422e-06,
+      "loss": 1.0919,
+      "step": 14670
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.794926518998487,
+      "learning_rate": 9.089656003365765e-06,
+      "loss": 1.1409,
+      "step": 14675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.01187948942198,
+      "learning_rate": 9.088712833752522e-06,
+      "loss": 1.1168,
+      "step": 14680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.358538361911016,
+      "learning_rate": 9.08776922478904e-06,
+      "loss": 1.0703,
+      "step": 14685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.303843672961618,
+      "learning_rate": 9.086825176576714e-06,
+      "loss": 1.1206,
+      "step": 14690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.090847186154365,
+      "learning_rate": 9.085880689216986e-06,
+      "loss": 1.0842,
+      "step": 14695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.845417577175287,
+      "learning_rate": 9.084935762811346e-06,
+      "loss": 1.0998,
+      "step": 14700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.935343897262458,
+      "learning_rate": 9.08399039746133e-06,
+      "loss": 1.1412,
+      "step": 14705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 27.903758243616746,
+      "learning_rate": 9.083044593268522e-06,
+      "loss": 1.1039,
+      "step": 14710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.379569474444917,
+      "learning_rate": 9.082098350334552e-06,
+      "loss": 1.1555,
+      "step": 14715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.374143094036183,
+      "learning_rate": 9.081151668761097e-06,
+      "loss": 1.1063,
+      "step": 14720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 26.567546670372046,
+      "learning_rate": 9.080204548649887e-06,
+      "loss": 1.1441,
+      "step": 14725
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 17.5783194755378,
+      "learning_rate": 9.07925699010269e-06,
+      "loss": 1.1372,
+      "step": 14730
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.64054869897644,
+      "learning_rate": 9.078308993221325e-06,
+      "loss": 1.0967,
+      "step": 14735
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 36.904593891318484,
+      "learning_rate": 9.07736055810766e-06,
+      "loss": 1.1641,
+      "step": 14740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 60.90044606885157,
+      "learning_rate": 9.076411684863609e-06,
+      "loss": 1.2104,
+      "step": 14745
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 31.515403806869745,
+      "learning_rate": 9.075462373591131e-06,
+      "loss": 1.1401,
+      "step": 14750
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.527393229252933,
+      "learning_rate": 9.074512624392237e-06,
+      "loss": 1.1814,
+      "step": 14755
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 25.559324892237377,
+      "learning_rate": 9.073562437368978e-06,
+      "loss": 1.2215,
+      "step": 14760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.17703256392607,
+      "learning_rate": 9.072611812623455e-06,
+      "loss": 1.114,
+      "step": 14765
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.62022547087072,
+      "learning_rate": 9.07166075025782e-06,
+      "loss": 1.1284,
+      "step": 14770
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 47.64907967284982,
+      "learning_rate": 9.07070925037427e-06,
+      "loss": 1.0944,
+      "step": 14775
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.732313223783803,
+      "learning_rate": 9.069757313075042e-06,
+      "loss": 1.1701,
+      "step": 14780
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 41.91910561016749,
+      "learning_rate": 9.068804938462432e-06,
+      "loss": 1.166,
+      "step": 14785
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 28.458670265029813,
+      "learning_rate": 9.067852126638775e-06,
+      "loss": 1.0961,
+      "step": 14790
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 43.422055609801596,
+      "learning_rate": 9.066898877706452e-06,
+      "loss": 1.1488,
+      "step": 14795
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 45.83463846175388,
+      "learning_rate": 9.065945191767897e-06,
+      "loss": 1.143,
+      "step": 14800
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.366062323915122,
+      "learning_rate": 9.064991068925587e-06,
+      "loss": 1.1366,
+      "step": 14805
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 48.11725412910709,
+      "learning_rate": 9.064036509282046e-06,
+      "loss": 1.1632,
+      "step": 14810
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.427492522436365,
+      "learning_rate": 9.063081512939847e-06,
+      "loss": 1.1133,
+      "step": 14815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 29.720651040287922,
+      "learning_rate": 9.062126080001608e-06,
+      "loss": 1.0939,
+      "step": 14820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 23.438743515526202,
+      "learning_rate": 9.061170210569992e-06,
+      "loss": 1.159,
+      "step": 14825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 24.303463946418535,
+      "learning_rate": 9.060213904747715e-06,
+      "loss": 1.1413,
+      "step": 14830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 39.2810075255219,
+      "learning_rate": 9.059257162637535e-06,
+      "loss": 1.1823,
+      "step": 14835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.22613215485907,
+      "learning_rate": 9.058299984342258e-06,
+      "loss": 1.1167,
+      "step": 14840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.6565528713030995,
+      "learning_rate": 9.057342369964737e-06,
+      "loss": 1.122,
+      "step": 14845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.097227707630147,
+      "learning_rate": 9.05638431960787e-06,
+      "loss": 1.1826,
+      "step": 14850
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.724964176778249,
+      "learning_rate": 9.055425833374609e-06,
+      "loss": 1.122,
+      "step": 14855
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.054356700617171,
+      "learning_rate": 9.05446691136794e-06,
+      "loss": 1.1403,
+      "step": 14860
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.170345265610834,
+      "learning_rate": 9.053507553690912e-06,
+      "loss": 1.1732,
+      "step": 14865
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.123171957572076,
+      "learning_rate": 9.052547760446606e-06,
+      "loss": 1.1322,
+      "step": 14870
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.312985123014363,
+      "learning_rate": 9.051587531738158e-06,
+      "loss": 1.1324,
+      "step": 14875
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.153623725885799,
+      "learning_rate": 9.050626867668747e-06,
+      "loss": 1.0854,
+      "step": 14880
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.212587392556209,
+      "learning_rate": 9.049665768341604e-06,
+      "loss": 1.1518,
+      "step": 14885
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.601145005621945,
+      "learning_rate": 9.048704233860003e-06,
+      "loss": 1.1126,
+      "step": 14890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.737501036088073,
+      "learning_rate": 9.047742264327261e-06,
+      "loss": 1.1573,
+      "step": 14895
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.44119029656582,
+      "learning_rate": 9.04677985984675e-06,
+      "loss": 1.0828,
+      "step": 14900
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.019495679200885,
+      "learning_rate": 9.045817020521884e-06,
+      "loss": 1.0922,
+      "step": 14905
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.396485879866313,
+      "learning_rate": 9.044853746456123e-06,
+      "loss": 1.1129,
+      "step": 14910
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.985975196367451,
+      "learning_rate": 9.043890037752976e-06,
+      "loss": 1.1021,
+      "step": 14915
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.738204524180604,
+      "learning_rate": 9.042925894515998e-06,
+      "loss": 1.1154,
+      "step": 14920
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.748101190111107,
+      "learning_rate": 9.041961316848787e-06,
+      "loss": 1.089,
+      "step": 14925
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.73538533372155,
+      "learning_rate": 9.040996304854997e-06,
+      "loss": 1.1355,
+      "step": 14930
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.113794638447425,
+      "learning_rate": 9.040030858638319e-06,
+      "loss": 1.1045,
+      "step": 14935
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.7434952780550805,
+      "learning_rate": 9.039064978302493e-06,
+      "loss": 1.1086,
+      "step": 14940
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 15.603301122450585,
+      "learning_rate": 9.038098663951311e-06,
+      "loss": 1.1274,
+      "step": 14945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.757240851222182,
+      "learning_rate": 9.037131915688607e-06,
+      "loss": 1.1293,
+      "step": 14950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 31.923350278535334,
+      "learning_rate": 9.03616473361826e-06,
+      "loss": 1.1706,
+      "step": 14955
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.82388837537516,
+      "learning_rate": 9.035197117844199e-06,
+      "loss": 1.1331,
+      "step": 14960
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.319920335950254,
+      "learning_rate": 9.0342290684704e-06,
+      "loss": 1.1461,
+      "step": 14965
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 54.85723431684405,
+      "learning_rate": 9.033260585600883e-06,
+      "loss": 1.1265,
+      "step": 14970
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 28.824414345469417,
+      "learning_rate": 9.032291669339716e-06,
+      "loss": 1.104,
+      "step": 14975
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.40871623293331,
+      "learning_rate": 9.031322319791014e-06,
+      "loss": 1.1264,
+      "step": 14980
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.612526593600604,
+      "learning_rate": 9.030352537058936e-06,
+      "loss": 1.1385,
+      "step": 14985
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.914087751351797,
+      "learning_rate": 9.02938232124769e-06,
+      "loss": 1.1049,
+      "step": 14990
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.266647447125317,
+      "learning_rate": 9.028411672461532e-06,
+      "loss": 1.1158,
+      "step": 14995
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.04429399058661,
+      "learning_rate": 9.027440590804761e-06,
+      "loss": 1.1337,
+      "step": 15000
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.948380414885566,
+      "learning_rate": 9.026469076381724e-06,
+      "loss": 1.1182,
+      "step": 15005
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.574202684687753,
+      "learning_rate": 9.025497129296813e-06,
+      "loss": 1.1119,
+      "step": 15010
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.087659409092865,
+      "learning_rate": 9.024524749654472e-06,
+      "loss": 1.1082,
+      "step": 15015
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.427771613553315,
+      "learning_rate": 9.023551937559186e-06,
+      "loss": 1.1108,
+      "step": 15020
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.490019223859416,
+      "learning_rate": 9.022578693115484e-06,
+      "loss": 1.1397,
+      "step": 15025
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 15.256419026868922,
+      "learning_rate": 9.021605016427949e-06,
+      "loss": 1.127,
+      "step": 15030
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.452780712753931,
+      "learning_rate": 9.020630907601207e-06,
+      "loss": 1.0729,
+      "step": 15035
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.4594919433949745,
+      "learning_rate": 9.019656366739929e-06,
+      "loss": 1.0662,
+      "step": 15040
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 19.75808217651537,
+      "learning_rate": 9.018681393948836e-06,
+      "loss": 1.008,
+      "step": 15045
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.744640468159116,
+      "learning_rate": 9.017705989332687e-06,
+      "loss": 1.0693,
+      "step": 15050
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.8766407741970905,
+      "learning_rate": 9.016730152996302e-06,
+      "loss": 1.1441,
+      "step": 15055
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.488179358832632,
+      "learning_rate": 9.015753885044534e-06,
+      "loss": 1.1546,
+      "step": 15060
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.268097923372006,
+      "learning_rate": 9.014777185582289e-06,
+      "loss": 1.0512,
+      "step": 15065
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.656687178551373,
+      "learning_rate": 9.013800054714515e-06,
+      "loss": 1.1458,
+      "step": 15070
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.970250634168265,
+      "learning_rate": 9.012822492546211e-06,
+      "loss": 1.1421,
+      "step": 15075
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.541114823315593,
+      "learning_rate": 9.011844499182423e-06,
+      "loss": 1.1049,
+      "step": 15080
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.162148932902298,
+      "learning_rate": 9.010866074728236e-06,
+      "loss": 1.1292,
+      "step": 15085
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.487928112419654,
+      "learning_rate": 9.009887219288786e-06,
+      "loss": 1.1189,
+      "step": 15090
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.42430035428869,
+      "learning_rate": 9.008907932969261e-06,
+      "loss": 1.1098,
+      "step": 15095
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.338598817675365,
+      "learning_rate": 9.007928215874884e-06,
+      "loss": 1.1538,
+      "step": 15100
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.13543336542322,
+      "learning_rate": 9.00694806811093e-06,
+      "loss": 1.1179,
+      "step": 15105
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.009629898302627,
+      "learning_rate": 9.005967489782725e-06,
+      "loss": 1.0913,
+      "step": 15110
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.163157815657705,
+      "learning_rate": 9.004986480995634e-06,
+      "loss": 1.0746,
+      "step": 15115
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.398435618814801,
+      "learning_rate": 9.004005041855068e-06,
+      "loss": 1.1442,
+      "step": 15120
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.309657495107698,
+      "learning_rate": 9.003023172466489e-06,
+      "loss": 1.1605,
+      "step": 15125
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.444696379623364,
+      "learning_rate": 9.002040872935405e-06,
+      "loss": 1.1448,
+      "step": 15130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.951662721228207,
+      "learning_rate": 9.001058143367368e-06,
+      "loss": 1.1309,
+      "step": 15135
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.934353463404143,
+      "learning_rate": 9.000074983867973e-06,
+      "loss": 1.095,
+      "step": 15140
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.996296854973321,
+      "learning_rate": 8.999091394542868e-06,
+      "loss": 1.1453,
+      "step": 15145
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.555201313261598,
+      "learning_rate": 8.998107375497744e-06,
+      "loss": 1.1219,
+      "step": 15150
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.4527144214470615,
+      "learning_rate": 8.997122926838337e-06,
+      "loss": 1.1429,
+      "step": 15155
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.053857604890467,
+      "learning_rate": 8.996138048670431e-06,
+      "loss": 1.1013,
+      "step": 15160
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.584512239337207,
+      "learning_rate": 8.995152741099854e-06,
+      "loss": 1.0783,
+      "step": 15165
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.359040859756247,
+      "learning_rate": 8.994167004232486e-06,
+      "loss": 1.158,
+      "step": 15170
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 17.558674170066805,
+      "learning_rate": 8.993180838174242e-06,
+      "loss": 1.1502,
+      "step": 15175
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.148872725818038,
+      "learning_rate": 8.992194243031097e-06,
+      "loss": 1.147,
+      "step": 15180
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.65830754136846,
+      "learning_rate": 8.99120721890906e-06,
+      "loss": 1.1536,
+      "step": 15185
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.295536196796188,
+      "learning_rate": 8.990219765914194e-06,
+      "loss": 1.1444,
+      "step": 15190
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 19.573950215677538,
+      "learning_rate": 8.989231884152603e-06,
+      "loss": 1.0894,
+      "step": 15195
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.424215213013229,
+      "learning_rate": 8.988243573730439e-06,
+      "loss": 1.133,
+      "step": 15200
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 40.37914304337508,
+      "learning_rate": 8.987254834753904e-06,
+      "loss": 1.1227,
+      "step": 15205
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 27.373384526939393,
+      "learning_rate": 8.986265667329239e-06,
+      "loss": 1.1093,
+      "step": 15210
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.569928707365623,
+      "learning_rate": 8.985276071562737e-06,
+      "loss": 1.1085,
+      "step": 15215
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.893884643105887,
+      "learning_rate": 8.984286047560729e-06,
+      "loss": 1.0573,
+      "step": 15220
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.084049740281538,
+      "learning_rate": 8.983295595429604e-06,
+      "loss": 1.1941,
+      "step": 15225
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.697120379518738,
+      "learning_rate": 8.982304715275788e-06,
+      "loss": 1.1134,
+      "step": 15230
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.153596659350104,
+      "learning_rate": 8.981313407205752e-06,
+      "loss": 1.1232,
+      "step": 15235
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.339287903534414,
+      "learning_rate": 8.980321671326022e-06,
+      "loss": 1.1423,
+      "step": 15240
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.220683732804061,
+      "learning_rate": 8.979329507743161e-06,
+      "loss": 1.0984,
+      "step": 15245
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.758846506839011,
+      "learning_rate": 8.978336916563783e-06,
+      "loss": 1.1382,
+      "step": 15250
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.10878136536204,
+      "learning_rate": 8.977343897894545e-06,
+      "loss": 1.1074,
+      "step": 15255
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.904917585744883,
+      "learning_rate": 8.976350451842152e-06,
+      "loss": 1.1039,
+      "step": 15260
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.953932824290469,
+      "learning_rate": 8.975356578513354e-06,
+      "loss": 1.1082,
+      "step": 15265
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.41885249619772,
+      "learning_rate": 8.974362278014946e-06,
+      "loss": 1.1211,
+      "step": 15270
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.8002738221505314,
+      "learning_rate": 8.973367550453774e-06,
+      "loss": 1.1127,
+      "step": 15275
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.206878051734687,
+      "learning_rate": 8.972372395936719e-06,
+      "loss": 1.1134,
+      "step": 15280
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.087959244254456,
+      "learning_rate": 8.97137681457072e-06,
+      "loss": 1.0968,
+      "step": 15285
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.415553792258701,
+      "learning_rate": 8.970380806462758e-06,
+      "loss": 1.1231,
+      "step": 15290
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.118829201084383,
+      "learning_rate": 8.969384371719852e-06,
+      "loss": 1.1112,
+      "step": 15295
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.928779573226809,
+      "learning_rate": 8.96838751044908e-06,
+      "loss": 1.0993,
+      "step": 15300
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 21.279270627251556,
+      "learning_rate": 8.967390222757553e-06,
+      "loss": 1.1204,
+      "step": 15305
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.733553955905094,
+      "learning_rate": 8.96639250875244e-06,
+      "loss": 1.1068,
+      "step": 15310
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 23.863312827531992,
+      "learning_rate": 8.965394368540945e-06,
+      "loss": 1.2045,
+      "step": 15315
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.54916337636012,
+      "learning_rate": 8.964395802230325e-06,
+      "loss": 1.1303,
+      "step": 15320
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.200584650921166,
+      "learning_rate": 8.96339680992788e-06,
+      "loss": 1.0871,
+      "step": 15325
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.0581056774615,
+      "learning_rate": 8.962397391740957e-06,
+      "loss": 1.1269,
+      "step": 15330
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 19.354208687587455,
+      "learning_rate": 8.961397547776946e-06,
+      "loss": 1.0688,
+      "step": 15335
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.372140860900933,
+      "learning_rate": 8.960397278143285e-06,
+      "loss": 1.0811,
+      "step": 15340
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.8651829016830215,
+      "learning_rate": 8.959396582947458e-06,
+      "loss": 1.1043,
+      "step": 15345
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 30.30903429132215,
+      "learning_rate": 8.958395462296995e-06,
+      "loss": 1.1203,
+      "step": 15350
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 23.306133502411026,
+      "learning_rate": 8.957393916299469e-06,
+      "loss": 1.0805,
+      "step": 15355
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.661451079673935,
+      "learning_rate": 8.956391945062501e-06,
+      "loss": 1.1396,
+      "step": 15360
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.86938027595896,
+      "learning_rate": 8.955389548693758e-06,
+      "loss": 1.1358,
+      "step": 15365
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 28.80689046818295,
+      "learning_rate": 8.954386727300953e-06,
+      "loss": 1.1024,
+      "step": 15370
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 29.535898479193374,
+      "learning_rate": 8.95338348099184e-06,
+      "loss": 1.17,
+      "step": 15375
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.511670803697672,
+      "learning_rate": 8.952379809874225e-06,
+      "loss": 1.1174,
+      "step": 15380
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.03831406520542,
+      "learning_rate": 8.951375714055956e-06,
+      "loss": 1.1425,
+      "step": 15385
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.170125157229293,
+      "learning_rate": 8.950371193644929e-06,
+      "loss": 1.0766,
+      "step": 15390
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.275570143929595,
+      "learning_rate": 8.949366248749084e-06,
+      "loss": 1.1446,
+      "step": 15395
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.146976157540005,
+      "learning_rate": 8.9483608794764e-06,
+      "loss": 1.1326,
+      "step": 15400
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.842079211086604,
+      "learning_rate": 8.947355085934918e-06,
+      "loss": 1.1216,
+      "step": 15405
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.185294680167168,
+      "learning_rate": 8.946348868232712e-06,
+      "loss": 1.1124,
+      "step": 15410
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.770946193843225,
+      "learning_rate": 8.945342226477903e-06,
+      "loss": 1.0472,
+      "step": 15415
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.094968685801851,
+      "learning_rate": 8.944335160778658e-06,
+      "loss": 1.1096,
+      "step": 15420
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 36.008993846535525,
+      "learning_rate": 8.943327671243194e-06,
+      "loss": 1.1069,
+      "step": 15425
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.102908445444992,
+      "learning_rate": 8.942319757979768e-06,
+      "loss": 1.0827,
+      "step": 15430
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 22.180074968865192,
+      "learning_rate": 8.941311421096686e-06,
+      "loss": 1.1042,
+      "step": 15435
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 25.342355504685877,
+      "learning_rate": 8.940302660702296e-06,
+      "loss": 1.1036,
+      "step": 15440
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.207078580665375,
+      "learning_rate": 8.939293476904996e-06,
+      "loss": 1.1118,
+      "step": 15445
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.961212029304685,
+      "learning_rate": 8.938283869813226e-06,
+      "loss": 1.0724,
+      "step": 15450
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 22.213912907198274,
+      "learning_rate": 8.937273839535474e-06,
+      "loss": 1.1287,
+      "step": 15455
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.205395591758585,
+      "learning_rate": 8.936263386180271e-06,
+      "loss": 1.105,
+      "step": 15460
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.466101834459902,
+      "learning_rate": 8.935252509856193e-06,
+      "loss": 1.1461,
+      "step": 15465
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.685012039390134,
+      "learning_rate": 8.934241210671868e-06,
+      "loss": 1.1156,
+      "step": 15470
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.445347741715205,
+      "learning_rate": 8.933229488735961e-06,
+      "loss": 1.1005,
+      "step": 15475
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.377434775353599,
+      "learning_rate": 8.932217344157188e-06,
+      "loss": 1.1346,
+      "step": 15480
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.217632836077673,
+      "learning_rate": 8.931204777044305e-06,
+      "loss": 1.1278,
+      "step": 15485
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.896865401653006,
+      "learning_rate": 8.93019178750612e-06,
+      "loss": 1.1265,
+      "step": 15490
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.429060194190058,
+      "learning_rate": 8.929178375651481e-06,
+      "loss": 1.084,
+      "step": 15495
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.56996667332036,
+      "learning_rate": 8.928164541589287e-06,
+      "loss": 1.1089,
+      "step": 15500
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.494176421990351,
+      "learning_rate": 8.927150285428475e-06,
+      "loss": 1.1051,
+      "step": 15505
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.920264190839941,
+      "learning_rate": 8.926135607278035e-06,
+      "loss": 1.1011,
+      "step": 15510
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.810185229857796,
+      "learning_rate": 8.925120507246993e-06,
+      "loss": 1.1335,
+      "step": 15515
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.242383280864297,
+      "learning_rate": 8.924104985444434e-06,
+      "loss": 1.0933,
+      "step": 15520
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.09670297381289,
+      "learning_rate": 8.923089041979474e-06,
+      "loss": 1.09,
+      "step": 15525
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.141979374923959,
+      "learning_rate": 8.922072676961282e-06,
+      "loss": 1.1145,
+      "step": 15530
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.114749823235573,
+      "learning_rate": 8.92105589049907e-06,
+      "loss": 1.1569,
+      "step": 15535
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.534260385387185,
+      "learning_rate": 8.9200386827021e-06,
+      "loss": 1.1126,
+      "step": 15540
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.362250458323114,
+      "learning_rate": 8.919021053679672e-06,
+      "loss": 1.1224,
+      "step": 15545
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.309395112552444,
+      "learning_rate": 8.918003003541136e-06,
+      "loss": 1.0922,
+      "step": 15550
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 22.45995518545166,
+      "learning_rate": 8.916984532395887e-06,
+      "loss": 1.1151,
+      "step": 15555
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.799846626012507,
+      "learning_rate": 8.91596564035336e-06,
+      "loss": 1.1321,
+      "step": 15560
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.819301489285724,
+      "learning_rate": 8.914946327523047e-06,
+      "loss": 1.1421,
+      "step": 15565
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.37561174852505,
+      "learning_rate": 8.91392659401447e-06,
+      "loss": 1.1423,
+      "step": 15570
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.8702545885999955,
+      "learning_rate": 8.912906439937207e-06,
+      "loss": 1.1309,
+      "step": 15575
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.042355847210189,
+      "learning_rate": 8.91188586540088e-06,
+      "loss": 1.0924,
+      "step": 15580
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.682587939526758,
+      "learning_rate": 8.91086487051515e-06,
+      "loss": 1.1059,
+      "step": 15585
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.871931530926953,
+      "learning_rate": 8.90984345538973e-06,
+      "loss": 1.1232,
+      "step": 15590
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.112442975979869,
+      "learning_rate": 8.908821620134377e-06,
+      "loss": 1.0661,
+      "step": 15595
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.912873380119569,
+      "learning_rate": 8.907799364858888e-06,
+      "loss": 1.1066,
+      "step": 15600
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.72976963135992,
+      "learning_rate": 8.906776689673113e-06,
+      "loss": 1.0747,
+      "step": 15605
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.297488178624966,
+      "learning_rate": 8.905753594686938e-06,
+      "loss": 1.0851,
+      "step": 15610
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.769513972727212,
+      "learning_rate": 8.904730080010304e-06,
+      "loss": 1.155,
+      "step": 15615
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.209314719264187,
+      "learning_rate": 8.90370614575319e-06,
+      "loss": 1.1069,
+      "step": 15620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.869716104495888,
+      "learning_rate": 8.902681792025621e-06,
+      "loss": 1.0908,
+      "step": 15625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.494067016564845,
+      "learning_rate": 8.90165701893767e-06,
+      "loss": 1.1306,
+      "step": 15630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.616906490157792,
+      "learning_rate": 8.900631826599454e-06,
+      "loss": 1.1103,
+      "step": 15635
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.4077577363818,
+      "learning_rate": 8.899606215121132e-06,
+      "loss": 1.1245,
+      "step": 15640
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.634824470674596,
+      "learning_rate": 8.898580184612913e-06,
+      "loss": 1.1558,
+      "step": 15645
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.699519493918952,
+      "learning_rate": 8.897553735185048e-06,
+      "loss": 1.1039,
+      "step": 15650
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.862963576619954,
+      "learning_rate": 8.89652686694783e-06,
+      "loss": 1.0783,
+      "step": 15655
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 22.144900784416958,
+      "learning_rate": 8.895499580011606e-06,
+      "loss": 1.1025,
+      "step": 15660
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 23.832881740551286,
+      "learning_rate": 8.894471874486759e-06,
+      "loss": 1.0998,
+      "step": 15665
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.5828596885912,
+      "learning_rate": 8.893443750483722e-06,
+      "loss": 1.1212,
+      "step": 15670
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 38.823298467421445,
+      "learning_rate": 8.89241520811297e-06,
+      "loss": 1.1026,
+      "step": 15675
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 43.279108304643266,
+      "learning_rate": 8.891386247485027e-06,
+      "loss": 1.1717,
+      "step": 15680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 23.204916311755625,
+      "learning_rate": 8.890356868710457e-06,
+      "loss": 1.16,
+      "step": 15685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.237581085215133,
+      "learning_rate": 8.889327071899872e-06,
+      "loss": 1.1444,
+      "step": 15690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.03368700824802,
+      "learning_rate": 8.888296857163927e-06,
+      "loss": 1.1257,
+      "step": 15695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 48.306439237789114,
+      "learning_rate": 8.887266224613326e-06,
+      "loss": 1.11,
+      "step": 15700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.132291292277202,
+      "learning_rate": 8.886235174358812e-06,
+      "loss": 1.1581,
+      "step": 15705
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 57.65949417291116,
+      "learning_rate": 8.885203706511178e-06,
+      "loss": 1.134,
+      "step": 15710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 19.569696607607145,
+      "learning_rate": 8.88417182118126e-06,
+      "loss": 1.0934,
+      "step": 15715
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 27.037448824784263,
+      "learning_rate": 8.883139518479936e-06,
+      "loss": 1.1227,
+      "step": 15720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 35.81651807981647,
+      "learning_rate": 8.882106798518135e-06,
+      "loss": 1.1356,
+      "step": 15725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.657523340271581,
+      "learning_rate": 8.881073661406822e-06,
+      "loss": 1.1112,
+      "step": 15730
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.151989068657166,
+      "learning_rate": 8.88004010725702e-06,
+      "loss": 1.1329,
+      "step": 15735
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.62975572551237,
+      "learning_rate": 8.879006136179781e-06,
+      "loss": 1.1044,
+      "step": 15740
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.057841844382403,
+      "learning_rate": 8.877971748286215e-06,
+      "loss": 1.0976,
+      "step": 15745
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.280863510880357,
+      "learning_rate": 8.87693694368747e-06,
+      "loss": 1.1132,
+      "step": 15750
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.365025414091182,
+      "learning_rate": 8.87590172249474e-06,
+      "loss": 1.1198,
+      "step": 15755
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 35.9105215347335,
+      "learning_rate": 8.874866084819264e-06,
+      "loss": 1.1248,
+      "step": 15760
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.372504021182381,
+      "learning_rate": 8.873830030772326e-06,
+      "loss": 1.2148,
+      "step": 15765
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 21.25397187021847,
+      "learning_rate": 8.872793560465254e-06,
+      "loss": 1.0871,
+      "step": 15770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.514453140096677,
+      "learning_rate": 8.871756674009424e-06,
+      "loss": 1.0848,
+      "step": 15775
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 23.66195328791996,
+      "learning_rate": 8.87071937151625e-06,
+      "loss": 1.1194,
+      "step": 15780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 26.19155099835197,
+      "learning_rate": 8.869681653097197e-06,
+      "loss": 1.1232,
+      "step": 15785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.85542025290746,
+      "learning_rate": 8.868643518863773e-06,
+      "loss": 1.1544,
+      "step": 15790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.51186349299305,
+      "learning_rate": 8.86760496892753e-06,
+      "loss": 1.1807,
+      "step": 15795
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 19.91108235467044,
+      "learning_rate": 8.866566003400063e-06,
+      "loss": 1.1473,
+      "step": 15800
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.732440182991025,
+      "learning_rate": 8.865526622393014e-06,
+      "loss": 1.1547,
+      "step": 15805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.109443017549859,
+      "learning_rate": 8.864486826018071e-06,
+      "loss": 1.0966,
+      "step": 15810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.255265210407417,
+      "learning_rate": 8.86344661438696e-06,
+      "loss": 1.1086,
+      "step": 15815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.0010913863033375,
+      "learning_rate": 8.862405987611462e-06,
+      "loss": 1.1323,
+      "step": 15820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.027189162578985,
+      "learning_rate": 8.861364945803395e-06,
+      "loss": 1.1185,
+      "step": 15825
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.516654026014469,
+      "learning_rate": 8.860323489074623e-06,
+      "loss": 1.1079,
+      "step": 15830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.650182475404735,
+      "learning_rate": 8.859281617537054e-06,
+      "loss": 1.1647,
+      "step": 15835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.986235347722728,
+      "learning_rate": 8.858239331302644e-06,
+      "loss": 1.1173,
+      "step": 15840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.23175185845024,
+      "learning_rate": 8.857196630483389e-06,
+      "loss": 1.1243,
+      "step": 15845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.945782923891434,
+      "learning_rate": 8.856153515191335e-06,
+      "loss": 1.1062,
+      "step": 15850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.685618784108,
+      "learning_rate": 8.855109985538566e-06,
+      "loss": 1.1132,
+      "step": 15855
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 25.11690911364613,
+      "learning_rate": 8.854066041637213e-06,
+      "loss": 1.1178,
+      "step": 15860
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.213614309771685,
+      "learning_rate": 8.853021683599459e-06,
+      "loss": 1.1257,
+      "step": 15865
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.781984162951699,
+      "learning_rate": 8.851976911537517e-06,
+      "loss": 1.128,
+      "step": 15870
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 18.240618915281846,
+      "learning_rate": 8.850931725563656e-06,
+      "loss": 1.1809,
+      "step": 15875
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.52773367410937,
+      "learning_rate": 8.849886125790186e-06,
+      "loss": 1.1418,
+      "step": 15880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.509552966492867,
+      "learning_rate": 8.848840112329462e-06,
+      "loss": 1.0787,
+      "step": 15885
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.612424373792031,
+      "learning_rate": 8.847793685293882e-06,
+      "loss": 1.1333,
+      "step": 15890
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.985888668119721,
+      "learning_rate": 8.846746844795888e-06,
+      "loss": 1.0613,
+      "step": 15895
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 18.08355129876919,
+      "learning_rate": 8.845699590947968e-06,
+      "loss": 1.0849,
+      "step": 15900
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.862136138661226,
+      "learning_rate": 8.844651923862655e-06,
+      "loss": 1.1323,
+      "step": 15905
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 42.95501916755341,
+      "learning_rate": 8.843603843652526e-06,
+      "loss": 1.1118,
+      "step": 15910
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 17.72326961263032,
+      "learning_rate": 8.8425553504302e-06,
+      "loss": 1.1186,
+      "step": 15915
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 17.028465241187952,
+      "learning_rate": 8.841506444308344e-06,
+      "loss": 1.1104,
+      "step": 15920
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.453218113393806,
+      "learning_rate": 8.840457125399664e-06,
+      "loss": 1.0451,
+      "step": 15925
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.375243031040855,
+      "learning_rate": 8.83940739381692e-06,
+      "loss": 1.1673,
+      "step": 15930
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.496741585782315,
+      "learning_rate": 8.838357249672904e-06,
+      "loss": 1.1325,
+      "step": 15935
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.022342603648964,
+      "learning_rate": 8.837306693080463e-06,
+      "loss": 1.1297,
+      "step": 15940
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.239180693152152,
+      "learning_rate": 8.836255724152482e-06,
+      "loss": 1.1209,
+      "step": 15945
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.476427111269794,
+      "learning_rate": 8.835204343001895e-06,
+      "loss": 1.047,
+      "step": 15950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.23621621332625,
+      "learning_rate": 8.834152549741672e-06,
+      "loss": 1.0892,
+      "step": 15955
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.9072491904722595,
+      "learning_rate": 8.833100344484839e-06,
+      "loss": 1.1609,
+      "step": 15960
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 17.171899439262887,
+      "learning_rate": 8.832047727344456e-06,
+      "loss": 1.1218,
+      "step": 15965
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.660014783667418,
+      "learning_rate": 8.830994698433633e-06,
+      "loss": 1.0618,
+      "step": 15970
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.965348345133702,
+      "learning_rate": 8.82994125786552e-06,
+      "loss": 1.1037,
+      "step": 15975
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 34.27148989398509,
+      "learning_rate": 8.828887405753319e-06,
+      "loss": 1.1039,
+      "step": 15980
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.381648603300448,
+      "learning_rate": 8.827833142210266e-06,
+      "loss": 1.1186,
+      "step": 15985
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.703178732356198,
+      "learning_rate": 8.826778467349649e-06,
+      "loss": 1.1484,
+      "step": 15990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 17.094715932962195,
+      "learning_rate": 8.825723381284797e-06,
+      "loss": 1.0973,
+      "step": 15995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.756907025277915,
+      "learning_rate": 8.824667884129084e-06,
+      "loss": 1.0878,
+      "step": 16000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 21.993517813012506,
+      "learning_rate": 8.823611975995926e-06,
+      "loss": 1.1696,
+      "step": 16005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.46932570998332,
+      "learning_rate": 8.822555656998787e-06,
+      "loss": 1.0832,
+      "step": 16010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.043052096345904,
+      "learning_rate": 8.821498927251173e-06,
+      "loss": 1.1203,
+      "step": 16015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.103538532001767,
+      "learning_rate": 8.820441786866633e-06,
+      "loss": 1.0753,
+      "step": 16020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.18136565211477,
+      "learning_rate": 8.819384235958763e-06,
+      "loss": 1.1236,
+      "step": 16025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 15.248905393904629,
+      "learning_rate": 8.8183262746412e-06,
+      "loss": 1.1073,
+      "step": 16030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.340618674349047,
+      "learning_rate": 8.817267903027626e-06,
+      "loss": 1.1415,
+      "step": 16035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.372859072118066,
+      "learning_rate": 8.81620912123177e-06,
+      "loss": 1.1508,
+      "step": 16040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.326172721309334,
+      "learning_rate": 8.815149929367401e-06,
+      "loss": 1.1244,
+      "step": 16045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.236125571217685,
+      "learning_rate": 8.814090327548337e-06,
+      "loss": 1.1394,
+      "step": 16050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.296148214222332,
+      "learning_rate": 8.813030315888433e-06,
+      "loss": 1.1009,
+      "step": 16055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.182229401804242,
+      "learning_rate": 8.811969894501592e-06,
+      "loss": 1.0911,
+      "step": 16060
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.998399376154743,
+      "learning_rate": 8.810909063501764e-06,
+      "loss": 1.0595,
+      "step": 16065
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 23.497996454399594,
+      "learning_rate": 8.809847823002938e-06,
+      "loss": 1.1009,
+      "step": 16070
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.322854357171817,
+      "learning_rate": 8.808786173119148e-06,
+      "loss": 1.0992,
+      "step": 16075
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.196916653237418,
+      "learning_rate": 8.807724113964476e-06,
+      "loss": 1.0928,
+      "step": 16080
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.100637163787419,
+      "learning_rate": 8.806661645653042e-06,
+      "loss": 1.0811,
+      "step": 16085
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.223392754663758,
+      "learning_rate": 8.805598768299016e-06,
+      "loss": 1.0792,
+      "step": 16090
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.232467412039565,
+      "learning_rate": 8.804535482016603e-06,
+      "loss": 1.0971,
+      "step": 16095
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.587657103940197,
+      "learning_rate": 8.803471786920066e-06,
+      "loss": 1.0939,
+      "step": 16100
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.405831133128997,
+      "learning_rate": 8.802407683123696e-06,
+      "loss": 1.0998,
+      "step": 16105
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.139479921017692,
+      "learning_rate": 8.801343170741841e-06,
+      "loss": 1.1771,
+      "step": 16110
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.091588251831105,
+      "learning_rate": 8.800278249888888e-06,
+      "loss": 1.1341,
+      "step": 16115
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.732088914984313,
+      "learning_rate": 8.79921292067926e-06,
+      "loss": 1.1608,
+      "step": 16120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 22.358520755565852,
+      "learning_rate": 8.798147183227441e-06,
+      "loss": 1.0895,
+      "step": 16125
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.021681756582822,
+      "learning_rate": 8.797081037647944e-06,
+      "loss": 1.1165,
+      "step": 16130
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.9118619006907,
+      "learning_rate": 8.796014484055328e-06,
+      "loss": 1.0875,
+      "step": 16135
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.21697134216012,
+      "learning_rate": 8.794947522564205e-06,
+      "loss": 1.1072,
+      "step": 16140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.7939752733004815,
+      "learning_rate": 8.793880153289224e-06,
+      "loss": 1.0968,
+      "step": 16145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.830503959797658,
+      "learning_rate": 8.792812376345074e-06,
+      "loss": 1.1274,
+      "step": 16150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.685047451008934,
+      "learning_rate": 8.7917441918465e-06,
+      "loss": 1.0992,
+      "step": 16155
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.420774631204822,
+      "learning_rate": 8.790675599908271e-06,
+      "loss": 1.1241,
+      "step": 16160
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.473468488099863,
+      "learning_rate": 8.789606600645225e-06,
+      "loss": 1.1349,
+      "step": 16165
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.817799435816081,
+      "learning_rate": 8.788537194172224e-06,
+      "loss": 1.1052,
+      "step": 16170
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 21.580602261212043,
+      "learning_rate": 8.78746738060418e-06,
+      "loss": 1.0905,
+      "step": 16175
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.060631817923278,
+      "learning_rate": 8.78639716005605e-06,
+      "loss": 1.1561,
+      "step": 16180
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.815627525466809,
+      "learning_rate": 8.785326532642837e-06,
+      "loss": 1.1507,
+      "step": 16185
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.026518156587827,
+      "learning_rate": 8.784255498479579e-06,
+      "loss": 1.157,
+      "step": 16190
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.07346093619661,
+      "learning_rate": 8.783184057681369e-06,
+      "loss": 1.1179,
+      "step": 16195
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 20.969875759523667,
+      "learning_rate": 8.782112210363334e-06,
+      "loss": 1.0961,
+      "step": 16200
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.833205612253515,
+      "learning_rate": 8.781039956640652e-06,
+      "loss": 1.1373,
+      "step": 16205
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.355534811501904,
+      "learning_rate": 8.779967296628538e-06,
+      "loss": 1.1,
+      "step": 16210
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.374499462893863,
+      "learning_rate": 8.778894230442256e-06,
+      "loss": 1.1991,
+      "step": 16215
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.895556724357824,
+      "learning_rate": 8.777820758197113e-06,
+      "loss": 1.2023,
+      "step": 16220
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.190352446121144,
+      "learning_rate": 8.776746880008456e-06,
+      "loss": 1.0906,
+      "step": 16225
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.572189145294844,
+      "learning_rate": 8.775672595991679e-06,
+      "loss": 1.0985,
+      "step": 16230
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 30.780148334145984,
+      "learning_rate": 8.774597906262216e-06,
+      "loss": 1.0957,
+      "step": 16235
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 20.840074732905205,
+      "learning_rate": 8.77352281093555e-06,
+      "loss": 1.1014,
+      "step": 16240
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 25.14932494567834,
+      "learning_rate": 8.772447310127205e-06,
+      "loss": 1.1366,
+      "step": 16245
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.524691081025394,
+      "learning_rate": 8.771371403952748e-06,
+      "loss": 1.1546,
+      "step": 16250
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.458253430618536,
+      "learning_rate": 8.77029509252779e-06,
+      "loss": 1.1583,
+      "step": 16255
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.343448104102179,
+      "learning_rate": 8.769218375967983e-06,
+      "loss": 1.0449,
+      "step": 16260
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.282051545362369,
+      "learning_rate": 8.768141254389026e-06,
+      "loss": 1.1447,
+      "step": 16265
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.273008482653508,
+      "learning_rate": 8.767063727906664e-06,
+      "loss": 1.1134,
+      "step": 16270
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.052302981351795,
+      "learning_rate": 8.765985796636677e-06,
+      "loss": 1.1085,
+      "step": 16275
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.170097022853318,
+      "learning_rate": 8.764907460694897e-06,
+      "loss": 1.1784,
+      "step": 16280
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.256415269142373,
+      "learning_rate": 8.763828720197193e-06,
+      "loss": 1.108,
+      "step": 16285
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 15.954914884126506,
+      "learning_rate": 8.76274957525948e-06,
+      "loss": 1.0677,
+      "step": 16290
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.77344232994075,
+      "learning_rate": 8.761670025997724e-06,
+      "loss": 1.1047,
+      "step": 16295
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.489888817776245,
+      "learning_rate": 8.76059007252792e-06,
+      "loss": 1.1786,
+      "step": 16300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.79756346741257,
+      "learning_rate": 8.759509714966114e-06,
+      "loss": 1.1012,
+      "step": 16305
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.970560448934954,
+      "learning_rate": 8.758428953428399e-06,
+      "loss": 1.0937,
+      "step": 16310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 15.696285012759176,
+      "learning_rate": 8.757347788030906e-06,
+      "loss": 1.1387,
+      "step": 16315
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.007036261122607,
+      "learning_rate": 8.75626621888981e-06,
+      "loss": 1.0979,
+      "step": 16320
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.03476086094843,
+      "learning_rate": 8.755184246121332e-06,
+      "loss": 1.0754,
+      "step": 16325
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.576555628021772,
+      "learning_rate": 8.754101869841733e-06,
+      "loss": 1.1414,
+      "step": 16330
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.362484837905388,
+      "learning_rate": 8.753019090167319e-06,
+      "loss": 1.14,
+      "step": 16335
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 76.12247768134962,
+      "learning_rate": 8.751935907214442e-06,
+      "loss": 1.139,
+      "step": 16340
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 24.24045767611907,
+      "learning_rate": 8.750852321099492e-06,
+      "loss": 1.1291,
+      "step": 16345
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 46.17592888130963,
+      "learning_rate": 8.749768331938908e-06,
+      "loss": 1.1227,
+      "step": 16350
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 44.6076177145666,
+      "learning_rate": 8.748683939849168e-06,
+      "loss": 1.0924,
+      "step": 16355
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 38.36581040746154,
+      "learning_rate": 8.747599144946792e-06,
+      "loss": 1.1413,
+      "step": 16360
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.272535418452899,
+      "learning_rate": 8.746513947348349e-06,
+      "loss": 1.1449,
+      "step": 16365
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 22.622182584692663,
+      "learning_rate": 8.745428347170449e-06,
+      "loss": 1.1333,
+      "step": 16370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.204687707319023,
+      "learning_rate": 8.744342344529743e-06,
+      "loss": 1.1012,
+      "step": 16375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.964796972326862,
+      "learning_rate": 8.743255939542928e-06,
+      "loss": 1.1267,
+      "step": 16380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.320735472324557,
+      "learning_rate": 8.74216913232674e-06,
+      "loss": 1.1239,
+      "step": 16385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 24.054811884597008,
+      "learning_rate": 8.741081922997964e-06,
+      "loss": 1.1229,
+      "step": 16390
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 18.22144563252079,
+      "learning_rate": 8.739994311673425e-06,
+      "loss": 1.1444,
+      "step": 16395
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.169098747662678,
+      "learning_rate": 8.73890629846999e-06,
+      "loss": 1.1194,
+      "step": 16400
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.202104055040978,
+      "learning_rate": 8.737817883504574e-06,
+      "loss": 1.1159,
+      "step": 16405
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.78528051776811,
+      "learning_rate": 8.73672906689413e-06,
+      "loss": 1.1106,
+      "step": 16410
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.062906823839292,
+      "learning_rate": 8.735639848755655e-06,
+      "loss": 1.108,
+      "step": 16415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.7989081569689045,
+      "learning_rate": 8.734550229206193e-06,
+      "loss": 1.1217,
+      "step": 16420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.928192640795595,
+      "learning_rate": 8.733460208362827e-06,
+      "loss": 1.0494,
+      "step": 16425
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.685317764958415,
+      "learning_rate": 8.732369786342684e-06,
+      "loss": 1.1242,
+      "step": 16430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 22.138909155666813,
+      "learning_rate": 8.731278963262934e-06,
+      "loss": 1.0998,
+      "step": 16435
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.005040029562854,
+      "learning_rate": 8.730187739240794e-06,
+      "loss": 1.0938,
+      "step": 16440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.659180914035443,
+      "learning_rate": 8.729096114393519e-06,
+      "loss": 1.095,
+      "step": 16445
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.040709143735745,
+      "learning_rate": 8.728004088838407e-06,
+      "loss": 1.1103,
+      "step": 16450
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.620165078357946,
+      "learning_rate": 8.726911662692804e-06,
+      "loss": 1.0346,
+      "step": 16455
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.4202464009306315,
+      "learning_rate": 8.725818836074094e-06,
+      "loss": 1.0585,
+      "step": 16460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.122519344833424,
+      "learning_rate": 8.724725609099708e-06,
+      "loss": 1.0677,
+      "step": 16465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 18.379173430495328,
+      "learning_rate": 8.723631981887115e-06,
+      "loss": 1.1095,
+      "step": 16470
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.807500606971903,
+      "learning_rate": 8.722537954553834e-06,
+      "loss": 1.1456,
+      "step": 16475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 19.057686553939874,
+      "learning_rate": 8.72144352721742e-06,
+      "loss": 1.1618,
+      "step": 16480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.129235480390931,
+      "learning_rate": 8.720348699995475e-06,
+      "loss": 1.1248,
+      "step": 16485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.865586804101719,
+      "learning_rate": 8.719253473005642e-06,
+      "loss": 1.1253,
+      "step": 16490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.141786377717228,
+      "learning_rate": 8.718157846365612e-06,
+      "loss": 1.1272,
+      "step": 16495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.014185685581909,
+      "learning_rate": 8.717061820193109e-06,
+      "loss": 1.0713,
+      "step": 16500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.39974596188113,
+      "learning_rate": 8.71596539460591e-06,
+      "loss": 1.0618,
+      "step": 16505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.767146051283026,
+      "learning_rate": 8.714868569721829e-06,
+      "loss": 1.1561,
+      "step": 16510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.77005098166936,
+      "learning_rate": 8.713771345658725e-06,
+      "loss": 1.086,
+      "step": 16515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.483453547078433,
+      "learning_rate": 8.7126737225345e-06,
+      "loss": 1.1,
+      "step": 16520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.028485044351108,
+      "learning_rate": 8.711575700467099e-06,
+      "loss": 1.1149,
+      "step": 16525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.776660395252255,
+      "learning_rate": 8.710477279574508e-06,
+      "loss": 1.0807,
+      "step": 16530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.321915649913745,
+      "learning_rate": 8.709378459974757e-06,
+      "loss": 1.0956,
+      "step": 16535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.015565458827222,
+      "learning_rate": 8.708279241785921e-06,
+      "loss": 1.0931,
+      "step": 16540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 38.165963473472665,
+      "learning_rate": 8.707179625126115e-06,
+      "loss": 1.1324,
+      "step": 16545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.334238440334481,
+      "learning_rate": 8.706079610113496e-06,
+      "loss": 1.1376,
+      "step": 16550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.098952970685284,
+      "learning_rate": 8.704979196866266e-06,
+      "loss": 1.1392,
+      "step": 16555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.854497211266295,
+      "learning_rate": 8.703878385502671e-06,
+      "loss": 1.1182,
+      "step": 16560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.669911348817724,
+      "learning_rate": 8.702777176140997e-06,
+      "loss": 1.0911,
+      "step": 16565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.752468121840435,
+      "learning_rate": 8.701675568899574e-06,
+      "loss": 1.0768,
+      "step": 16570
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.627644905402025,
+      "learning_rate": 8.700573563896774e-06,
+      "loss": 1.0934,
+      "step": 16575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 20.195405848785732,
+      "learning_rate": 8.699471161251012e-06,
+      "loss": 1.1107,
+      "step": 16580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.198940222449206,
+      "learning_rate": 8.698368361080749e-06,
+      "loss": 1.1203,
+      "step": 16585
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.2067746874275365,
+      "learning_rate": 8.69726516350448e-06,
+      "loss": 1.0956,
+      "step": 16590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.733743717287036,
+      "learning_rate": 8.696161568640756e-06,
+      "loss": 1.0886,
+      "step": 16595
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.89035233966049,
+      "learning_rate": 8.695057576608156e-06,
+      "loss": 1.1542,
+      "step": 16600
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.117215892219331,
+      "learning_rate": 8.693953187525312e-06,
+      "loss": 1.1133,
+      "step": 16605
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.567981163642601,
+      "learning_rate": 8.692848401510899e-06,
+      "loss": 1.1064,
+      "step": 16610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.234473257192498,
+      "learning_rate": 8.691743218683623e-06,
+      "loss": 1.0654,
+      "step": 16615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.476461636235864,
+      "learning_rate": 8.690637639162247e-06,
+      "loss": 1.0996,
+      "step": 16620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.89283221404281,
+      "learning_rate": 8.689531663065569e-06,
+      "loss": 1.0795,
+      "step": 16625
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.6064564707496665,
+      "learning_rate": 8.68842529051243e-06,
+      "loss": 1.1061,
+      "step": 16630
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.8832524553344046,
+      "learning_rate": 8.687318521621715e-06,
+      "loss": 1.1455,
+      "step": 16635
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.96740972184527,
+      "learning_rate": 8.686211356512353e-06,
+      "loss": 1.1003,
+      "step": 16640
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.563313662079871,
+      "learning_rate": 8.68510379530331e-06,
+      "loss": 1.1143,
+      "step": 16645
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.39720690050603,
+      "learning_rate": 8.683995838113602e-06,
+      "loss": 1.1014,
+      "step": 16650
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.758215647478176,
+      "learning_rate": 8.682887485062283e-06,
+      "loss": 1.1131,
+      "step": 16655
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.8197066067424235,
+      "learning_rate": 8.68177873626845e-06,
+      "loss": 1.0837,
+      "step": 16660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 20.021803830672226,
+      "learning_rate": 8.680669591851241e-06,
+      "loss": 1.1323,
+      "step": 16665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.48333234748205,
+      "learning_rate": 8.679560051929841e-06,
+      "loss": 1.1795,
+      "step": 16670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.004186377201357,
+      "learning_rate": 8.678450116623475e-06,
+      "loss": 1.0931,
+      "step": 16675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.78110602933926,
+      "learning_rate": 8.677339786051409e-06,
+      "loss": 1.0697,
+      "step": 16680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.05618061588711,
+      "learning_rate": 8.676229060332955e-06,
+      "loss": 1.133,
+      "step": 16685
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.34079703403976,
+      "learning_rate": 8.675117939587462e-06,
+      "loss": 1.0632,
+      "step": 16690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.57522360774888,
+      "learning_rate": 8.674006423934329e-06,
+      "loss": 1.0819,
+      "step": 16695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.981022926998485,
+      "learning_rate": 8.672894513492989e-06,
+      "loss": 1.0911,
+      "step": 16700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.377801934015054,
+      "learning_rate": 8.671782208382925e-06,
+      "loss": 1.0995,
+      "step": 16705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.55500338678948,
+      "learning_rate": 8.670669508723658e-06,
+      "loss": 1.1319,
+      "step": 16710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.287168833812008,
+      "learning_rate": 8.669556414634753e-06,
+      "loss": 1.093,
+      "step": 16715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.485193154142666,
+      "learning_rate": 8.668442926235815e-06,
+      "loss": 1.0344,
+      "step": 16720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.674636382508078,
+      "learning_rate": 8.667329043646495e-06,
+      "loss": 1.1413,
+      "step": 16725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.605226086662105,
+      "learning_rate": 8.666214766986484e-06,
+      "loss": 1.1142,
+      "step": 16730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.042553356278292,
+      "learning_rate": 8.665100096375517e-06,
+      "loss": 1.1072,
+      "step": 16735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.721074379071956,
+      "learning_rate": 8.663985031933367e-06,
+      "loss": 1.1298,
+      "step": 16740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.066261313152033,
+      "learning_rate": 8.662869573779856e-06,
+      "loss": 1.1151,
+      "step": 16745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.267548006539362,
+      "learning_rate": 8.661753722034844e-06,
+      "loss": 1.1654,
+      "step": 16750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.069320807507802,
+      "learning_rate": 8.660637476818235e-06,
+      "loss": 1.1548,
+      "step": 16755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.917268241092627,
+      "learning_rate": 8.659520838249972e-06,
+      "loss": 1.1078,
+      "step": 16760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.103795176539565,
+      "learning_rate": 8.658403806450045e-06,
+      "loss": 1.1569,
+      "step": 16765
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.039259669749306,
+      "learning_rate": 8.657286381538484e-06,
+      "loss": 1.0719,
+      "step": 16770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.64818010374265,
+      "learning_rate": 8.656168563635358e-06,
+      "loss": 1.0916,
+      "step": 16775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.667193970299143,
+      "learning_rate": 8.655050352860786e-06,
+      "loss": 1.1076,
+      "step": 16780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.629460052604731,
+      "learning_rate": 8.653931749334922e-06,
+      "loss": 1.1023,
+      "step": 16785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 33.732872738539115,
+      "learning_rate": 8.652812753177965e-06,
+      "loss": 1.1522,
+      "step": 16790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 34.37900182726514,
+      "learning_rate": 8.65169336451016e-06,
+      "loss": 1.1751,
+      "step": 16795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.114958382608194,
+      "learning_rate": 8.650573583451782e-06,
+      "loss": 1.1227,
+      "step": 16800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.55020664542636,
+      "learning_rate": 8.649453410123164e-06,
+      "loss": 1.1111,
+      "step": 16805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.922271126707626,
+      "learning_rate": 8.648332844644669e-06,
+      "loss": 1.087,
+      "step": 16810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.1183304135241885,
+      "learning_rate": 8.64721188713671e-06,
+      "loss": 1.1104,
+      "step": 16815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.800394367832217,
+      "learning_rate": 8.646090537719738e-06,
+      "loss": 1.1132,
+      "step": 16820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.588142176645427,
+      "learning_rate": 8.644968796514244e-06,
+      "loss": 1.124,
+      "step": 16825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 37.79013112368904,
+      "learning_rate": 8.643846663640767e-06,
+      "loss": 1.0859,
+      "step": 16830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.83873331471025,
+      "learning_rate": 8.642724139219887e-06,
+      "loss": 1.0483,
+      "step": 16835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.748327254509929,
+      "learning_rate": 8.641601223372219e-06,
+      "loss": 1.1463,
+      "step": 16840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.32810310898289,
+      "learning_rate": 8.640477916218427e-06,
+      "loss": 1.147,
+      "step": 16845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.781404767906865,
+      "learning_rate": 8.639354217879219e-06,
+      "loss": 1.1185,
+      "step": 16850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.948261878606812,
+      "learning_rate": 8.638230128475339e-06,
+      "loss": 1.1286,
+      "step": 16855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.774906837685148,
+      "learning_rate": 8.637105648127572e-06,
+      "loss": 1.1657,
+      "step": 16860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.829624181780629,
+      "learning_rate": 8.635980776956753e-06,
+      "loss": 1.1155,
+      "step": 16865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.265451296320978,
+      "learning_rate": 8.634855515083754e-06,
+      "loss": 1.0956,
+      "step": 16870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.919765183096331,
+      "learning_rate": 8.633729862629485e-06,
+      "loss": 1.0738,
+      "step": 16875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.275255046385121,
+      "learning_rate": 8.632603819714907e-06,
+      "loss": 1.1066,
+      "step": 16880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.779714646987017,
+      "learning_rate": 8.631477386461017e-06,
+      "loss": 1.1275,
+      "step": 16885
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.48483276220722,
+      "learning_rate": 8.630350562988854e-06,
+      "loss": 1.0948,
+      "step": 16890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.85938014572465,
+      "learning_rate": 8.629223349419502e-06,
+      "loss": 1.0943,
+      "step": 16895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.004591130647585,
+      "learning_rate": 8.628095745874084e-06,
+      "loss": 1.1074,
+      "step": 16900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.65457644471501,
+      "learning_rate": 8.626967752473766e-06,
+      "loss": 1.1077,
+      "step": 16905
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.182654798327075,
+      "learning_rate": 8.625839369339756e-06,
+      "loss": 1.1352,
+      "step": 16910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.047457540280567,
+      "learning_rate": 8.624710596593304e-06,
+      "loss": 1.0868,
+      "step": 16915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.733693314136023,
+      "learning_rate": 8.6235814343557e-06,
+      "loss": 1.1602,
+      "step": 16920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.463943057759602,
+      "learning_rate": 8.62245188274828e-06,
+      "loss": 1.0792,
+      "step": 16925
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.136447871619055,
+      "learning_rate": 8.621321941892418e-06,
+      "loss": 1.1297,
+      "step": 16930
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.101580015915936,
+      "learning_rate": 8.62019161190953e-06,
+      "loss": 1.1202,
+      "step": 16935
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.992471439897448,
+      "learning_rate": 8.619060892921077e-06,
+      "loss": 1.0861,
+      "step": 16940
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.160694331840603,
+      "learning_rate": 8.61792978504856e-06,
+      "loss": 1.0824,
+      "step": 16945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.43779713325381,
+      "learning_rate": 8.616798288413519e-06,
+      "loss": 1.1127,
+      "step": 16950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.919756875258514,
+      "learning_rate": 8.61566640313754e-06,
+      "loss": 1.1164,
+      "step": 16955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.73141976572175,
+      "learning_rate": 8.61453412934225e-06,
+      "loss": 1.0591,
+      "step": 16960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.1210669962111535,
+      "learning_rate": 8.613401467149313e-06,
+      "loss": 1.0836,
+      "step": 16965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.10212590273476,
+      "learning_rate": 8.612268416680445e-06,
+      "loss": 1.0952,
+      "step": 16970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.103311016410487,
+      "learning_rate": 8.61113497805739e-06,
+      "loss": 1.04,
+      "step": 16975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.858408644862056,
+      "learning_rate": 8.610001151401947e-06,
+      "loss": 1.1391,
+      "step": 16980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 28.537734196563537,
+      "learning_rate": 8.608866936835946e-06,
+      "loss": 1.1008,
+      "step": 16985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.678920307043942,
+      "learning_rate": 8.607732334481268e-06,
+      "loss": 1.1211,
+      "step": 16990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.049879390681618,
+      "learning_rate": 8.606597344459826e-06,
+      "loss": 1.0908,
+      "step": 16995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.000289047398722,
+      "learning_rate": 8.605461966893582e-06,
+      "loss": 1.0672,
+      "step": 17000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.300366652184351,
+      "learning_rate": 8.604326201904541e-06,
+      "loss": 1.1107,
+      "step": 17005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.03427331057001,
+      "learning_rate": 8.60319004961474e-06,
+      "loss": 1.0685,
+      "step": 17010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.520653156003474,
+      "learning_rate": 8.602053510146265e-06,
+      "loss": 1.0649,
+      "step": 17015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 21.63507991906392,
+      "learning_rate": 8.600916583621247e-06,
+      "loss": 1.1288,
+      "step": 17020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.793669670278979,
+      "learning_rate": 8.599779270161846e-06,
+      "loss": 1.1012,
+      "step": 17025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 22.626520487692865,
+      "learning_rate": 8.598641569890275e-06,
+      "loss": 1.1016,
+      "step": 17030
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 32.28364135134357,
+      "learning_rate": 8.597503482928788e-06,
+      "loss": 1.1309,
+      "step": 17035
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 32.0488042790708,
+      "learning_rate": 8.596365009399676e-06,
+      "loss": 1.114,
+      "step": 17040
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 16.97518127775392,
+      "learning_rate": 8.595226149425268e-06,
+      "loss": 1.1284,
+      "step": 17045
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.59692849599178,
+      "learning_rate": 8.594086903127946e-06,
+      "loss": 1.1647,
+      "step": 17050
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.677751460142839,
+      "learning_rate": 8.592947270630122e-06,
+      "loss": 1.0799,
+      "step": 17055
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 56.47850036664735,
+      "learning_rate": 8.591807252054261e-06,
+      "loss": 1.1001,
+      "step": 17060
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 16.033786859529016,
+      "learning_rate": 8.590666847522855e-06,
+      "loss": 1.0932,
+      "step": 17065
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 15.922058186030087,
+      "learning_rate": 8.58952605715845e-06,
+      "loss": 1.1408,
+      "step": 17070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.695337545016002,
+      "learning_rate": 8.58838488108363e-06,
+      "loss": 1.0902,
+      "step": 17075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.853955323983569,
+      "learning_rate": 8.587243319421018e-06,
+      "loss": 1.0831,
+      "step": 17080
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.282878247017411,
+      "learning_rate": 8.586101372293278e-06,
+      "loss": 1.1114,
+      "step": 17085
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.932947242973346,
+      "learning_rate": 8.584959039823118e-06,
+      "loss": 1.1423,
+      "step": 17090
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.271938691365637,
+      "learning_rate": 8.583816322133289e-06,
+      "loss": 1.1269,
+      "step": 17095
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.445713789115437,
+      "learning_rate": 8.58267321934658e-06,
+      "loss": 1.1257,
+      "step": 17100
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 29.421390018795922,
+      "learning_rate": 8.581529731585823e-06,
+      "loss": 1.0923,
+      "step": 17105
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.998397195340639,
+      "learning_rate": 8.580385858973888e-06,
+      "loss": 1.133,
+      "step": 17110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.831929265163076,
+      "learning_rate": 8.57924160163369e-06,
+      "loss": 1.0752,
+      "step": 17115
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.415976045958236,
+      "learning_rate": 8.578096959688186e-06,
+      "loss": 1.1306,
+      "step": 17120
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.972343347910804,
+      "learning_rate": 8.576951933260375e-06,
+      "loss": 1.1042,
+      "step": 17125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.943121935071054,
+      "learning_rate": 8.57580652247329e-06,
+      "loss": 1.1209,
+      "step": 17130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.278290573093377,
+      "learning_rate": 8.574660727450015e-06,
+      "loss": 1.1442,
+      "step": 17135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.003901926627689,
+      "learning_rate": 8.573514548313666e-06,
+      "loss": 1.089,
+      "step": 17140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.367489844559957,
+      "learning_rate": 8.572367985187408e-06,
+      "loss": 1.126,
+      "step": 17145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.023590407704637,
+      "learning_rate": 8.571221038194446e-06,
+      "loss": 1.0789,
+      "step": 17150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.418039829453431,
+      "learning_rate": 8.570073707458022e-06,
+      "loss": 1.1212,
+      "step": 17155
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.534951519053513,
+      "learning_rate": 8.56892599310142e-06,
+      "loss": 1.0669,
+      "step": 17160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.211321714989818,
+      "learning_rate": 8.56777789524797e-06,
+      "loss": 1.0577,
+      "step": 17165
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.349987216251385,
+      "learning_rate": 8.56662941402104e-06,
+      "loss": 1.12,
+      "step": 17170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.514759203836478,
+      "learning_rate": 8.56548054954404e-06,
+      "loss": 1.0546,
+      "step": 17175
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.264870038261684,
+      "learning_rate": 8.564331301940419e-06,
+      "loss": 1.0375,
+      "step": 17180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.773221916226845,
+      "learning_rate": 8.563181671333666e-06,
+      "loss": 1.1053,
+      "step": 17185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.392172356282338,
+      "learning_rate": 8.56203165784732e-06,
+      "loss": 1.1423,
+      "step": 17190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.528523758799187,
+      "learning_rate": 8.560881261604951e-06,
+      "loss": 1.143,
+      "step": 17195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.933484178575647,
+      "learning_rate": 8.559730482730176e-06,
+      "loss": 1.096,
+      "step": 17200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.325635838846004,
+      "learning_rate": 8.55857932134665e-06,
+      "loss": 1.0839,
+      "step": 17205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.721837717096488,
+      "learning_rate": 8.55742777757807e-06,
+      "loss": 1.122,
+      "step": 17210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.7142419398959134,
+      "learning_rate": 8.556275851548173e-06,
+      "loss": 1.0708,
+      "step": 17215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.614369692298798,
+      "learning_rate": 8.555123543380743e-06,
+      "loss": 1.0669,
+      "step": 17220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.420334429553925,
+      "learning_rate": 8.553970853199599e-06,
+      "loss": 1.1087,
+      "step": 17225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 21.343029747736015,
+      "learning_rate": 8.552817781128602e-06,
+      "loss": 1.1398,
+      "step": 17230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 18.319859430286797,
+      "learning_rate": 8.551664327291653e-06,
+      "loss": 1.1277,
+      "step": 17235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.849825184667681,
+      "learning_rate": 8.550510491812697e-06,
+      "loss": 1.1342,
+      "step": 17240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.964489914643135,
+      "learning_rate": 8.549356274815721e-06,
+      "loss": 1.1359,
+      "step": 17245
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.42300877290792,
+      "learning_rate": 8.548201676424748e-06,
+      "loss": 1.1076,
+      "step": 17250
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.462190775163863,
+      "learning_rate": 8.547046696763845e-06,
+      "loss": 1.1187,
+      "step": 17255
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.216531661712807,
+      "learning_rate": 8.54589133595712e-06,
+      "loss": 1.1116,
+      "step": 17260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 24.40952282571262,
+      "learning_rate": 8.544735594128723e-06,
+      "loss": 1.0711,
+      "step": 17265
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 23.459383355754376,
+      "learning_rate": 8.54357947140284e-06,
+      "loss": 1.126,
+      "step": 17270
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.065887056262643,
+      "learning_rate": 8.542422967903707e-06,
+      "loss": 1.1083,
+      "step": 17275
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.537942460173479,
+      "learning_rate": 8.541266083755591e-06,
+      "loss": 1.1231,
+      "step": 17280
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.351557397708405,
+      "learning_rate": 8.540108819082805e-06,
+      "loss": 1.1319,
+      "step": 17285
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.852202759882676,
+      "learning_rate": 8.538951174009702e-06,
+      "loss": 1.0799,
+      "step": 17290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.61855316270117,
+      "learning_rate": 8.537793148660678e-06,
+      "loss": 1.1442,
+      "step": 17295
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.2349835295622,
+      "learning_rate": 8.536634743160166e-06,
+      "loss": 1.0929,
+      "step": 17300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.587824493928512,
+      "learning_rate": 8.535475957632643e-06,
+      "loss": 1.0586,
+      "step": 17305
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.667751947461904,
+      "learning_rate": 8.534316792202625e-06,
+      "loss": 1.1205,
+      "step": 17310
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 26.370449991021072,
+      "learning_rate": 8.533157246994669e-06,
+      "loss": 1.0946,
+      "step": 17315
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 36.973689379481165,
+      "learning_rate": 8.531997322133375e-06,
+      "loss": 1.1253,
+      "step": 17320
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.53314149309355,
+      "learning_rate": 8.530837017743378e-06,
+      "loss": 1.1086,
+      "step": 17325
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 25.817016166659737,
+      "learning_rate": 8.529676333949363e-06,
+      "loss": 1.1238,
+      "step": 17330
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.6761358372072435,
+      "learning_rate": 8.528515270876048e-06,
+      "loss": 1.0697,
+      "step": 17335
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.469243813776712,
+      "learning_rate": 8.527353828648194e-06,
+      "loss": 1.1164,
+      "step": 17340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.649172251010545,
+      "learning_rate": 8.526192007390606e-06,
+      "loss": 1.1045,
+      "step": 17345
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.4949183545598075,
+      "learning_rate": 8.525029807228123e-06,
+      "loss": 1.0692,
+      "step": 17350
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.811352187241923,
+      "learning_rate": 8.52386722828563e-06,
+      "loss": 1.0709,
+      "step": 17355
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.980562540623275,
+      "learning_rate": 8.522704270688052e-06,
+      "loss": 1.1251,
+      "step": 17360
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 22.60227073390142,
+      "learning_rate": 8.52154093456035e-06,
+      "loss": 1.1097,
+      "step": 17365
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.706177546209313,
+      "learning_rate": 8.520377220027536e-06,
+      "loss": 1.1025,
+      "step": 17370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 24.111639874634594,
+      "learning_rate": 8.519213127214652e-06,
+      "loss": 1.0489,
+      "step": 17375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.64188633456058,
+      "learning_rate": 8.518048656246787e-06,
+      "loss": 1.1185,
+      "step": 17380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.108092664502759,
+      "learning_rate": 8.516883807249068e-06,
+      "loss": 1.122,
+      "step": 17385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.117524805180457,
+      "learning_rate": 8.515718580346661e-06,
+      "loss": 1.1076,
+      "step": 17390
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 22.24577824837758,
+      "learning_rate": 8.514552975664777e-06,
+      "loss": 1.1248,
+      "step": 17395
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.026810677379471,
+      "learning_rate": 8.513386993328665e-06,
+      "loss": 1.1077,
+      "step": 17400
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.1915382204436025,
+      "learning_rate": 8.512220633463614e-06,
+      "loss": 1.0916,
+      "step": 17405
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.839240989711329,
+      "learning_rate": 8.511053896194957e-06,
+      "loss": 1.1187,
+      "step": 17410
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 26.15965631789435,
+      "learning_rate": 8.509886781648063e-06,
+      "loss": 1.1202,
+      "step": 17415
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 22.241398466093468,
+      "learning_rate": 8.508719289948344e-06,
+      "loss": 1.1342,
+      "step": 17420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.247589317940527,
+      "learning_rate": 8.507551421221249e-06,
+      "loss": 1.0671,
+      "step": 17425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 22.653854170506236,
+      "learning_rate": 8.506383175592277e-06,
+      "loss": 1.1051,
+      "step": 17430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 19.541347250509766,
+      "learning_rate": 8.505214553186958e-06,
+      "loss": 1.1214,
+      "step": 17435
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 25.45608404430628,
+      "learning_rate": 8.504045554130865e-06,
+      "loss": 1.1242,
+      "step": 17440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.94343309854376,
+      "learning_rate": 8.502876178549613e-06,
+      "loss": 1.0825,
+      "step": 17445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.730566313918818,
+      "learning_rate": 8.501706426568855e-06,
+      "loss": 1.0683,
+      "step": 17450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.819286866830012,
+      "learning_rate": 8.500536298314289e-06,
+      "loss": 1.036,
+      "step": 17455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.960477254420302,
+      "learning_rate": 8.499365793911647e-06,
+      "loss": 1.1349,
+      "step": 17460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.312382476193674,
+      "learning_rate": 8.498194913486708e-06,
+      "loss": 1.0864,
+      "step": 17465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.997993720101222,
+      "learning_rate": 8.497023657165286e-06,
+      "loss": 1.1126,
+      "step": 17470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.806390923878613,
+      "learning_rate": 8.495852025073239e-06,
+      "loss": 1.089,
+      "step": 17475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 27.91487422067381,
+      "learning_rate": 8.494680017336463e-06,
+      "loss": 1.0606,
+      "step": 17480
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.6369038730092225,
+      "learning_rate": 8.493507634080897e-06,
+      "loss": 1.0521,
+      "step": 17485
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.734492768079987,
+      "learning_rate": 8.492334875432515e-06,
+      "loss": 1.1017,
+      "step": 17490
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 23.74806671070999,
+      "learning_rate": 8.491161741517339e-06,
+      "loss": 1.0697,
+      "step": 17495
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 23.599203453778514,
+      "learning_rate": 8.489988232461429e-06,
+      "loss": 1.1092,
+      "step": 17500
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.452834439130612,
+      "learning_rate": 8.48881434839088e-06,
+      "loss": 1.132,
+      "step": 17505
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 23.06522444093552,
+      "learning_rate": 8.487640089431828e-06,
+      "loss": 1.0897,
+      "step": 17510
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.220613719612215,
+      "learning_rate": 8.48646545571046e-06,
+      "loss": 1.1332,
+      "step": 17515
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.881418264358748,
+      "learning_rate": 8.485290447352991e-06,
+      "loss": 1.0482,
+      "step": 17520
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.721331269933362,
+      "learning_rate": 8.48411506448568e-06,
+      "loss": 1.0661,
+      "step": 17525
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 38.611690420170696,
+      "learning_rate": 8.482939307234832e-06,
+      "loss": 1.1265,
+      "step": 17530
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 20.76774680312566,
+      "learning_rate": 8.481763175726783e-06,
+      "loss": 1.1176,
+      "step": 17535
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.548987196078617,
+      "learning_rate": 8.480586670087912e-06,
+      "loss": 1.1401,
+      "step": 17540
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 58.47056233784023,
+      "learning_rate": 8.479409790444645e-06,
+      "loss": 1.0948,
+      "step": 17545
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 39.16992664642986,
+      "learning_rate": 8.478232536923439e-06,
+      "loss": 1.0975,
+      "step": 17550
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.935799863947528,
+      "learning_rate": 8.477054909650798e-06,
+      "loss": 1.136,
+      "step": 17555
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 44.17679517978767,
+      "learning_rate": 8.47587690875326e-06,
+      "loss": 1.0761,
+      "step": 17560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 28.713438754897968,
+      "learning_rate": 8.474698534357408e-06,
+      "loss": 1.1264,
+      "step": 17565
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 27.395694336471735,
+      "learning_rate": 8.473519786589863e-06,
+      "loss": 1.0881,
+      "step": 17570
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 32.9082789260856,
+      "learning_rate": 8.472340665577288e-06,
+      "loss": 1.1478,
+      "step": 17575
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.325827596095246,
+      "learning_rate": 8.471161171446383e-06,
+      "loss": 1.0463,
+      "step": 17580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.935308252137894,
+      "learning_rate": 8.469981304323893e-06,
+      "loss": 1.0782,
+      "step": 17585
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 27.91690380967443,
+      "learning_rate": 8.468801064336596e-06,
+      "loss": 1.1201,
+      "step": 17590
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 17.16370033358599,
+      "learning_rate": 8.467620451611314e-06,
+      "loss": 1.0942,
+      "step": 17595
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 29.914032628110014,
+      "learning_rate": 8.466439466274914e-06,
+      "loss": 1.1052,
+      "step": 17600
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 25.549628881727934,
+      "learning_rate": 8.465258108454292e-06,
+      "loss": 1.1478,
+      "step": 17605
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 34.184125937523916,
+      "learning_rate": 8.464076378276393e-06,
+      "loss": 1.1315,
+      "step": 17610
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 17.90086949768453,
+      "learning_rate": 8.4628942758682e-06,
+      "loss": 1.12,
+      "step": 17615
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 38.545213076593434,
+      "learning_rate": 8.461711801356735e-06,
+      "loss": 1.1479,
+      "step": 17620
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.061326289715497,
+      "learning_rate": 8.460528954869059e-06,
+      "loss": 1.0875,
+      "step": 17625
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.976899938074336,
+      "learning_rate": 8.459345736532273e-06,
+      "loss": 1.1053,
+      "step": 17630
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.303790615050954,
+      "learning_rate": 8.45816214647352e-06,
+      "loss": 1.1428,
+      "step": 17635
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.00709108138671,
+      "learning_rate": 8.456978184819986e-06,
+      "loss": 1.1406,
+      "step": 17640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.59102350372654,
+      "learning_rate": 8.455793851698886e-06,
+      "loss": 1.0858,
+      "step": 17645
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.443895983664765,
+      "learning_rate": 8.454609147237486e-06,
+      "loss": 1.1019,
+      "step": 17650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.518424946798726,
+      "learning_rate": 8.453424071563086e-06,
+      "loss": 1.0569,
+      "step": 17655
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.922045969264012,
+      "learning_rate": 8.45223862480303e-06,
+      "loss": 1.0906,
+      "step": 17660
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.311494406030087,
+      "learning_rate": 8.451052807084696e-06,
+      "loss": 1.1251,
+      "step": 17665
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.631810292606946,
+      "learning_rate": 8.449866618535509e-06,
+      "loss": 1.1309,
+      "step": 17670
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.296432872445173,
+      "learning_rate": 8.448680059282926e-06,
+      "loss": 1.078,
+      "step": 17675
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 25.743316666721856,
+      "learning_rate": 8.447493129454452e-06,
+      "loss": 1.0751,
+      "step": 17680
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.243500906845636,
+      "learning_rate": 8.446305829177626e-06,
+      "loss": 1.0927,
+      "step": 17685
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 16.192924157932385,
+      "learning_rate": 8.44511815858003e-06,
+      "loss": 1.0935,
+      "step": 17690
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.157845779594874,
+      "learning_rate": 8.443930117789282e-06,
+      "loss": 1.0926,
+      "step": 17695
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.384549416657215,
+      "learning_rate": 8.442741706933044e-06,
+      "loss": 1.1331,
+      "step": 17700
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.405902870706582,
+      "learning_rate": 8.441552926139015e-06,
+      "loss": 1.0611,
+      "step": 17705
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.270131696636666,
+      "learning_rate": 8.440363775534937e-06,
+      "loss": 1.0998,
+      "step": 17710
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.934156379264266,
+      "learning_rate": 8.439174255248587e-06,
+      "loss": 1.1216,
+      "step": 17715
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 16.03286631358554,
+      "learning_rate": 8.437984365407785e-06,
+      "loss": 1.1187,
+      "step": 17720
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.442715788040404,
+      "learning_rate": 8.43679410614039e-06,
+      "loss": 1.0706,
+      "step": 17725
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 21.182337587915736,
+      "learning_rate": 8.435603477574302e-06,
+      "loss": 1.0991,
+      "step": 17730
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.9859308497457615,
+      "learning_rate": 8.434412479837456e-06,
+      "loss": 1.089,
+      "step": 17735
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.37699563985268,
+      "learning_rate": 8.43322111305783e-06,
+      "loss": 1.0968,
+      "step": 17740
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.438308011952263,
+      "learning_rate": 8.432029377363447e-06,
+      "loss": 1.0785,
+      "step": 17745
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.260210394859314,
+      "eval_runtime": 25.4511,
+      "eval_samples_per_second": 31.669,
+      "eval_steps_per_second": 3.968,
+      "step": 17748
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.626869052426889,
+      "learning_rate": 8.430837272882359e-06,
+      "loss": 1.0274,
+      "step": 17750
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.182969299055262,
+      "learning_rate": 8.429644799742664e-06,
+      "loss": 0.9516,
+      "step": 17755
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.716810807239526,
+      "learning_rate": 8.4284519580725e-06,
+      "loss": 0.9606,
+      "step": 17760
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.131151759499285,
+      "learning_rate": 8.427258748000042e-06,
+      "loss": 0.9358,
+      "step": 17765
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.035344289845632,
+      "learning_rate": 8.426065169653506e-06,
+      "loss": 0.9879,
+      "step": 17770
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.266042187503562,
+      "learning_rate": 8.424871223161149e-06,
+      "loss": 0.9095,
+      "step": 17775
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.179896703542951,
+      "learning_rate": 8.423676908651262e-06,
+      "loss": 0.9865,
+      "step": 17780
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.362839299506145,
+      "learning_rate": 8.42248222625218e-06,
+      "loss": 0.9546,
+      "step": 17785
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.888365966924951,
+      "learning_rate": 8.421287176092281e-06,
+      "loss": 0.945,
+      "step": 17790
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.672762024171263,
+      "learning_rate": 8.420091758299973e-06,
+      "loss": 0.9378,
+      "step": 17795
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 15.506551720105456,
+      "learning_rate": 8.418895973003714e-06,
+      "loss": 0.9938,
+      "step": 17800
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.486516634692079,
+      "learning_rate": 8.417699820331993e-06,
+      "loss": 0.9082,
+      "step": 17805
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.364703652262363,
+      "learning_rate": 8.416503300413343e-06,
+      "loss": 0.9372,
+      "step": 17810
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.545462843248188,
+      "learning_rate": 8.415306413376334e-06,
+      "loss": 0.9836,
+      "step": 17815
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.605583686936086,
+      "learning_rate": 8.41410915934958e-06,
+      "loss": 0.9373,
+      "step": 17820
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.517101393133185,
+      "learning_rate": 8.412911538461728e-06,
+      "loss": 0.9467,
+      "step": 17825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 26.77141370776423,
+      "learning_rate": 8.411713550841468e-06,
+      "loss": 1.0081,
+      "step": 17830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.113379449744265,
+      "learning_rate": 8.410515196617531e-06,
+      "loss": 0.9514,
+      "step": 17835
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 14.14320939335982,
+      "learning_rate": 8.409316475918685e-06,
+      "loss": 0.9376,
+      "step": 17840
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.796556111361008,
+      "learning_rate": 8.408117388873736e-06,
+      "loss": 0.9535,
+      "step": 17845
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.034232956285297,
+      "learning_rate": 8.406917935611534e-06,
+      "loss": 0.9787,
+      "step": 17850
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.086630477089237,
+      "learning_rate": 8.405718116260961e-06,
+      "loss": 0.984,
+      "step": 17855
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.455304663864185,
+      "learning_rate": 8.404517930950948e-06,
+      "loss": 0.9526,
+      "step": 17860
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 24.500988484822788,
+      "learning_rate": 8.403317379810457e-06,
+      "loss": 0.9772,
+      "step": 17865
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 19.7152269612736,
+      "learning_rate": 8.402116462968494e-06,
+      "loss": 0.9556,
+      "step": 17870
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.355661289210014,
+      "learning_rate": 8.400915180554103e-06,
+      "loss": 0.9564,
+      "step": 17875
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 14.78776367650486,
+      "learning_rate": 8.399713532696367e-06,
+      "loss": 0.9458,
+      "step": 17880
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.81553224748409,
+      "learning_rate": 8.398511519524407e-06,
+      "loss": 0.9935,
+      "step": 17885
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 15.93516160483249,
+      "learning_rate": 8.397309141167385e-06,
+      "loss": 0.8953,
+      "step": 17890
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.285809425161123,
+      "learning_rate": 8.396106397754502e-06,
+      "loss": 0.9396,
+      "step": 17895
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.369381116123431,
+      "learning_rate": 8.394903289415e-06,
+      "loss": 0.9678,
+      "step": 17900
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.775482011076765,
+      "learning_rate": 8.393699816278156e-06,
+      "loss": 0.9888,
+      "step": 17905
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 21.196293851671367,
+      "learning_rate": 8.392495978473288e-06,
+      "loss": 0.9507,
+      "step": 17910
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.04637881645689,
+      "learning_rate": 8.391291776129756e-06,
+      "loss": 1.0118,
+      "step": 17915
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.197501902986017,
+      "learning_rate": 8.390087209376955e-06,
+      "loss": 1.0271,
+      "step": 17920
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.8033078762451735,
+      "learning_rate": 8.388882278344323e-06,
+      "loss": 0.9333,
+      "step": 17925
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.189558038761195,
+      "learning_rate": 8.387676983161334e-06,
+      "loss": 0.9384,
+      "step": 17930
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.855087387346607,
+      "learning_rate": 8.3864713239575e-06,
+      "loss": 0.9485,
+      "step": 17935
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.760669896597583,
+      "learning_rate": 8.38526530086238e-06,
+      "loss": 0.9504,
+      "step": 17940
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.160041078365477,
+      "learning_rate": 8.384058914005559e-06,
+      "loss": 1.002,
+      "step": 17945
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.497187362194074,
+      "learning_rate": 8.382852163516674e-06,
+      "loss": 0.918,
+      "step": 17950
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.454154753435161,
+      "learning_rate": 8.381645049525397e-06,
+      "loss": 0.9583,
+      "step": 17955
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.243586210893909,
+      "learning_rate": 8.380437572161433e-06,
+      "loss": 0.9782,
+      "step": 17960
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.544112672005654,
+      "learning_rate": 8.379229731554535e-06,
+      "loss": 0.9446,
+      "step": 17965
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.091305152673275,
+      "learning_rate": 8.378021527834486e-06,
+      "loss": 0.9595,
+      "step": 17970
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.483751533615201,
+      "learning_rate": 8.376812961131116e-06,
+      "loss": 0.9638,
+      "step": 17975
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.784214784890503,
+      "learning_rate": 8.37560403157429e-06,
+      "loss": 0.9441,
+      "step": 17980
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 23.59859939988344,
+      "learning_rate": 8.374394739293915e-06,
+      "loss": 0.974,
+      "step": 17985
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.229782337216466,
+      "learning_rate": 8.37318508441993e-06,
+      "loss": 1.0004,
+      "step": 17990
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.72974584502807,
+      "learning_rate": 8.371975067082323e-06,
+      "loss": 0.9712,
+      "step": 17995
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 21.746766732841987,
+      "learning_rate": 8.370764687411112e-06,
+      "loss": 0.9699,
+      "step": 18000
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.286187714329804,
+      "learning_rate": 8.369553945536362e-06,
+      "loss": 1.003,
+      "step": 18005
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.928871579151746,
+      "learning_rate": 8.368342841588166e-06,
+      "loss": 0.9202,
+      "step": 18010
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.274170476041927,
+      "learning_rate": 8.36713137569667e-06,
+      "loss": 0.9862,
+      "step": 18015
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.512660323560134,
+      "learning_rate": 8.365919547992044e-06,
+      "loss": 0.9561,
+      "step": 18020
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.195093630974645,
+      "learning_rate": 8.364707358604507e-06,
+      "loss": 0.924,
+      "step": 18025
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.1052810062634,
+      "learning_rate": 8.363494807664317e-06,
+      "loss": 0.9424,
+      "step": 18030
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.602480660263411,
+      "learning_rate": 8.362281895301764e-06,
+      "loss": 0.9831,
+      "step": 18035
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.421941861714285,
+      "learning_rate": 8.361068621647183e-06,
+      "loss": 0.9708,
+      "step": 18040
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.941533958485444,
+      "learning_rate": 8.359854986830946e-06,
+      "loss": 0.9572,
+      "step": 18045
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.174726028923948,
+      "learning_rate": 8.358640990983462e-06,
+      "loss": 0.9572,
+      "step": 18050
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.8112198563692115,
+      "learning_rate": 8.357426634235181e-06,
+      "loss": 0.9365,
+      "step": 18055
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.010485347737228,
+      "learning_rate": 8.356211916716591e-06,
+      "loss": 0.9303,
+      "step": 18060
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.629349724775571,
+      "learning_rate": 8.354996838558216e-06,
+      "loss": 1.0019,
+      "step": 18065
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.8006384018527415,
+      "learning_rate": 8.353781399890627e-06,
+      "loss": 0.955,
+      "step": 18070
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.691838965415852,
+      "learning_rate": 8.352565600844423e-06,
+      "loss": 0.9841,
+      "step": 18075
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 17.90274075326026,
+      "learning_rate": 8.35134944155025e-06,
+      "loss": 0.9797,
+      "step": 18080
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.722774412280813,
+      "learning_rate": 8.35013292213879e-06,
+      "loss": 0.9849,
+      "step": 18085
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.918455321710201,
+      "learning_rate": 8.34891604274076e-06,
+      "loss": 0.9636,
+      "step": 18090
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 17.268835468399153,
+      "learning_rate": 8.347698803486923e-06,
+      "loss": 0.9758,
+      "step": 18095
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 30.127043018401352,
+      "learning_rate": 8.346481204508074e-06,
+      "loss": 0.919,
+      "step": 18100
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 39.89111038215658,
+      "learning_rate": 8.34526324593505e-06,
+      "loss": 1.0071,
+      "step": 18105
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.074029187578066,
+      "learning_rate": 8.34404492789873e-06,
+      "loss": 0.9443,
+      "step": 18110
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 21.07756951105243,
+      "learning_rate": 8.34282625053002e-06,
+      "loss": 0.9777,
+      "step": 18115
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.261502736007123,
+      "learning_rate": 8.341607213959878e-06,
+      "loss": 0.9467,
+      "step": 18120
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 45.67904376279292,
+      "learning_rate": 8.340387818319293e-06,
+      "loss": 0.9187,
+      "step": 18125
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 24.91464582420456,
+      "learning_rate": 8.339168063739296e-06,
+      "loss": 0.9602,
+      "step": 18130
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 20.608723898861903,
+      "learning_rate": 8.337947950350952e-06,
+      "loss": 0.9051,
+      "step": 18135
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 30.400324583792425,
+      "learning_rate": 8.33672747828537e-06,
+      "loss": 0.9359,
+      "step": 18140
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 17.50665466740123,
+      "learning_rate": 8.335506647673694e-06,
+      "loss": 0.9758,
+      "step": 18145
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 39.0997047554433,
+      "learning_rate": 8.334285458647107e-06,
+      "loss": 0.9474,
+      "step": 18150
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.64631163623313,
+      "learning_rate": 8.333063911336834e-06,
+      "loss": 0.9864,
+      "step": 18155
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 30.842942128044452,
+      "learning_rate": 8.331842005874133e-06,
+      "loss": 0.9267,
+      "step": 18160
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.724000587486835,
+      "learning_rate": 8.330619742390305e-06,
+      "loss": 0.9693,
+      "step": 18165
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 28.029245836093526,
+      "learning_rate": 8.329397121016685e-06,
+      "loss": 0.9846,
+      "step": 18170
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.986480048543006,
+      "learning_rate": 8.32817414188465e-06,
+      "loss": 0.9658,
+      "step": 18175
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 14.736614094456346,
+      "learning_rate": 8.326950805125616e-06,
+      "loss": 0.9817,
+      "step": 18180
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.629078013752862,
+      "learning_rate": 8.325727110871037e-06,
+      "loss": 0.9667,
+      "step": 18185
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 25.056818404039877,
+      "learning_rate": 8.3245030592524e-06,
+      "loss": 0.9677,
+      "step": 18190
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.78687448613767,
+      "learning_rate": 8.323278650401239e-06,
+      "loss": 0.9261,
+      "step": 18195
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.668641636699055,
+      "learning_rate": 8.322053884449118e-06,
+      "loss": 0.9773,
+      "step": 18200
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.144517806073648,
+      "learning_rate": 8.320828761527646e-06,
+      "loss": 0.9642,
+      "step": 18205
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 19.87183113438539,
+      "learning_rate": 8.319603281768468e-06,
+      "loss": 0.9635,
+      "step": 18210
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.087429965457005,
+      "learning_rate": 8.31837744530327e-06,
+      "loss": 0.9548,
+      "step": 18215
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 60.09733804016439,
+      "learning_rate": 8.317151252263767e-06,
+      "loss": 0.93,
+      "step": 18220
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 39.00904417089919,
+      "learning_rate": 8.315924702781722e-06,
+      "loss": 0.9532,
+      "step": 18225
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.029073412497254,
+      "learning_rate": 8.314697796988935e-06,
+      "loss": 0.9248,
+      "step": 18230
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 42.28286385404939,
+      "learning_rate": 8.313470535017238e-06,
+      "loss": 0.9838,
+      "step": 18235
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.955384374997312,
+      "learning_rate": 8.312242916998512e-06,
+      "loss": 0.9115,
+      "step": 18240
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 34.29604175273556,
+      "learning_rate": 8.311014943064666e-06,
+      "loss": 0.9836,
+      "step": 18245
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.155447289367002,
+      "learning_rate": 8.30978661334765e-06,
+      "loss": 0.9535,
+      "step": 18250
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 22.786695422905964,
+      "learning_rate": 8.308557927979457e-06,
+      "loss": 0.9502,
+      "step": 18255
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.835472742132069,
+      "learning_rate": 8.307328887092113e-06,
+      "loss": 0.9517,
+      "step": 18260
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 23.41340268501469,
+      "learning_rate": 8.30609949081768e-06,
+      "loss": 1.0081,
+      "step": 18265
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 14.790843337054666,
+      "learning_rate": 8.30486973928827e-06,
+      "loss": 0.9768,
+      "step": 18270
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.022962350746162,
+      "learning_rate": 8.30363963263602e-06,
+      "loss": 0.9343,
+      "step": 18275
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.901891764383109,
+      "learning_rate": 8.302409170993111e-06,
+      "loss": 0.9641,
+      "step": 18280
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.76624124271663,
+      "learning_rate": 8.301178354491761e-06,
+      "loss": 0.8977,
+      "step": 18285
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 21.27823795057321,
+      "learning_rate": 8.299947183264228e-06,
+      "loss": 0.9807,
+      "step": 18290
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 21.249876163768274,
+      "learning_rate": 8.298715657442806e-06,
+      "loss": 0.9157,
+      "step": 18295
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.004007173788033,
+      "learning_rate": 8.297483777159832e-06,
+      "loss": 0.9421,
+      "step": 18300
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.392554791068047,
+      "learning_rate": 8.296251542547669e-06,
+      "loss": 0.986,
+      "step": 18305
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.615683155154661,
+      "learning_rate": 8.295018953738731e-06,
+      "loss": 0.9661,
+      "step": 18310
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.114232138557812,
+      "learning_rate": 8.293786010865466e-06,
+      "loss": 0.988,
+      "step": 18315
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.30042125520741,
+      "learning_rate": 8.292552714060356e-06,
+      "loss": 0.9409,
+      "step": 18320
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.5353023364251,
+      "learning_rate": 8.291319063455926e-06,
+      "loss": 0.9558,
+      "step": 18325
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 25.610488752863407,
+      "learning_rate": 8.290085059184738e-06,
+      "loss": 0.9496,
+      "step": 18330
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 23.270991061708862,
+      "learning_rate": 8.288850701379389e-06,
+      "loss": 0.9814,
+      "step": 18335
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.308313373842472,
+      "learning_rate": 8.287615990172518e-06,
+      "loss": 0.9542,
+      "step": 18340
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 19.33781984905678,
+      "learning_rate": 8.286380925696798e-06,
+      "loss": 0.967,
+      "step": 18345
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.208328854455573,
+      "learning_rate": 8.285145508084946e-06,
+      "loss": 0.9375,
+      "step": 18350
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 22.416980708087756,
+      "learning_rate": 8.283909737469708e-06,
+      "loss": 0.9893,
+      "step": 18355
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.192097478384558,
+      "learning_rate": 8.282673613983879e-06,
+      "loss": 0.9131,
+      "step": 18360
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.378178719432283,
+      "learning_rate": 8.28143713776028e-06,
+      "loss": 0.9457,
+      "step": 18365
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.290966136744645,
+      "learning_rate": 8.28020030893178e-06,
+      "loss": 0.9822,
+      "step": 18370
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.307433316113864,
+      "learning_rate": 8.278963127631279e-06,
+      "loss": 1.0093,
+      "step": 18375
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.957628210901851,
+      "learning_rate": 8.27772559399172e-06,
+      "loss": 0.9393,
+      "step": 18380
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.356604342839569,
+      "learning_rate": 8.276487708146079e-06,
+      "loss": 0.9358,
+      "step": 18385
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.7197882949054195,
+      "learning_rate": 8.275249470227375e-06,
+      "loss": 1.0173,
+      "step": 18390
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.923678298763505,
+      "learning_rate": 8.274010880368659e-06,
+      "loss": 0.9534,
+      "step": 18395
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.614201875233425,
+      "learning_rate": 8.272771938703026e-06,
+      "loss": 0.9382,
+      "step": 18400
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.504115565168725,
+      "learning_rate": 8.271532645363603e-06,
+      "loss": 0.959,
+      "step": 18405
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.04604956495506,
+      "learning_rate": 8.270293000483558e-06,
+      "loss": 0.9537,
+      "step": 18410
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.5964264965602855,
+      "learning_rate": 8.269053004196101e-06,
+      "loss": 0.9786,
+      "step": 18415
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 23.66660192681161,
+      "learning_rate": 8.26781265663447e-06,
+      "loss": 0.9706,
+      "step": 18420
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.796563103202072,
+      "learning_rate": 8.266571957931944e-06,
+      "loss": 0.9459,
+      "step": 18425
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.570452622301191,
+      "learning_rate": 8.265330908221847e-06,
+      "loss": 0.9281,
+      "step": 18430
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.176551398651894,
+      "learning_rate": 8.264089507637532e-06,
+      "loss": 0.9126,
+      "step": 18435
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 14.338693718246894,
+      "learning_rate": 8.262847756312395e-06,
+      "loss": 0.9521,
+      "step": 18440
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.637525785879936,
+      "learning_rate": 8.261605654379867e-06,
+      "loss": 1.017,
+      "step": 18445
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 16.01039001176703,
+      "learning_rate": 8.260363201973414e-06,
+      "loss": 0.9709,
+      "step": 18450
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.14051163935431,
+      "learning_rate": 8.259120399226549e-06,
+      "loss": 0.997,
+      "step": 18455
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.98945406377979,
+      "learning_rate": 8.257877246272812e-06,
+      "loss": 0.9537,
+      "step": 18460
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.090962525715208,
+      "learning_rate": 8.256633743245785e-06,
+      "loss": 0.9723,
+      "step": 18465
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.515364227793765,
+      "learning_rate": 8.255389890279092e-06,
+      "loss": 0.9487,
+      "step": 18470
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 14.510355697987137,
+      "learning_rate": 8.25414568750639e-06,
+      "loss": 0.9522,
+      "step": 18475
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.103611961373405,
+      "learning_rate": 8.252901135061368e-06,
+      "loss": 0.988,
+      "step": 18480
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.770202819196264,
+      "learning_rate": 8.251656233077765e-06,
+      "loss": 0.9149,
+      "step": 18485
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.991801820714467,
+      "learning_rate": 8.25041098168935e-06,
+      "loss": 0.9593,
+      "step": 18490
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.650659264452085,
+      "learning_rate": 8.24916538102993e-06,
+      "loss": 0.9414,
+      "step": 18495
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.966841558848007,
+      "learning_rate": 8.247919431233348e-06,
+      "loss": 0.955,
+      "step": 18500
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 15.225653690132798,
+      "learning_rate": 8.24667313243349e-06,
+      "loss": 0.9534,
+      "step": 18505
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.631953567911231,
+      "learning_rate": 8.245426484764276e-06,
+      "loss": 0.8882,
+      "step": 18510
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.726119781198639,
+      "learning_rate": 8.244179488359662e-06,
+      "loss": 1.0197,
+      "step": 18515
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 16.174482327233644,
+      "learning_rate": 8.242932143353644e-06,
+      "loss": 0.9582,
+      "step": 18520
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.534930144614199,
+      "learning_rate": 8.241684449880256e-06,
+      "loss": 0.976,
+      "step": 18525
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.856211269291136,
+      "learning_rate": 8.240436408073567e-06,
+      "loss": 0.9495,
+      "step": 18530
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.98076858251849,
+      "learning_rate": 8.239188018067685e-06,
+      "loss": 1.0105,
+      "step": 18535
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.07406647910944,
+      "learning_rate": 8.237939279996756e-06,
+      "loss": 0.8891,
+      "step": 18540
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.515147583058341,
+      "learning_rate": 8.236690193994958e-06,
+      "loss": 0.9179,
+      "step": 18545
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.714080446754478,
+      "learning_rate": 8.235440760196515e-06,
+      "loss": 0.9494,
+      "step": 18550
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.492272147032065,
+      "learning_rate": 8.234190978735685e-06,
+      "loss": 0.9619,
+      "step": 18555
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.568265001262244,
+      "learning_rate": 8.23294084974676e-06,
+      "loss": 0.9696,
+      "step": 18560
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.3419706985076845,
+      "learning_rate": 8.231690373364073e-06,
+      "loss": 1.0006,
+      "step": 18565
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.564488376137859,
+      "learning_rate": 8.23043954972199e-06,
+      "loss": 0.9454,
+      "step": 18570
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.048791559594555,
+      "learning_rate": 8.229188378954924e-06,
+      "loss": 0.9429,
+      "step": 18575
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.597372581230177,
+      "learning_rate": 8.227936861197315e-06,
+      "loss": 0.8992,
+      "step": 18580
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.154848619959518,
+      "learning_rate": 8.226684996583643e-06,
+      "loss": 0.9725,
+      "step": 18585
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.668712866256631,
+      "learning_rate": 8.225432785248428e-06,
+      "loss": 0.924,
+      "step": 18590
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 14.1545260742609,
+      "learning_rate": 8.224180227326225e-06,
+      "loss": 0.9462,
+      "step": 18595
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.001727228534417,
+      "learning_rate": 8.222927322951627e-06,
+      "loss": 0.9561,
+      "step": 18600
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.339464771272855,
+      "learning_rate": 8.221674072259264e-06,
+      "loss": 0.9805,
+      "step": 18605
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.465257003196277,
+      "learning_rate": 8.220420475383806e-06,
+      "loss": 0.9831,
+      "step": 18610
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.309880652239877,
+      "learning_rate": 8.219166532459952e-06,
+      "loss": 0.9339,
+      "step": 18615
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.303498663406126,
+      "learning_rate": 8.217912243622448e-06,
+      "loss": 0.9385,
+      "step": 18620
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.601022428278913,
+      "learning_rate": 8.216657609006074e-06,
+      "loss": 0.9489,
+      "step": 18625
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.263553704963337,
+      "learning_rate": 8.215402628745641e-06,
+      "loss": 0.9579,
+      "step": 18630
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.934401613463754,
+      "learning_rate": 8.214147302976006e-06,
+      "loss": 0.939,
+      "step": 18635
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.7049324252921885,
+      "learning_rate": 8.212891631832056e-06,
+      "loss": 0.9588,
+      "step": 18640
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.209531753799407,
+      "learning_rate": 8.211635615448724e-06,
+      "loss": 0.9399,
+      "step": 18645
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.827967546412326,
+      "learning_rate": 8.210379253960968e-06,
+      "loss": 0.9697,
+      "step": 18650
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.527337209997317,
+      "learning_rate": 8.209122547503794e-06,
+      "loss": 0.928,
+      "step": 18655
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.3975283433171315,
+      "learning_rate": 8.20786549621224e-06,
+      "loss": 0.9518,
+      "step": 18660
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.86373472604526,
+      "learning_rate": 8.206608100221381e-06,
+      "loss": 0.9493,
+      "step": 18665
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.305465192759861,
+      "learning_rate": 8.205350359666328e-06,
+      "loss": 0.9427,
+      "step": 18670
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.213564348863166,
+      "learning_rate": 8.204092274682234e-06,
+      "loss": 0.9364,
+      "step": 18675
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.333924477842656,
+      "learning_rate": 8.202833845404283e-06,
+      "loss": 0.9285,
+      "step": 18680
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.860249703771888,
+      "learning_rate": 8.201575071967702e-06,
+      "loss": 0.9049,
+      "step": 18685
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.474337294939785,
+      "learning_rate": 8.20031595450775e-06,
+      "loss": 0.9035,
+      "step": 18690
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.196275356019304,
+      "learning_rate": 8.199056493159723e-06,
+      "loss": 0.964,
+      "step": 18695
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.4508756345650315,
+      "learning_rate": 8.197796688058958e-06,
+      "loss": 0.9324,
+      "step": 18700
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.844182868137905,
+      "learning_rate": 8.196536539340826e-06,
+      "loss": 0.9888,
+      "step": 18705
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 18.547983933573764,
+      "learning_rate": 8.195276047140736e-06,
+      "loss": 1.0005,
+      "step": 18710
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 13.488927970940656,
+      "learning_rate": 8.194015211594132e-06,
+      "loss": 0.9491,
+      "step": 18715
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.705863112768338,
+      "learning_rate": 8.192754032836497e-06,
+      "loss": 0.9034,
+      "step": 18720
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.145966580637015,
+      "learning_rate": 8.191492511003352e-06,
+      "loss": 0.9655,
+      "step": 18725
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.20031713580108,
+      "learning_rate": 8.19023064623025e-06,
+      "loss": 0.9553,
+      "step": 18730
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.639991001447758,
+      "learning_rate": 8.188968438652786e-06,
+      "loss": 0.9624,
+      "step": 18735
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 21.608267164285888,
+      "learning_rate": 8.18770588840659e-06,
+      "loss": 0.9002,
+      "step": 18740
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.945583228113203,
+      "learning_rate": 8.186442995627326e-06,
+      "loss": 0.9304,
+      "step": 18745
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 28.34890652585614,
+      "learning_rate": 8.185179760450701e-06,
+      "loss": 0.9739,
+      "step": 18750
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.069300970650084,
+      "learning_rate": 8.183916183012454e-06,
+      "loss": 0.8806,
+      "step": 18755
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 36.11615174510507,
+      "learning_rate": 8.18265226344836e-06,
+      "loss": 0.9402,
+      "step": 18760
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.215858061573503,
+      "learning_rate": 8.181388001894235e-06,
+      "loss": 0.9577,
+      "step": 18765
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 22.219075572387343,
+      "learning_rate": 8.180123398485928e-06,
+      "loss": 0.9962,
+      "step": 18770
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 30.866794792385654,
+      "learning_rate": 8.178858453359325e-06,
+      "loss": 0.9723,
+      "step": 18775
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.349053682449533,
+      "learning_rate": 8.177593166650355e-06,
+      "loss": 1.0173,
+      "step": 18780
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 18.764661962599007,
+      "learning_rate": 8.176327538494974e-06,
+      "loss": 0.9737,
+      "step": 18785
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.157296726846901,
+      "learning_rate": 8.175061569029184e-06,
+      "loss": 0.967,
+      "step": 18790
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.996580215077445,
+      "learning_rate": 8.17379525838901e-06,
+      "loss": 0.9716,
+      "step": 18795
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.957251562807729,
+      "learning_rate": 8.172528606710532e-06,
+      "loss": 0.9563,
+      "step": 18800
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.233281480523723,
+      "learning_rate": 8.171261614129853e-06,
+      "loss": 0.9245,
+      "step": 18805
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.3928794184713915,
+      "learning_rate": 8.169994280783118e-06,
+      "loss": 0.9789,
+      "step": 18810
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.849178797058468,
+      "learning_rate": 8.168726606806509e-06,
+      "loss": 0.941,
+      "step": 18815
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.978465793251911,
+      "learning_rate": 8.16745859233624e-06,
+      "loss": 0.9328,
+      "step": 18820
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.995401138400932,
+      "learning_rate": 8.166190237508564e-06,
+      "loss": 0.976,
+      "step": 18825
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.280665478113042,
+      "learning_rate": 8.164921542459775e-06,
+      "loss": 0.9521,
+      "step": 18830
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.471328976224394,
+      "learning_rate": 8.1636525073262e-06,
+      "loss": 0.9743,
+      "step": 18835
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.82507813033722,
+      "learning_rate": 8.1623831322442e-06,
+      "loss": 1.003,
+      "step": 18840
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.591937122079878,
+      "learning_rate": 8.161113417350175e-06,
+      "loss": 0.949,
+      "step": 18845
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.920017579385835,
+      "learning_rate": 8.159843362780561e-06,
+      "loss": 0.9835,
+      "step": 18850
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.151195001545371,
+      "learning_rate": 8.158572968671836e-06,
+      "loss": 0.9278,
+      "step": 18855
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.16530122699321,
+      "learning_rate": 8.157302235160503e-06,
+      "loss": 0.9589,
+      "step": 18860
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.795817208230676,
+      "learning_rate": 8.156031162383112e-06,
+      "loss": 0.9317,
+      "step": 18865
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.669749041217251,
+      "learning_rate": 8.154759750476243e-06,
+      "loss": 0.9238,
+      "step": 18870
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.223008124934531,
+      "learning_rate": 8.153487999576515e-06,
+      "loss": 0.9893,
+      "step": 18875
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.195875131379625,
+      "learning_rate": 8.152215909820585e-06,
+      "loss": 0.9693,
+      "step": 18880
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.460823359114162,
+      "learning_rate": 8.150943481345144e-06,
+      "loss": 0.9553,
+      "step": 18885
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.194043017006326,
+      "learning_rate": 8.149670714286918e-06,
+      "loss": 0.9884,
+      "step": 18890
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.370120778961474,
+      "learning_rate": 8.148397608782674e-06,
+      "loss": 0.9534,
+      "step": 18895
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.44001417886006,
+      "learning_rate": 8.147124164969215e-06,
+      "loss": 0.9226,
+      "step": 18900
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.367211857488165,
+      "learning_rate": 8.145850382983371e-06,
+      "loss": 0.97,
+      "step": 18905
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.771808296700199,
+      "learning_rate": 8.144576262962021e-06,
+      "loss": 0.9357,
+      "step": 18910
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.366267655412884,
+      "learning_rate": 8.143301805042075e-06,
+      "loss": 0.9368,
+      "step": 18915
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.554192544212176,
+      "learning_rate": 8.142027009360476e-06,
+      "loss": 0.9708,
+      "step": 18920
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.221417115603707,
+      "learning_rate": 8.14075187605421e-06,
+      "loss": 0.9195,
+      "step": 18925
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.9617101973673,
+      "learning_rate": 8.139476405260291e-06,
+      "loss": 0.9513,
+      "step": 18930
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.965495849954547,
+      "learning_rate": 8.138200597115777e-06,
+      "loss": 0.9652,
+      "step": 18935
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.128574959625924,
+      "learning_rate": 8.13692445175776e-06,
+      "loss": 0.9638,
+      "step": 18940
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.027231065166058,
+      "learning_rate": 8.135647969323365e-06,
+      "loss": 0.9341,
+      "step": 18945
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.670297375594766,
+      "learning_rate": 8.13437114994976e-06,
+      "loss": 0.906,
+      "step": 18950
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.137314314312876,
+      "learning_rate": 8.133093993774139e-06,
+      "loss": 0.9647,
+      "step": 18955
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.7281437061946265,
+      "learning_rate": 8.131816500933743e-06,
+      "loss": 0.9215,
+      "step": 18960
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 23.789460071290314,
+      "learning_rate": 8.130538671565842e-06,
+      "loss": 0.9854,
+      "step": 18965
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 30.65871741662231,
+      "learning_rate": 8.129260505807743e-06,
+      "loss": 0.9303,
+      "step": 18970
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.632581448626001,
+      "learning_rate": 8.127982003796793e-06,
+      "loss": 0.9178,
+      "step": 18975
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 17.618529898929967,
+      "learning_rate": 8.126703165670372e-06,
+      "loss": 1.0079,
+      "step": 18980
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.324825991342937,
+      "learning_rate": 8.125423991565897e-06,
+      "loss": 0.9703,
+      "step": 18985
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.506536800116093,
+      "learning_rate": 8.124144481620819e-06,
+      "loss": 0.9108,
+      "step": 18990
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.205021872546387,
+      "learning_rate": 8.12286463597263e-06,
+      "loss": 0.9596,
+      "step": 18995
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.12751175739675,
+      "learning_rate": 8.121584454758853e-06,
+      "loss": 0.9372,
+      "step": 19000
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.204895926529103,
+      "learning_rate": 8.12030393811705e-06,
+      "loss": 0.9751,
+      "step": 19005
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.641658872199885,
+      "learning_rate": 8.119023086184819e-06,
+      "loss": 0.9427,
+      "step": 19010
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.460254379406491,
+      "learning_rate": 8.117741899099792e-06,
+      "loss": 0.9642,
+      "step": 19015
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.496584675510707,
+      "learning_rate": 8.116460376999635e-06,
+      "loss": 0.9969,
+      "step": 19020
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.332517697442171,
+      "learning_rate": 8.11517852002206e-06,
+      "loss": 0.9598,
+      "step": 19025
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.114972784844499,
+      "learning_rate": 8.113896328304803e-06,
+      "loss": 0.9483,
+      "step": 19030
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.348835767403838,
+      "learning_rate": 8.112613801985643e-06,
+      "loss": 0.9471,
+      "step": 19035
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 30.686926455930507,
+      "learning_rate": 8.111330941202394e-06,
+      "loss": 0.9906,
+      "step": 19040
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 64.61487049476504,
+      "learning_rate": 8.110047746092902e-06,
+      "loss": 0.9905,
+      "step": 19045
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 26.250353886694356,
+      "learning_rate": 8.108764216795056e-06,
+      "loss": 0.9956,
+      "step": 19050
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 60.48397806017447,
+      "learning_rate": 8.107480353446774e-06,
+      "loss": 1.0264,
+      "step": 19055
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.496287744009525,
+      "learning_rate": 8.106196156186014e-06,
+      "loss": 0.9552,
+      "step": 19060
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 46.872428608283386,
+      "learning_rate": 8.104911625150769e-06,
+      "loss": 1.0021,
+      "step": 19065
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 12.489594361741183,
+      "learning_rate": 8.103626760479066e-06,
+      "loss": 0.9865,
+      "step": 19070
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 25.851004185686303,
+      "learning_rate": 8.10234156230897e-06,
+      "loss": 0.9875,
+      "step": 19075
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.685957859549898,
+      "learning_rate": 8.101056030778583e-06,
+      "loss": 0.961,
+      "step": 19080
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.472298903540388,
+      "learning_rate": 8.099770166026038e-06,
+      "loss": 0.965,
+      "step": 19085
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 15.045807887758574,
+      "learning_rate": 8.098483968189508e-06,
+      "loss": 1.0052,
+      "step": 19090
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.657552807974352,
+      "learning_rate": 8.0971974374072e-06,
+      "loss": 0.9837,
+      "step": 19095
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.024360037175807,
+      "learning_rate": 8.09591057381736e-06,
+      "loss": 0.9882,
+      "step": 19100
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.421729973972606,
+      "learning_rate": 8.094623377558267e-06,
+      "loss": 0.9836,
+      "step": 19105
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.870301147408619,
+      "learning_rate": 8.093335848768233e-06,
+      "loss": 0.9749,
+      "step": 19110
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.307985263352527,
+      "learning_rate": 8.09204798758561e-06,
+      "loss": 0.9897,
+      "step": 19115
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 30.308768334493212,
+      "learning_rate": 8.090759794148784e-06,
+      "loss": 0.9734,
+      "step": 19120
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.812322916269569,
+      "learning_rate": 8.089471268596179e-06,
+      "loss": 0.9719,
+      "step": 19125
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.281872072100057,
+      "learning_rate": 8.08818241106625e-06,
+      "loss": 0.9711,
+      "step": 19130
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.350191912606665,
+      "learning_rate": 8.086893221697492e-06,
+      "loss": 0.9444,
+      "step": 19135
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.905795692103847,
+      "learning_rate": 8.085603700628433e-06,
+      "loss": 0.9115,
+      "step": 19140
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.494627237059104,
+      "learning_rate": 8.08431384799764e-06,
+      "loss": 0.8975,
+      "step": 19145
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.504272272834612,
+      "learning_rate": 8.083023663943711e-06,
+      "loss": 0.9935,
+      "step": 19150
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.466082915234127,
+      "learning_rate": 8.081733148605284e-06,
+      "loss": 0.9135,
+      "step": 19155
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.128940782938963,
+      "learning_rate": 8.080442302121028e-06,
+      "loss": 0.9419,
+      "step": 19160
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.905885961095287,
+      "learning_rate": 8.079151124629651e-06,
+      "loss": 0.9847,
+      "step": 19165
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.14963058167278,
+      "learning_rate": 8.077859616269899e-06,
+      "loss": 0.9679,
+      "step": 19170
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.156502034570472,
+      "learning_rate": 8.076567777180547e-06,
+      "loss": 0.9999,
+      "step": 19175
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.999590755179119,
+      "learning_rate": 8.075275607500407e-06,
+      "loss": 0.9442,
+      "step": 19180
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.474147826255185,
+      "learning_rate": 8.073983107368334e-06,
+      "loss": 0.9343,
+      "step": 19185
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.978192324670545,
+      "learning_rate": 8.072690276923208e-06,
+      "loss": 0.9922,
+      "step": 19190
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.040302910399815,
+      "learning_rate": 8.07139711630395e-06,
+      "loss": 0.9598,
+      "step": 19195
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.3148028608519375,
+      "learning_rate": 8.070103625649518e-06,
+      "loss": 0.9404,
+      "step": 19200
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 27.689607326731224,
+      "learning_rate": 8.068809805098903e-06,
+      "loss": 0.9692,
+      "step": 19205
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.275963124357617,
+      "learning_rate": 8.06751565479113e-06,
+      "loss": 0.9202,
+      "step": 19210
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.496705319574194,
+      "learning_rate": 8.066221174865261e-06,
+      "loss": 0.9988,
+      "step": 19215
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 31.37702244184853,
+      "learning_rate": 8.064926365460394e-06,
+      "loss": 0.9333,
+      "step": 19220
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.7848024561920335,
+      "learning_rate": 8.063631226715663e-06,
+      "loss": 0.8728,
+      "step": 19225
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.981401505795201,
+      "learning_rate": 8.062335758770237e-06,
+      "loss": 0.9123,
+      "step": 19230
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.463761471424651,
+      "learning_rate": 8.061039961763317e-06,
+      "loss": 0.9598,
+      "step": 19235
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.227621295533597,
+      "learning_rate": 8.059743835834146e-06,
+      "loss": 0.9562,
+      "step": 19240
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 22.251037260477297,
+      "learning_rate": 8.058447381121994e-06,
+      "loss": 0.9289,
+      "step": 19245
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.459511512872369,
+      "learning_rate": 8.057150597766175e-06,
+      "loss": 0.9958,
+      "step": 19250
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.330265059106776,
+      "learning_rate": 8.055853485906029e-06,
+      "loss": 0.9903,
+      "step": 19255
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.670618505402853,
+      "learning_rate": 8.054556045680941e-06,
+      "loss": 0.9569,
+      "step": 19260
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 16.362562558245227,
+      "learning_rate": 8.053258277230324e-06,
+      "loss": 0.9019,
+      "step": 19265
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.88995697002873,
+      "learning_rate": 8.051960180693633e-06,
+      "loss": 0.9523,
+      "step": 19270
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.741444655475432,
+      "learning_rate": 8.05066175621035e-06,
+      "loss": 0.9264,
+      "step": 19275
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.072142501407967,
+      "learning_rate": 8.049363003919997e-06,
+      "loss": 0.946,
+      "step": 19280
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.649458524454594,
+      "learning_rate": 8.04806392396213e-06,
+      "loss": 0.9676,
+      "step": 19285
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.008869414362337,
+      "learning_rate": 8.046764516476344e-06,
+      "loss": 0.9662,
+      "step": 19290
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.940476007854292,
+      "learning_rate": 8.045464781602265e-06,
+      "loss": 0.9511,
+      "step": 19295
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.533736787794461,
+      "learning_rate": 8.044164719479555e-06,
+      "loss": 0.9082,
+      "step": 19300
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.97256028421361,
+      "learning_rate": 8.042864330247908e-06,
+      "loss": 0.9621,
+      "step": 19305
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.719141334024216,
+      "learning_rate": 8.041563614047062e-06,
+      "loss": 0.965,
+      "step": 19310
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.192133825637401,
+      "learning_rate": 8.040262571016781e-06,
+      "loss": 0.959,
+      "step": 19315
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.743848115116697,
+      "learning_rate": 8.03896120129687e-06,
+      "loss": 0.944,
+      "step": 19320
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.112480487988186,
+      "learning_rate": 8.037659505027166e-06,
+      "loss": 0.9175,
+      "step": 19325
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.368858588572888,
+      "learning_rate": 8.03635748234754e-06,
+      "loss": 0.9366,
+      "step": 19330
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.698209830999122,
+      "learning_rate": 8.035055133397905e-06,
+      "loss": 0.9499,
+      "step": 19335
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.464124207181301,
+      "learning_rate": 8.033752458318199e-06,
+      "loss": 0.9452,
+      "step": 19340
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.195499561112465,
+      "learning_rate": 8.032449457248402e-06,
+      "loss": 0.9832,
+      "step": 19345
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 30.492675598290408,
+      "learning_rate": 8.031146130328529e-06,
+      "loss": 0.9359,
+      "step": 19350
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.567005081859069,
+      "learning_rate": 8.029842477698627e-06,
+      "loss": 0.9332,
+      "step": 19355
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.461193392688758,
+      "learning_rate": 8.02853849949878e-06,
+      "loss": 1.0088,
+      "step": 19360
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.56565444359108,
+      "learning_rate": 8.027234195869103e-06,
+      "loss": 0.9748,
+      "step": 19365
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 15.050053366170115,
+      "learning_rate": 8.025929566949751e-06,
+      "loss": 0.9733,
+      "step": 19370
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.653637879963531,
+      "learning_rate": 8.024624612880913e-06,
+      "loss": 0.9393,
+      "step": 19375
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.798955707180229,
+      "learning_rate": 8.023319333802813e-06,
+      "loss": 0.9355,
+      "step": 19380
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.565570627134732,
+      "learning_rate": 8.022013729855708e-06,
+      "loss": 0.9703,
+      "step": 19385
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.826299823673436,
+      "learning_rate": 8.02070780117989e-06,
+      "loss": 0.9325,
+      "step": 19390
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.0159554325501094,
+      "learning_rate": 8.019401547915685e-06,
+      "loss": 0.9844,
+      "step": 19395
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.261569313375878,
+      "learning_rate": 8.018094970203461e-06,
+      "loss": 0.9494,
+      "step": 19400
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.004110579814553,
+      "learning_rate": 8.016788068183613e-06,
+      "loss": 0.9833,
+      "step": 19405
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.451316943774587,
+      "learning_rate": 8.01548084199657e-06,
+      "loss": 0.967,
+      "step": 19410
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.630238035382916,
+      "learning_rate": 8.014173291782806e-06,
+      "loss": 0.9702,
+      "step": 19415
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 24.126467039884865,
+      "learning_rate": 8.012865417682815e-06,
+      "loss": 0.951,
+      "step": 19420
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.683487608165553,
+      "learning_rate": 8.011557219837141e-06,
+      "loss": 0.9963,
+      "step": 19425
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.349705191691237,
+      "learning_rate": 8.010248698386352e-06,
+      "loss": 0.9462,
+      "step": 19430
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.711331585817193,
+      "learning_rate": 8.008939853471055e-06,
+      "loss": 0.9605,
+      "step": 19435
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.98077388159023,
+      "learning_rate": 8.007630685231892e-06,
+      "loss": 0.9497,
+      "step": 19440
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.91240858152615,
+      "learning_rate": 8.006321193809538e-06,
+      "loss": 0.9629,
+      "step": 19445
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.066176753467188,
+      "learning_rate": 8.005011379344702e-06,
+      "loss": 0.9564,
+      "step": 19450
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.698201041761696,
+      "learning_rate": 8.003701241978133e-06,
+      "loss": 1.0054,
+      "step": 19455
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.421053631687196,
+      "learning_rate": 8.002390781850608e-06,
+      "loss": 0.9293,
+      "step": 19460
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.13309559011947,
+      "learning_rate": 8.001079999102944e-06,
+      "loss": 0.953,
+      "step": 19465
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.496890683342123,
+      "learning_rate": 7.99976889387599e-06,
+      "loss": 0.9352,
+      "step": 19470
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 14.474961844980022,
+      "learning_rate": 7.998457466310627e-06,
+      "loss": 0.9317,
+      "step": 19475
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.439084447002402,
+      "learning_rate": 7.997145716547777e-06,
+      "loss": 0.992,
+      "step": 19480
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 24.455108082467582,
+      "learning_rate": 7.995833644728392e-06,
+      "loss": 0.9719,
+      "step": 19485
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.766452783205073,
+      "learning_rate": 7.994521250993462e-06,
+      "loss": 0.9482,
+      "step": 19490
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 36.68519370259944,
+      "learning_rate": 7.993208535484007e-06,
+      "loss": 0.9433,
+      "step": 19495
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.86956301167978,
+      "learning_rate": 7.991895498341084e-06,
+      "loss": 0.9447,
+      "step": 19500
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 25.656579481434097,
+      "learning_rate": 7.990582139705788e-06,
+      "loss": 0.9273,
+      "step": 19505
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.770256508544099,
+      "learning_rate": 7.98926845971924e-06,
+      "loss": 0.9583,
+      "step": 19510
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.550009874566956,
+      "learning_rate": 7.987954458522602e-06,
+      "loss": 0.9801,
+      "step": 19515
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.906244823367583,
+      "learning_rate": 7.986640136257073e-06,
+      "loss": 0.8999,
+      "step": 19520
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.8235374691748785,
+      "learning_rate": 7.98532549306388e-06,
+      "loss": 0.9643,
+      "step": 19525
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.699269384621902,
+      "learning_rate": 7.984010529084288e-06,
+      "loss": 0.8983,
+      "step": 19530
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.227718940194455,
+      "learning_rate": 7.982695244459596e-06,
+      "loss": 0.9525,
+      "step": 19535
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.713407954850574,
+      "learning_rate": 7.981379639331133e-06,
+      "loss": 0.97,
+      "step": 19540
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.503803093212527,
+      "learning_rate": 7.980063713840271e-06,
+      "loss": 0.9277,
+      "step": 19545
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.05706236845638,
+      "learning_rate": 7.978747468128412e-06,
+      "loss": 0.9435,
+      "step": 19550
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.36127654938038,
+      "learning_rate": 7.977430902336992e-06,
+      "loss": 0.9456,
+      "step": 19555
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.428830630164985,
+      "learning_rate": 7.97611401660748e-06,
+      "loss": 0.9216,
+      "step": 19560
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.011920096570101,
+      "learning_rate": 7.974796811081382e-06,
+      "loss": 0.9783,
+      "step": 19565
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.199607915822191,
+      "learning_rate": 7.973479285900239e-06,
+      "loss": 0.8955,
+      "step": 19570
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.890940547817857,
+      "learning_rate": 7.972161441205623e-06,
+      "loss": 0.9498,
+      "step": 19575
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.70694765558659,
+      "learning_rate": 7.970843277139142e-06,
+      "loss": 0.9534,
+      "step": 19580
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.305971650187844,
+      "learning_rate": 7.969524793842442e-06,
+      "loss": 0.964,
+      "step": 19585
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.13277345477748,
+      "learning_rate": 7.968205991457196e-06,
+      "loss": 0.9355,
+      "step": 19590
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 15.825923515805263,
+      "learning_rate": 7.966886870125117e-06,
+      "loss": 0.9677,
+      "step": 19595
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.975135735608088,
+      "learning_rate": 7.96556742998795e-06,
+      "loss": 0.9352,
+      "step": 19600
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.591421233230836,
+      "learning_rate": 7.964247671187474e-06,
+      "loss": 0.9637,
+      "step": 19605
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.718634106108007,
+      "learning_rate": 7.962927593865504e-06,
+      "loss": 0.9085,
+      "step": 19610
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 16.955638635767677,
+      "learning_rate": 7.961607198163888e-06,
+      "loss": 0.9228,
+      "step": 19615
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 43.602053972987015,
+      "learning_rate": 7.960286484224506e-06,
+      "loss": 0.9427,
+      "step": 19620
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.16533227672473,
+      "learning_rate": 7.958965452189278e-06,
+      "loss": 0.9279,
+      "step": 19625
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.829481821488702,
+      "learning_rate": 7.957644102200153e-06,
+      "loss": 0.9973,
+      "step": 19630
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.970687551034499,
+      "learning_rate": 7.956322434399116e-06,
+      "loss": 0.9837,
+      "step": 19635
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.190972374786007,
+      "learning_rate": 7.955000448928188e-06,
+      "loss": 0.9744,
+      "step": 19640
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.588181451304402,
+      "learning_rate": 7.953678145929417e-06,
+      "loss": 0.926,
+      "step": 19645
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.381918459585516,
+      "learning_rate": 7.952355525544896e-06,
+      "loss": 0.9735,
+      "step": 19650
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.80693524917704,
+      "learning_rate": 7.951032587916742e-06,
+      "loss": 1.0078,
+      "step": 19655
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.3503338628040575,
+      "learning_rate": 7.949709333187116e-06,
+      "loss": 0.959,
+      "step": 19660
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.237006720925747,
+      "learning_rate": 7.9483857614982e-06,
+      "loss": 0.9856,
+      "step": 19665
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.816087659148016,
+      "learning_rate": 7.947061872992226e-06,
+      "loss": 0.9693,
+      "step": 19670
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.677939183677492,
+      "learning_rate": 7.945737667811444e-06,
+      "loss": 0.9488,
+      "step": 19675
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.773405593237955,
+      "learning_rate": 7.94441314609815e-06,
+      "loss": 0.8974,
+      "step": 19680
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.885395413739726,
+      "learning_rate": 7.94308830799467e-06,
+      "loss": 0.9466,
+      "step": 19685
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.230162314206446,
+      "learning_rate": 7.94176315364336e-06,
+      "loss": 0.9619,
+      "step": 19690
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.010389732362329,
+      "learning_rate": 7.94043768318662e-06,
+      "loss": 0.9265,
+      "step": 19695
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.326671113480245,
+      "learning_rate": 7.939111896766871e-06,
+      "loss": 0.9616,
+      "step": 19700
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.940817824927529,
+      "learning_rate": 7.937785794526579e-06,
+      "loss": 0.9498,
+      "step": 19705
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.123994633748377,
+      "learning_rate": 7.936459376608235e-06,
+      "loss": 0.9683,
+      "step": 19710
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.732863798232363,
+      "learning_rate": 7.935132643154372e-06,
+      "loss": 0.9558,
+      "step": 19715
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 21.54114130340925,
+      "learning_rate": 7.933805594307554e-06,
+      "loss": 0.9622,
+      "step": 19720
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.338570610213885,
+      "learning_rate": 7.932478230210379e-06,
+      "loss": 0.9975,
+      "step": 19725
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.1643460718766,
+      "learning_rate": 7.931150551005474e-06,
+      "loss": 0.9754,
+      "step": 19730
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.794767634573762,
+      "learning_rate": 7.929822556835504e-06,
+      "loss": 0.9992,
+      "step": 19735
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.358733421042912,
+      "learning_rate": 7.928494247843173e-06,
+      "loss": 0.962,
+      "step": 19740
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.441516126333959,
+      "learning_rate": 7.92716562417121e-06,
+      "loss": 0.9343,
+      "step": 19745
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.005697348918134,
+      "learning_rate": 7.925836685962381e-06,
+      "loss": 0.94,
+      "step": 19750
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.954779456667186,
+      "learning_rate": 7.92450743335949e-06,
+      "loss": 0.9297,
+      "step": 19755
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 21.96543735644006,
+      "learning_rate": 7.923177866505367e-06,
+      "loss": 0.9999,
+      "step": 19760
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.439114656591249,
+      "learning_rate": 7.92184798554288e-06,
+      "loss": 0.9381,
+      "step": 19765
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.437159688487576,
+      "learning_rate": 7.920517790614932e-06,
+      "loss": 0.9548,
+      "step": 19770
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.31164624093142,
+      "learning_rate": 7.91918728186446e-06,
+      "loss": 0.9739,
+      "step": 19775
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.568427451030278,
+      "learning_rate": 7.91785645943443e-06,
+      "loss": 0.9471,
+      "step": 19780
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 25.524585604423766,
+      "learning_rate": 7.916525323467847e-06,
+      "loss": 0.9716,
+      "step": 19785
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 35.19814614535263,
+      "learning_rate": 7.915193874107745e-06,
+      "loss": 0.8945,
+      "step": 19790
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.81733098741581,
+      "learning_rate": 7.913862111497197e-06,
+      "loss": 0.9891,
+      "step": 19795
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.765431762253932,
+      "learning_rate": 7.912530035779305e-06,
+      "loss": 0.9364,
+      "step": 19800
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.650495804124496,
+      "learning_rate": 7.911197647097205e-06,
+      "loss": 0.9527,
+      "step": 19805
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.759189748283042,
+      "learning_rate": 7.90986494559407e-06,
+      "loss": 0.9488,
+      "step": 19810
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.432197284445616,
+      "learning_rate": 7.908531931413106e-06,
+      "loss": 0.9614,
+      "step": 19815
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.144157754736919,
+      "learning_rate": 7.907198604697548e-06,
+      "loss": 0.9442,
+      "step": 19820
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.58622891155811,
+      "learning_rate": 7.90586496559067e-06,
+      "loss": 0.9407,
+      "step": 19825
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.256606032959596,
+      "learning_rate": 7.904531014235778e-06,
+      "loss": 0.9402,
+      "step": 19830
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 18.995752133576723,
+      "learning_rate": 7.90319675077621e-06,
+      "loss": 0.9276,
+      "step": 19835
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.08826606164148,
+      "learning_rate": 7.901862175355336e-06,
+      "loss": 0.9613,
+      "step": 19840
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.823681467049705,
+      "learning_rate": 7.900527288116567e-06,
+      "loss": 0.9438,
+      "step": 19845
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.458283910186196,
+      "learning_rate": 7.899192089203339e-06,
+      "loss": 0.9377,
+      "step": 19850
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 20.56108461369566,
+      "learning_rate": 7.897856578759125e-06,
+      "loss": 0.9592,
+      "step": 19855
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.117862529409534,
+      "learning_rate": 7.896520756927433e-06,
+      "loss": 0.9397,
+      "step": 19860
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.540676078214242,
+      "learning_rate": 7.895184623851802e-06,
+      "loss": 0.9409,
+      "step": 19865
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.922870762478146,
+      "learning_rate": 7.893848179675805e-06,
+      "loss": 0.945,
+      "step": 19870
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.026419476200202,
+      "learning_rate": 7.892511424543052e-06,
+      "loss": 0.961,
+      "step": 19875
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.098333436086616,
+      "learning_rate": 7.891174358597177e-06,
+      "loss": 0.9516,
+      "step": 19880
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.174111229345392,
+      "learning_rate": 7.889836981981862e-06,
+      "loss": 0.9242,
+      "step": 19885
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.583304008671309,
+      "learning_rate": 7.888499294840807e-06,
+      "loss": 0.9575,
+      "step": 19890
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.976771479838358,
+      "learning_rate": 7.887161297317756e-06,
+      "loss": 0.9566,
+      "step": 19895
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.552608534180628,
+      "learning_rate": 7.88582298955648e-06,
+      "loss": 0.9596,
+      "step": 19900
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.614054283870222,
+      "learning_rate": 7.884484371700788e-06,
+      "loss": 0.9002,
+      "step": 19905
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.567773616178767,
+      "learning_rate": 7.88314544389452e-06,
+      "loss": 0.9542,
+      "step": 19910
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.335168192737689,
+      "learning_rate": 7.88180620628155e-06,
+      "loss": 0.9427,
+      "step": 19915
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.476534656304151,
+      "learning_rate": 7.880466659005786e-06,
+      "loss": 0.9583,
+      "step": 19920
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.664795105240772,
+      "learning_rate": 7.879126802211165e-06,
+      "loss": 0.9185,
+      "step": 19925
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 26.065707131841684,
+      "learning_rate": 7.877786636041664e-06,
+      "loss": 0.912,
+      "step": 19930
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 17.60635611885136,
+      "learning_rate": 7.87644616064129e-06,
+      "loss": 0.9218,
+      "step": 19935
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.444135083948037,
+      "learning_rate": 7.875105376154078e-06,
+      "loss": 0.9589,
+      "step": 19940
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.963626629764778,
+      "learning_rate": 7.873764282724104e-06,
+      "loss": 0.9531,
+      "step": 19945
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.953867406038498,
+      "learning_rate": 7.872422880495476e-06,
+      "loss": 0.9422,
+      "step": 19950
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.281832810997978,
+      "learning_rate": 7.871081169612335e-06,
+      "loss": 0.9363,
+      "step": 19955
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.67327295981892,
+      "learning_rate": 7.869739150218849e-06,
+      "loss": 0.973,
+      "step": 19960
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.492111430740662,
+      "learning_rate": 7.868396822459227e-06,
+      "loss": 0.9387,
+      "step": 19965
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.82283212288816,
+      "learning_rate": 7.867054186477707e-06,
+      "loss": 0.9438,
+      "step": 19970
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.228653455876002,
+      "learning_rate": 7.865711242418561e-06,
+      "loss": 0.9475,
+      "step": 19975
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.178307091181455,
+      "learning_rate": 7.864367990426095e-06,
+      "loss": 0.9576,
+      "step": 19980
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.142175396053066,
+      "learning_rate": 7.863024430644647e-06,
+      "loss": 0.9847,
+      "step": 19985
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.199772689796203,
+      "learning_rate": 7.861680563218588e-06,
+      "loss": 0.9865,
+      "step": 19990
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.44718736131562,
+      "learning_rate": 7.860336388292324e-06,
+      "loss": 0.9453,
+      "step": 19995
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.565393743290027,
+      "learning_rate": 7.858991906010289e-06,
+      "loss": 0.9078,
+      "step": 20000
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.229254186701492,
+      "learning_rate": 7.857647116516959e-06,
+      "loss": 0.9054,
+      "step": 20005
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.699017336532892,
+      "learning_rate": 7.856302019956833e-06,
+      "loss": 0.9626,
+      "step": 20010
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.219672531293912,
+      "learning_rate": 7.85495661647445e-06,
+      "loss": 0.9441,
+      "step": 20015
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.426445358055107,
+      "learning_rate": 7.853610906214376e-06,
+      "loss": 0.9438,
+      "step": 20020
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.130086153119944,
+      "learning_rate": 7.852264889321219e-06,
+      "loss": 0.9131,
+      "step": 20025
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.950594775885904,
+      "learning_rate": 7.850918565939611e-06,
+      "loss": 0.9071,
+      "step": 20030
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.683234471569929,
+      "learning_rate": 7.849571936214221e-06,
+      "loss": 0.9831,
+      "step": 20035
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.918927231712517,
+      "learning_rate": 7.84822500028975e-06,
+      "loss": 0.9563,
+      "step": 20040
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.723994762414341,
+      "learning_rate": 7.846877758310933e-06,
+      "loss": 0.933,
+      "step": 20045
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.704931324078806,
+      "learning_rate": 7.845530210422537e-06,
+      "loss": 0.953,
+      "step": 20050
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 28.593829438922224,
+      "learning_rate": 7.84418235676936e-06,
+      "loss": 0.9777,
+      "step": 20055
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.884887942381204,
+      "learning_rate": 7.842834197496238e-06,
+      "loss": 0.9455,
+      "step": 20060
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.992617975755011,
+      "learning_rate": 7.841485732748035e-06,
+      "loss": 0.9018,
+      "step": 20065
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.638986263329572,
+      "learning_rate": 7.84013696266965e-06,
+      "loss": 0.9249,
+      "step": 20070
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.268277699407994,
+      "learning_rate": 7.838787887406012e-06,
+      "loss": 0.9544,
+      "step": 20075
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.252205588716973,
+      "learning_rate": 7.83743850710209e-06,
+      "loss": 0.9261,
+      "step": 20080
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 28.813291291039274,
+      "learning_rate": 7.836088821902874e-06,
+      "loss": 1.047,
+      "step": 20085
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.072403660756706,
+      "learning_rate": 7.8347388319534e-06,
+      "loss": 0.954,
+      "step": 20090
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.777456113095121,
+      "learning_rate": 7.833388537398728e-06,
+      "loss": 0.9591,
+      "step": 20095
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.021113854385633,
+      "learning_rate": 7.832037938383951e-06,
+      "loss": 0.9556,
+      "step": 20100
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.788626586060873,
+      "learning_rate": 7.830687035054202e-06,
+      "loss": 0.9563,
+      "step": 20105
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.205718897745614,
+      "learning_rate": 7.829335827554634e-06,
+      "loss": 0.9358,
+      "step": 20110
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.827859640256731,
+      "learning_rate": 7.827984316030447e-06,
+      "loss": 0.9815,
+      "step": 20115
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 18.496962409757696,
+      "learning_rate": 7.826632500626863e-06,
+      "loss": 0.9493,
+      "step": 20120
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.612688104431891,
+      "learning_rate": 7.825280381489145e-06,
+      "loss": 0.9321,
+      "step": 20125
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.301062602979739,
+      "learning_rate": 7.823927958762578e-06,
+      "loss": 1.0011,
+      "step": 20130
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.476625104434346,
+      "learning_rate": 7.822575232592491e-06,
+      "loss": 0.8694,
+      "step": 20135
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.1728480126820635,
+      "learning_rate": 7.821222203124238e-06,
+      "loss": 0.9675,
+      "step": 20140
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.186787079320183,
+      "learning_rate": 7.819868870503207e-06,
+      "loss": 0.9378,
+      "step": 20145
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.02753075537611,
+      "learning_rate": 7.81851523487482e-06,
+      "loss": 0.9412,
+      "step": 20150
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.966718193736909,
+      "learning_rate": 7.817161296384535e-06,
+      "loss": 0.9182,
+      "step": 20155
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.133481628479382,
+      "learning_rate": 7.815807055177833e-06,
+      "loss": 0.9422,
+      "step": 20160
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 23.655816673631616,
+      "learning_rate": 7.814452511400236e-06,
+      "loss": 0.8711,
+      "step": 20165
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 26.946670626715143,
+      "learning_rate": 7.813097665197298e-06,
+      "loss": 0.9733,
+      "step": 20170
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.224702603212419,
+      "learning_rate": 7.811742516714597e-06,
+      "loss": 0.9133,
+      "step": 20175
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 27.390459826633204,
+      "learning_rate": 7.810387066097757e-06,
+      "loss": 0.9404,
+      "step": 20180
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.010382622704545,
+      "learning_rate": 7.809031313492424e-06,
+      "loss": 0.8973,
+      "step": 20185
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 30.271574718468564,
+      "learning_rate": 7.807675259044276e-06,
+      "loss": 0.9459,
+      "step": 20190
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 16.320562856419855,
+      "learning_rate": 7.80631890289903e-06,
+      "loss": 0.9574,
+      "step": 20195
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 47.9206358179213,
+      "learning_rate": 7.804962245202435e-06,
+      "loss": 0.9591,
+      "step": 20200
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 20.098770244061356,
+      "learning_rate": 7.803605286100267e-06,
+      "loss": 0.9812,
+      "step": 20205
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 25.01687722783874,
+      "learning_rate": 7.802248025738337e-06,
+      "loss": 0.949,
+      "step": 20210
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 22.772754727830986,
+      "learning_rate": 7.80089046426249e-06,
+      "loss": 0.9643,
+      "step": 20215
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 20.757226003243538,
+      "learning_rate": 7.7995326018186e-06,
+      "loss": 0.9686,
+      "step": 20220
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 23.62684872274855,
+      "learning_rate": 7.798174438552579e-06,
+      "loss": 0.9553,
+      "step": 20225
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 13.882771029211606,
+      "learning_rate": 7.796815974610364e-06,
+      "loss": 0.9107,
+      "step": 20230
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 20.457454530687464,
+      "learning_rate": 7.79545721013793e-06,
+      "loss": 0.997,
+      "step": 20235
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.00567752149711,
+      "learning_rate": 7.794098145281278e-06,
+      "loss": 0.9093,
+      "step": 20240
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 16.015518065801285,
+      "learning_rate": 7.792738780186454e-06,
+      "loss": 0.9648,
+      "step": 20245
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 20.447413868315554,
+      "learning_rate": 7.791379114999519e-06,
+      "loss": 0.8944,
+      "step": 20250
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.709060533492773,
+      "learning_rate": 7.790019149866581e-06,
+      "loss": 0.9712,
+      "step": 20255
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 21.06159144898778,
+      "learning_rate": 7.78865888493377e-06,
+      "loss": 0.8869,
+      "step": 20260
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.418773761436006,
+      "learning_rate": 7.787298320347256e-06,
+      "loss": 0.9299,
+      "step": 20265
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 22.04331772266648,
+      "learning_rate": 7.785937456253237e-06,
+      "loss": 0.9819,
+      "step": 20270
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 20.04280913224276,
+      "learning_rate": 7.784576292797942e-06,
+      "loss": 0.9749,
+      "step": 20275
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 19.096545780104982,
+      "learning_rate": 7.783214830127635e-06,
+      "loss": 0.9623,
+      "step": 20280
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.837282050928299,
+      "learning_rate": 7.781853068388612e-06,
+      "loss": 0.948,
+      "step": 20285
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.823047837922731,
+      "learning_rate": 7.780491007727199e-06,
+      "loss": 0.9361,
+      "step": 20290
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.69324400255971,
+      "learning_rate": 7.779128648289756e-06,
+      "loss": 0.9205,
+      "step": 20295
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.876811524927135,
+      "learning_rate": 7.777765990222674e-06,
+      "loss": 0.9432,
+      "step": 20300
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.62810775402767,
+      "learning_rate": 7.776403033672377e-06,
+      "loss": 0.972,
+      "step": 20305
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.855485166436244,
+      "learning_rate": 7.775039778785323e-06,
+      "loss": 0.9392,
+      "step": 20310
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.91010169971048,
+      "learning_rate": 7.773676225707996e-06,
+      "loss": 0.9639,
+      "step": 20315
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.493402247240592,
+      "learning_rate": 7.772312374586917e-06,
+      "loss": 0.9441,
+      "step": 20320
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.117611514101663,
+      "learning_rate": 7.770948225568638e-06,
+      "loss": 0.9031,
+      "step": 20325
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.4333155763605,
+      "learning_rate": 7.769583778799743e-06,
+      "loss": 0.9402,
+      "step": 20330
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.814839172345621,
+      "learning_rate": 7.76821903442685e-06,
+      "loss": 0.9168,
+      "step": 20335
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.994237784498635,
+      "learning_rate": 7.766853992596603e-06,
+      "loss": 0.9239,
+      "step": 20340
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.5864266968455745,
+      "learning_rate": 7.765488653455684e-06,
+      "loss": 0.8941,
+      "step": 20345
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.030732770562809,
+      "learning_rate": 7.764123017150803e-06,
+      "loss": 0.9493,
+      "step": 20350
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.812501584881364,
+      "learning_rate": 7.762757083828706e-06,
+      "loss": 0.8914,
+      "step": 20355
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.5567430448249615,
+      "learning_rate": 7.761390853636166e-06,
+      "loss": 0.9132,
+      "step": 20360
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.99304509738203,
+      "learning_rate": 7.760024326719992e-06,
+      "loss": 0.9529,
+      "step": 20365
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.917609481835917,
+      "learning_rate": 7.758657503227026e-06,
+      "loss": 0.9552,
+      "step": 20370
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.693355948175669,
+      "learning_rate": 7.757290383304134e-06,
+      "loss": 0.9275,
+      "step": 20375
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.295712353407465,
+      "learning_rate": 7.755922967098222e-06,
+      "loss": 1.0089,
+      "step": 20380
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.108495593307115,
+      "learning_rate": 7.754555254756223e-06,
+      "loss": 0.9478,
+      "step": 20385
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.504904737179936,
+      "learning_rate": 7.753187246425107e-06,
+      "loss": 0.9633,
+      "step": 20390
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.300866739301536,
+      "learning_rate": 7.75181894225187e-06,
+      "loss": 0.9455,
+      "step": 20395
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.201812749460246,
+      "learning_rate": 7.750450342383543e-06,
+      "loss": 0.9413,
+      "step": 20400
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.026197017140966,
+      "learning_rate": 7.749081446967185e-06,
+      "loss": 0.9313,
+      "step": 20405
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.669060099154594,
+      "learning_rate": 7.747712256149899e-06,
+      "loss": 0.9103,
+      "step": 20410
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.294498548513014,
+      "learning_rate": 7.7463427700788e-06,
+      "loss": 0.9381,
+      "step": 20415
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.866788856883709,
+      "learning_rate": 7.744972988901052e-06,
+      "loss": 0.9512,
+      "step": 20420
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.957223837013078,
+      "learning_rate": 7.74360291276384e-06,
+      "loss": 0.9544,
+      "step": 20425
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.670840019573902,
+      "learning_rate": 7.742232541814387e-06,
+      "loss": 0.9358,
+      "step": 20430
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.67131364133245,
+      "learning_rate": 7.740861876199945e-06,
+      "loss": 0.9053,
+      "step": 20435
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.972472203850405,
+      "learning_rate": 7.7394909160678e-06,
+      "loss": 0.9314,
+      "step": 20440
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.11134998831311,
+      "learning_rate": 7.738119661565265e-06,
+      "loss": 0.8935,
+      "step": 20445
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.452260015075453,
+      "learning_rate": 7.736748112839687e-06,
+      "loss": 0.9954,
+      "step": 20450
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.667512558201727,
+      "learning_rate": 7.735376270038445e-06,
+      "loss": 0.9535,
+      "step": 20455
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.460760878877248,
+      "learning_rate": 7.734004133308954e-06,
+      "loss": 0.9174,
+      "step": 20460
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.292102266339746,
+      "learning_rate": 7.73263170279865e-06,
+      "loss": 0.9333,
+      "step": 20465
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.486230922070714,
+      "learning_rate": 7.731258978655011e-06,
+      "loss": 0.9345,
+      "step": 20470
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.199946706587383,
+      "learning_rate": 7.729885961025541e-06,
+      "loss": 0.9325,
+      "step": 20475
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.473331183840534,
+      "learning_rate": 7.728512650057775e-06,
+      "loss": 0.932,
+      "step": 20480
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.312204523691917,
+      "learning_rate": 7.727139045899283e-06,
+      "loss": 0.9185,
+      "step": 20485
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.14242500288437,
+      "learning_rate": 7.725765148697666e-06,
+      "loss": 0.9197,
+      "step": 20490
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.2522951212072,
+      "learning_rate": 7.724390958600554e-06,
+      "loss": 0.9463,
+      "step": 20495
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.519152645189666,
+      "learning_rate": 7.723016475755609e-06,
+      "loss": 0.9647,
+      "step": 20500
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.05566160577291,
+      "learning_rate": 7.721641700310526e-06,
+      "loss": 0.9518,
+      "step": 20505
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.178885415473305,
+      "learning_rate": 7.720266632413032e-06,
+      "loss": 0.9649,
+      "step": 20510
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.391420205741028,
+      "learning_rate": 7.718891272210882e-06,
+      "loss": 0.9804,
+      "step": 20515
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.572287710982629,
+      "learning_rate": 7.717515619851865e-06,
+      "loss": 0.9662,
+      "step": 20520
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.462636354099574,
+      "learning_rate": 7.716139675483802e-06,
+      "loss": 0.9602,
+      "step": 20525
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.124819557747461,
+      "learning_rate": 7.714763439254544e-06,
+      "loss": 0.9533,
+      "step": 20530
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.1256298027688985,
+      "learning_rate": 7.713386911311973e-06,
+      "loss": 0.9414,
+      "step": 20535
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.832376192386037,
+      "learning_rate": 7.712010091804003e-06,
+      "loss": 0.9107,
+      "step": 20540
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.252263820472098,
+      "learning_rate": 7.710632980878579e-06,
+      "loss": 0.9456,
+      "step": 20545
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.871561810712612,
+      "learning_rate": 7.70925557868368e-06,
+      "loss": 0.8975,
+      "step": 20550
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.994001323908641,
+      "learning_rate": 7.707877885367312e-06,
+      "loss": 0.9204,
+      "step": 20555
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.13124068629601,
+      "learning_rate": 7.706499901077514e-06,
+      "loss": 0.953,
+      "step": 20560
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 21.7476257801919,
+      "learning_rate": 7.705121625962355e-06,
+      "loss": 0.9819,
+      "step": 20565
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.1000944255777,
+      "learning_rate": 7.70374306016994e-06,
+      "loss": 0.929,
+      "step": 20570
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.350684996938273,
+      "learning_rate": 7.702364203848402e-06,
+      "loss": 0.9646,
+      "step": 20575
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.88530740213351,
+      "learning_rate": 7.700985057145901e-06,
+      "loss": 0.9111,
+      "step": 20580
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.142530855593234,
+      "learning_rate": 7.69960562021064e-06,
+      "loss": 0.9244,
+      "step": 20585
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.998818714740994,
+      "learning_rate": 7.698225893190836e-06,
+      "loss": 0.9895,
+      "step": 20590
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 36.05026297211486,
+      "learning_rate": 7.696845876234754e-06,
+      "loss": 0.9407,
+      "step": 20595
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.379932775451739,
+      "learning_rate": 7.69546556949068e-06,
+      "loss": 0.9594,
+      "step": 20600
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.11858098548503,
+      "learning_rate": 7.694084973106935e-06,
+      "loss": 0.9896,
+      "step": 20605
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.401269268204945,
+      "learning_rate": 7.69270408723187e-06,
+      "loss": 0.9432,
+      "step": 20610
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.1797372417099155,
+      "learning_rate": 7.691322912013865e-06,
+      "loss": 0.9629,
+      "step": 20615
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.705054453128703,
+      "learning_rate": 7.689941447601337e-06,
+      "loss": 0.9644,
+      "step": 20620
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.516573079744466,
+      "learning_rate": 7.688559694142727e-06,
+      "loss": 0.8937,
+      "step": 20625
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 20.21457343072996,
+      "learning_rate": 7.687177651786515e-06,
+      "loss": 0.9655,
+      "step": 20630
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.996892522397096,
+      "learning_rate": 7.685795320681204e-06,
+      "loss": 0.9521,
+      "step": 20635
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.237187821918866,
+      "learning_rate": 7.684412700975333e-06,
+      "loss": 0.9249,
+      "step": 20640
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.159260906106853,
+      "learning_rate": 7.683029792817469e-06,
+      "loss": 0.9656,
+      "step": 20645
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.635056787249885,
+      "learning_rate": 7.681646596356213e-06,
+      "loss": 0.966,
+      "step": 20650
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 17.31157956588242,
+      "learning_rate": 7.680263111740196e-06,
+      "loss": 0.9198,
+      "step": 20655
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.340323196905999,
+      "learning_rate": 7.678879339118076e-06,
+      "loss": 0.978,
+      "step": 20660
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.83179833470517,
+      "learning_rate": 7.67749527863855e-06,
+      "loss": 0.9953,
+      "step": 20665
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.0343859653169085,
+      "learning_rate": 7.676110930450341e-06,
+      "loss": 0.9748,
+      "step": 20670
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.153752771498908,
+      "learning_rate": 7.6747262947022e-06,
+      "loss": 0.9653,
+      "step": 20675
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.02771660668748,
+      "learning_rate": 7.673341371542915e-06,
+      "loss": 0.9989,
+      "step": 20680
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 48.11076183925116,
+      "learning_rate": 7.671956161121301e-06,
+      "loss": 0.9655,
+      "step": 20685
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 54.67378814486978,
+      "learning_rate": 7.670570663586206e-06,
+      "loss": 0.9601,
+      "step": 20690
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.005885281122075,
+      "learning_rate": 7.669184879086506e-06,
+      "loss": 0.9371,
+      "step": 20695
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 50.83827742882587,
+      "learning_rate": 7.667798807771113e-06,
+      "loss": 0.9901,
+      "step": 20700
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.550982513404612,
+      "learning_rate": 7.666412449788962e-06,
+      "loss": 0.9272,
+      "step": 20705
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.923425970068918,
+      "learning_rate": 7.665025805289024e-06,
+      "loss": 0.9607,
+      "step": 20710
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 19.118786487252336,
+      "learning_rate": 7.663638874420304e-06,
+      "loss": 0.9325,
+      "step": 20715
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 18.430024969069798,
+      "learning_rate": 7.66225165733183e-06,
+      "loss": 0.951,
+      "step": 20720
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.617103972216178,
+      "learning_rate": 7.66086415417267e-06,
+      "loss": 0.9237,
+      "step": 20725
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.905601849945198,
+      "learning_rate": 7.65947636509191e-06,
+      "loss": 0.9565,
+      "step": 20730
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.404541839571632,
+      "learning_rate": 7.658088290238676e-06,
+      "loss": 0.9162,
+      "step": 20735
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.419445163593089,
+      "learning_rate": 7.656699929762126e-06,
+      "loss": 0.9464,
+      "step": 20740
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.430992931986451,
+      "learning_rate": 7.655311283811445e-06,
+      "loss": 0.8827,
+      "step": 20745
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.065737646040784,
+      "learning_rate": 7.653922352535846e-06,
+      "loss": 0.9677,
+      "step": 20750
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 18.77086705350957,
+      "learning_rate": 7.652533136084576e-06,
+      "loss": 1.0184,
+      "step": 20755
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 24.260042994602436,
+      "learning_rate": 7.651143634606916e-06,
+      "loss": 0.9341,
+      "step": 20760
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.426390430477827,
+      "learning_rate": 7.649753848252172e-06,
+      "loss": 0.928,
+      "step": 20765
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 27.097812573409072,
+      "learning_rate": 7.648363777169683e-06,
+      "loss": 0.9227,
+      "step": 20770
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.242724605152857,
+      "learning_rate": 7.646973421508817e-06,
+      "loss": 0.9526,
+      "step": 20775
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.374244987877411,
+      "learning_rate": 7.645582781418976e-06,
+      "loss": 0.986,
+      "step": 20780
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.228570739391233,
+      "learning_rate": 7.644191857049589e-06,
+      "loss": 0.9002,
+      "step": 20785
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.83638250702849,
+      "learning_rate": 7.642800648550118e-06,
+      "loss": 0.9775,
+      "step": 20790
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.336442751327782,
+      "learning_rate": 7.641409156070051e-06,
+      "loss": 0.9485,
+      "step": 20795
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 19.797444900674478,
+      "learning_rate": 7.640017379758915e-06,
+      "loss": 0.9464,
+      "step": 20800
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.420222215295391,
+      "learning_rate": 7.638625319766259e-06,
+      "loss": 0.9486,
+      "step": 20805
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.96637359152699,
+      "learning_rate": 7.637232976241668e-06,
+      "loss": 0.9428,
+      "step": 20810
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.5993339538563145,
+      "learning_rate": 7.635840349334756e-06,
+      "loss": 0.9472,
+      "step": 20815
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.506197599095325,
+      "learning_rate": 7.634447439195163e-06,
+      "loss": 0.9636,
+      "step": 20820
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.271958654692396,
+      "learning_rate": 7.633054245972568e-06,
+      "loss": 0.9427,
+      "step": 20825
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.9127308558959,
+      "learning_rate": 7.631660769816674e-06,
+      "loss": 0.9478,
+      "step": 20830
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.61839773462849,
+      "learning_rate": 7.630267010877214e-06,
+      "loss": 0.9005,
+      "step": 20835
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.569140640620592,
+      "learning_rate": 7.628872969303957e-06,
+      "loss": 1.0022,
+      "step": 20840
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.386916154329136,
+      "learning_rate": 7.6274786452466956e-06,
+      "loss": 0.971,
+      "step": 20845
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.026513499182911,
+      "learning_rate": 7.626084038855259e-06,
+      "loss": 0.9737,
+      "step": 20850
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.045049475414952,
+      "learning_rate": 7.624689150279503e-06,
+      "loss": 0.9591,
+      "step": 20855
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.256779559712614,
+      "learning_rate": 7.6232939796693125e-06,
+      "loss": 0.8882,
+      "step": 20860
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.319763241650598,
+      "learning_rate": 7.621898527174608e-06,
+      "loss": 0.9075,
+      "step": 20865
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.832794111019263,
+      "learning_rate": 7.620502792945337e-06,
+      "loss": 0.9585,
+      "step": 20870
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 24.556392955013226,
+      "learning_rate": 7.619106777131473e-06,
+      "loss": 0.9639,
+      "step": 20875
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.460080309283446,
+      "learning_rate": 7.617710479883027e-06,
+      "loss": 0.9385,
+      "step": 20880
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.226225363274759,
+      "learning_rate": 7.616313901350039e-06,
+      "loss": 0.9857,
+      "step": 20885
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 21.772307778678126,
+      "learning_rate": 7.614917041682574e-06,
+      "loss": 0.9984,
+      "step": 20890
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.614314404859782,
+      "learning_rate": 7.613519901030732e-06,
+      "loss": 0.9581,
+      "step": 20895
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.963333458178771,
+      "learning_rate": 7.612122479544643e-06,
+      "loss": 0.9694,
+      "step": 20900
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.695766001888146,
+      "learning_rate": 7.610724777374465e-06,
+      "loss": 0.8983,
+      "step": 20905
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 37.120298451575536,
+      "learning_rate": 7.609326794670388e-06,
+      "loss": 0.9425,
+      "step": 20910
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 21.478949909311023,
+      "learning_rate": 7.6079285315826315e-06,
+      "loss": 0.9318,
+      "step": 20915
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.427274097955634,
+      "learning_rate": 7.606529988261444e-06,
+      "loss": 0.9863,
+      "step": 20920
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.10229688250555,
+      "learning_rate": 7.605131164857107e-06,
+      "loss": 0.9543,
+      "step": 20925
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.778828961352319,
+      "learning_rate": 7.603732061519928e-06,
+      "loss": 0.941,
+      "step": 20930
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.1941642933224035,
+      "learning_rate": 7.602332678400248e-06,
+      "loss": 0.9274,
+      "step": 20935
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.879156158554867,
+      "learning_rate": 7.6009330156484354e-06,
+      "loss": 0.94,
+      "step": 20940
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.057378736263681,
+      "learning_rate": 7.5995330734148935e-06,
+      "loss": 0.9357,
+      "step": 20945
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.615523835821663,
+      "learning_rate": 7.598132851850048e-06,
+      "loss": 0.9163,
+      "step": 20950
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.296204281967132,
+      "learning_rate": 7.5967323511043635e-06,
+      "loss": 0.9588,
+      "step": 20955
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.269780691696742,
+      "learning_rate": 7.595331571328326e-06,
+      "loss": 0.9373,
+      "step": 20960
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 19.5791004987557,
+      "learning_rate": 7.593930512672459e-06,
+      "loss": 0.9454,
+      "step": 20965
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.579961147436789,
+      "learning_rate": 7.592529175287309e-06,
+      "loss": 0.9707,
+      "step": 20970
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.361535473011104,
+      "learning_rate": 7.591127559323457e-06,
+      "loss": 0.9777,
+      "step": 20975
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.935288657655972,
+      "learning_rate": 7.589725664931515e-06,
+      "loss": 0.9889,
+      "step": 20980
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 18.982290661280405,
+      "learning_rate": 7.588323492262119e-06,
+      "loss": 0.9921,
+      "step": 20985
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.236125844110351,
+      "learning_rate": 7.586921041465942e-06,
+      "loss": 0.9336,
+      "step": 20990
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.854237049488104,
+      "learning_rate": 7.58551831269368e-06,
+      "loss": 0.9502,
+      "step": 20995
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.0312585768173905,
+      "learning_rate": 7.584115306096067e-06,
+      "loss": 0.9287,
+      "step": 21000
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.346789162775415,
+      "learning_rate": 7.5827120218238595e-06,
+      "loss": 0.8808,
+      "step": 21005
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.504551615242983,
+      "learning_rate": 7.581308460027847e-06,
+      "loss": 0.9754,
+      "step": 21010
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.901105302557985,
+      "learning_rate": 7.579904620858847e-06,
+      "loss": 0.966,
+      "step": 21015
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.525381604737314,
+      "learning_rate": 7.578500504467709e-06,
+      "loss": 0.9657,
+      "step": 21020
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.326836924303326,
+      "learning_rate": 7.577096111005314e-06,
+      "loss": 0.9968,
+      "step": 21025
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.398970644385333,
+      "learning_rate": 7.575691440622568e-06,
+      "loss": 0.9566,
+      "step": 21030
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.001077947538999,
+      "learning_rate": 7.574286493470408e-06,
+      "loss": 0.9313,
+      "step": 21035
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.791664139210824,
+      "learning_rate": 7.572881269699805e-06,
+      "loss": 0.888,
+      "step": 21040
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.22368264307788,
+      "learning_rate": 7.571475769461752e-06,
+      "loss": 0.9074,
+      "step": 21045
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.9213401332893145,
+      "learning_rate": 7.570069992907281e-06,
+      "loss": 0.8855,
+      "step": 21050
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.82363035520253,
+      "learning_rate": 7.568663940187446e-06,
+      "loss": 0.9138,
+      "step": 21055
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.680108664597871,
+      "learning_rate": 7.567257611453334e-06,
+      "loss": 0.9481,
+      "step": 21060
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.627214895767383,
+      "learning_rate": 7.565851006856062e-06,
+      "loss": 0.9298,
+      "step": 21065
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.12112975983704,
+      "learning_rate": 7.564444126546775e-06,
+      "loss": 0.9185,
+      "step": 21070
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.583574689912782,
+      "learning_rate": 7.5630369706766475e-06,
+      "loss": 0.9142,
+      "step": 21075
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.27272563823132,
+      "learning_rate": 7.561629539396888e-06,
+      "loss": 0.9386,
+      "step": 21080
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.290102595138414,
+      "learning_rate": 7.5602218328587276e-06,
+      "loss": 0.9312,
+      "step": 21085
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.313924833437394,
+      "learning_rate": 7.5588138512134336e-06,
+      "loss": 0.931,
+      "step": 21090
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 13.190736544064523,
+      "learning_rate": 7.5574055946122985e-06,
+      "loss": 0.9084,
+      "step": 21095
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.6721350480756465,
+      "learning_rate": 7.555997063206644e-06,
+      "loss": 0.9605,
+      "step": 21100
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.521803459395151,
+      "learning_rate": 7.554588257147826e-06,
+      "loss": 0.9516,
+      "step": 21105
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.17096644772755,
+      "learning_rate": 7.553179176587224e-06,
+      "loss": 0.9885,
+      "step": 21110
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.025598792959729,
+      "learning_rate": 7.5517698216762524e-06,
+      "loss": 0.9559,
+      "step": 21115
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.96985925137222,
+      "learning_rate": 7.550360192566352e-06,
+      "loss": 0.9067,
+      "step": 21120
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.77244294940966,
+      "learning_rate": 7.548950289408992e-06,
+      "loss": 0.9356,
+      "step": 21125
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.119061544556094,
+      "learning_rate": 7.547540112355676e-06,
+      "loss": 0.891,
+      "step": 21130
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.777681245297325,
+      "learning_rate": 7.546129661557931e-06,
+      "loss": 0.9037,
+      "step": 21135
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.032709466254947,
+      "learning_rate": 7.544718937167316e-06,
+      "loss": 0.8482,
+      "step": 21140
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.529650137045095,
+      "learning_rate": 7.54330793933542e-06,
+      "loss": 0.9236,
+      "step": 21145
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.360157463969724,
+      "learning_rate": 7.541896668213863e-06,
+      "loss": 0.9694,
+      "step": 21150
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.318288730102102,
+      "learning_rate": 7.540485123954291e-06,
+      "loss": 0.9218,
+      "step": 21155
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.789017705133796,
+      "learning_rate": 7.5390733067083795e-06,
+      "loss": 0.9922,
+      "step": 21160
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.418787983032518,
+      "learning_rate": 7.537661216627836e-06,
+      "loss": 0.9248,
+      "step": 21165
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.534560377496262,
+      "learning_rate": 7.536248853864394e-06,
+      "loss": 0.9383,
+      "step": 21170
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.449451161814772,
+      "learning_rate": 7.5348362185698195e-06,
+      "loss": 0.932,
+      "step": 21175
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.34518898711807,
+      "learning_rate": 7.533423310895909e-06,
+      "loss": 0.9081,
+      "step": 21180
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.198989088263259,
+      "learning_rate": 7.53201013099448e-06,
+      "loss": 0.9204,
+      "step": 21185
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.046402607382952,
+      "learning_rate": 7.530596679017388e-06,
+      "loss": 0.9475,
+      "step": 21190
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.303387107951348,
+      "learning_rate": 7.529182955116516e-06,
+      "loss": 0.952,
+      "step": 21195
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.185400883833296,
+      "learning_rate": 7.527768959443772e-06,
+      "loss": 0.9683,
+      "step": 21200
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.214167831779232,
+      "learning_rate": 7.526354692151098e-06,
+      "loss": 0.9326,
+      "step": 21205
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.274897879322006,
+      "learning_rate": 7.524940153390462e-06,
+      "loss": 0.9173,
+      "step": 21210
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 20.37349961726135,
+      "learning_rate": 7.523525343313864e-06,
+      "loss": 0.8891,
+      "step": 21215
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.138431594683539,
+      "learning_rate": 7.52211026207333e-06,
+      "loss": 0.948,
+      "step": 21220
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.763648864022242,
+      "learning_rate": 7.520694909820918e-06,
+      "loss": 0.8954,
+      "step": 21225
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.48402849122678,
+      "learning_rate": 7.519279286708713e-06,
+      "loss": 0.9023,
+      "step": 21230
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.763390265065176,
+      "learning_rate": 7.517863392888829e-06,
+      "loss": 0.909,
+      "step": 21235
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.440875544119596,
+      "learning_rate": 7.516447228513414e-06,
+      "loss": 0.932,
+      "step": 21240
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.305933083943643,
+      "learning_rate": 7.515030793734637e-06,
+      "loss": 0.9577,
+      "step": 21245
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.556069501823375,
+      "learning_rate": 7.5136140887047016e-06,
+      "loss": 0.9316,
+      "step": 21250
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.41055193641902,
+      "learning_rate": 7.512197113575839e-06,
+      "loss": 0.9318,
+      "step": 21255
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.195446472655102,
+      "learning_rate": 7.510779868500309e-06,
+      "loss": 0.9352,
+      "step": 21260
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.844228105465982,
+      "learning_rate": 7.509362353630401e-06,
+      "loss": 0.9383,
+      "step": 21265
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.837580825549168,
+      "learning_rate": 7.507944569118433e-06,
+      "loss": 0.9371,
+      "step": 21270
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.876513523293854,
+      "learning_rate": 7.506526515116753e-06,
+      "loss": 0.9466,
+      "step": 21275
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 22.946683256536083,
+      "learning_rate": 7.5051081917777364e-06,
+      "loss": 0.9951,
+      "step": 21280
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.975980402928209,
+      "learning_rate": 7.5036895992537905e-06,
+      "loss": 0.9339,
+      "step": 21285
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.188663623906173,
+      "learning_rate": 7.5022707376973455e-06,
+      "loss": 0.9763,
+      "step": 21290
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.7174814513903005,
+      "learning_rate": 7.500851607260869e-06,
+      "loss": 0.9151,
+      "step": 21295
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.202518259656641,
+      "learning_rate": 7.4994322080968486e-06,
+      "loss": 0.9213,
+      "step": 21300
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.760280416397578,
+      "learning_rate": 7.498012540357806e-06,
+      "loss": 0.9213,
+      "step": 21305
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.546815554784757,
+      "learning_rate": 7.4965926041962925e-06,
+      "loss": 0.934,
+      "step": 21310
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.195872702862795,
+      "learning_rate": 7.495172399764884e-06,
+      "loss": 0.8817,
+      "step": 21315
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.558399093100658,
+      "learning_rate": 7.49375192721619e-06,
+      "loss": 0.8942,
+      "step": 21320
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.761034264753144,
+      "learning_rate": 7.4923311867028456e-06,
+      "loss": 0.9103,
+      "step": 21325
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.009522992860807,
+      "learning_rate": 7.490910178377516e-06,
+      "loss": 0.9288,
+      "step": 21330
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 29.430300437780726,
+      "learning_rate": 7.489488902392894e-06,
+      "loss": 0.9224,
+      "step": 21335
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.850009935907973,
+      "learning_rate": 7.488067358901703e-06,
+      "loss": 0.9613,
+      "step": 21340
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 23.770443191064665,
+      "learning_rate": 7.486645548056694e-06,
+      "loss": 0.9567,
+      "step": 21345
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.809567856585966,
+      "learning_rate": 7.485223470010645e-06,
+      "loss": 0.8932,
+      "step": 21350
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.508141063152426,
+      "learning_rate": 7.4838011249163665e-06,
+      "loss": 0.9094,
+      "step": 21355
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.768026330311038,
+      "learning_rate": 7.482378512926696e-06,
+      "loss": 0.8966,
+      "step": 21360
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.05579203141165,
+      "learning_rate": 7.480955634194498e-06,
+      "loss": 0.9502,
+      "step": 21365
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.104265135011229,
+      "learning_rate": 7.479532488872669e-06,
+      "loss": 0.9326,
+      "step": 21370
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.467875400644987,
+      "learning_rate": 7.4781090771141295e-06,
+      "loss": 0.9207,
+      "step": 21375
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.78996501239046,
+      "learning_rate": 7.476685399071834e-06,
+      "loss": 0.9533,
+      "step": 21380
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.328382881534821,
+      "learning_rate": 7.475261454898761e-06,
+      "loss": 0.9038,
+      "step": 21385
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.785160044829789,
+      "learning_rate": 7.473837244747921e-06,
+      "loss": 0.9649,
+      "step": 21390
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.399980164405193,
+      "learning_rate": 7.472412768772351e-06,
+      "loss": 0.9478,
+      "step": 21395
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.415761600421625,
+      "learning_rate": 7.470988027125118e-06,
+      "loss": 0.9874,
+      "step": 21400
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.732511549440401,
+      "learning_rate": 7.469563019959315e-06,
+      "loss": 0.96,
+      "step": 21405
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.220960037971349,
+      "learning_rate": 7.468137747428068e-06,
+      "loss": 0.9026,
+      "step": 21410
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 30.03476848739393,
+      "learning_rate": 7.466712209684527e-06,
+      "loss": 0.9466,
+      "step": 21415
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.309195809786209,
+      "learning_rate": 7.465286406881872e-06,
+      "loss": 0.9114,
+      "step": 21420
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.219077562920409,
+      "learning_rate": 7.463860339173312e-06,
+      "loss": 0.9188,
+      "step": 21425
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.142719146553722,
+      "learning_rate": 7.462434006712086e-06,
+      "loss": 0.9276,
+      "step": 21430
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.711057394672944,
+      "learning_rate": 7.461007409651458e-06,
+      "loss": 0.9247,
+      "step": 21435
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.127355597514079,
+      "learning_rate": 7.459580548144722e-06,
+      "loss": 0.9162,
+      "step": 21440
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.040901079276335,
+      "learning_rate": 7.458153422345201e-06,
+      "loss": 0.9605,
+      "step": 21445
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.797276859772071,
+      "learning_rate": 7.456726032406248e-06,
+      "loss": 0.9631,
+      "step": 21450
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 21.903885998288168,
+      "learning_rate": 7.455298378481239e-06,
+      "loss": 0.9342,
+      "step": 21455
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.759976237902354,
+      "learning_rate": 7.453870460723582e-06,
+      "loss": 0.932,
+      "step": 21460
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.48102783908614,
+      "learning_rate": 7.452442279286717e-06,
+      "loss": 0.8964,
+      "step": 21465
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.488482415759527,
+      "learning_rate": 7.451013834324103e-06,
+      "loss": 0.9109,
+      "step": 21470
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.331598137248048,
+      "learning_rate": 7.449585125989237e-06,
+      "loss": 0.9643,
+      "step": 21475
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.654772828867728,
+      "learning_rate": 7.4481561544356385e-06,
+      "loss": 0.982,
+      "step": 21480
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.096943034014457,
+      "learning_rate": 7.446726919816856e-06,
+      "loss": 0.8857,
+      "step": 21485
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.639059797088008,
+      "learning_rate": 7.445297422286466e-06,
+      "loss": 0.9105,
+      "step": 21490
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.773557868297447,
+      "learning_rate": 7.443867661998079e-06,
+      "loss": 0.8886,
+      "step": 21495
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.9337824298745,
+      "learning_rate": 7.442437639105325e-06,
+      "loss": 0.9833,
+      "step": 21500
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.773334279231266,
+      "learning_rate": 7.4410073537618675e-06,
+      "loss": 0.934,
+      "step": 21505
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 26.863583338083775,
+      "learning_rate": 7.439576806121397e-06,
+      "loss": 0.9491,
+      "step": 21510
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.421138787209415,
+      "learning_rate": 7.438145996337631e-06,
+      "loss": 0.9447,
+      "step": 21515
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.72126609755657,
+      "learning_rate": 7.436714924564318e-06,
+      "loss": 0.9403,
+      "step": 21520
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.682958017245936,
+      "learning_rate": 7.4352835909552335e-06,
+      "loss": 0.8809,
+      "step": 21525
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.099390859872644,
+      "learning_rate": 7.433851995664176e-06,
+      "loss": 0.9099,
+      "step": 21530
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.537055772326045,
+      "learning_rate": 7.432420138844983e-06,
+      "loss": 0.9575,
+      "step": 21535
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.920949571431883,
+      "learning_rate": 7.4309880206515105e-06,
+      "loss": 0.9711,
+      "step": 21540
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.757306521300025,
+      "learning_rate": 7.4295556412376464e-06,
+      "loss": 0.9453,
+      "step": 21545
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.033007300876887,
+      "learning_rate": 7.4281230007573056e-06,
+      "loss": 0.9286,
+      "step": 21550
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.266553157499411,
+      "learning_rate": 7.426690099364434e-06,
+      "loss": 0.9516,
+      "step": 21555
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.135476926450496,
+      "learning_rate": 7.425256937213e-06,
+      "loss": 0.9494,
+      "step": 21560
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.423487480789941,
+      "learning_rate": 7.423823514457005e-06,
+      "loss": 0.9238,
+      "step": 21565
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.736953209377333,
+      "learning_rate": 7.4223898312504765e-06,
+      "loss": 0.9434,
+      "step": 21570
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.3092760122703,
+      "learning_rate": 7.420955887747469e-06,
+      "loss": 0.9125,
+      "step": 21575
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.908835585561126,
+      "learning_rate": 7.419521684102067e-06,
+      "loss": 0.9766,
+      "step": 21580
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.8815841406236,
+      "learning_rate": 7.418087220468382e-06,
+      "loss": 0.879,
+      "step": 21585
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.906136238731885,
+      "learning_rate": 7.4166524970005536e-06,
+      "loss": 0.9168,
+      "step": 21590
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.799089867629139,
+      "learning_rate": 7.415217513852748e-06,
+      "loss": 0.9258,
+      "step": 21595
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.953161289922542,
+      "learning_rate": 7.413782271179162e-06,
+      "loss": 0.9014,
+      "step": 21600
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.923640837621932,
+      "learning_rate": 7.412346769134019e-06,
+      "loss": 0.9164,
+      "step": 21605
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.616238957462768,
+      "learning_rate": 7.410911007871568e-06,
+      "loss": 0.9133,
+      "step": 21610
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.579105032762657,
+      "learning_rate": 7.4094749875460906e-06,
+      "loss": 0.9495,
+      "step": 21615
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.97138268909543,
+      "learning_rate": 7.408038708311891e-06,
+      "loss": 0.9299,
+      "step": 21620
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.012023377241391,
+      "learning_rate": 7.406602170323303e-06,
+      "loss": 0.9391,
+      "step": 21625
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.149749929869508,
+      "learning_rate": 7.405165373734692e-06,
+      "loss": 0.9527,
+      "step": 21630
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.553269110539492,
+      "learning_rate": 7.4037283187004475e-06,
+      "loss": 0.9511,
+      "step": 21635
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.591005957525322,
+      "learning_rate": 7.402291005374986e-06,
+      "loss": 0.9664,
+      "step": 21640
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.285002619746813,
+      "learning_rate": 7.400853433912754e-06,
+      "loss": 0.915,
+      "step": 21645
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.896674721777744,
+      "learning_rate": 7.3994156044682255e-06,
+      "loss": 0.9036,
+      "step": 21650
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.636618262462036,
+      "learning_rate": 7.3979775171959e-06,
+      "loss": 0.9748,
+      "step": 21655
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.722986156127867,
+      "learning_rate": 7.396539172250306e-06,
+      "loss": 0.9033,
+      "step": 21660
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.127526630379863,
+      "learning_rate": 7.395100569786004e-06,
+      "loss": 0.8972,
+      "step": 21665
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.817841883447162,
+      "learning_rate": 7.393661709957573e-06,
+      "loss": 0.9164,
+      "step": 21670
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.949975362196888,
+      "learning_rate": 7.392222592919628e-06,
+      "loss": 0.8962,
+      "step": 21675
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.791113841050933,
+      "learning_rate": 7.390783218826808e-06,
+      "loss": 0.9042,
+      "step": 21680
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.47892949124888,
+      "learning_rate": 7.389343587833779e-06,
+      "loss": 0.938,
+      "step": 21685
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 27.056777358833994,
+      "learning_rate": 7.387903700095238e-06,
+      "loss": 0.9364,
+      "step": 21690
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.166185386748777,
+      "learning_rate": 7.386463555765906e-06,
+      "loss": 0.8985,
+      "step": 21695
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.714153460423674,
+      "learning_rate": 7.38502315500053e-06,
+      "loss": 0.911,
+      "step": 21700
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.387898092159356,
+      "learning_rate": 7.383582497953893e-06,
+      "loss": 0.9294,
+      "step": 21705
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.956411073682496,
+      "learning_rate": 7.382141584780795e-06,
+      "loss": 0.9063,
+      "step": 21710
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.356845930989436,
+      "learning_rate": 7.380700415636071e-06,
+      "loss": 0.9089,
+      "step": 21715
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.568343275081348,
+      "learning_rate": 7.379258990674581e-06,
+      "loss": 0.9163,
+      "step": 21720
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.280649080030962,
+      "learning_rate": 7.377817310051212e-06,
+      "loss": 0.9707,
+      "step": 21725
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.924771144652284,
+      "learning_rate": 7.376375373920879e-06,
+      "loss": 0.9232,
+      "step": 21730
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.244365524235691,
+      "learning_rate": 7.374933182438524e-06,
+      "loss": 0.89,
+      "step": 21735
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.816340246901364,
+      "learning_rate": 7.373490735759117e-06,
+      "loss": 0.9262,
+      "step": 21740
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.861367749097658,
+      "learning_rate": 7.372048034037655e-06,
+      "loss": 0.9358,
+      "step": 21745
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.850857046818518,
+      "learning_rate": 7.370605077429165e-06,
+      "loss": 0.9211,
+      "step": 21750
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.384409376966101,
+      "learning_rate": 7.369161866088695e-06,
+      "loss": 0.959,
+      "step": 21755
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.583561928352578,
+      "learning_rate": 7.367718400171326e-06,
+      "loss": 0.9876,
+      "step": 21760
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.657757260203732,
+      "learning_rate": 7.366274679832167e-06,
+      "loss": 0.9265,
+      "step": 21765
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.278416652005298,
+      "learning_rate": 7.364830705226351e-06,
+      "loss": 0.9088,
+      "step": 21770
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.627095533447477,
+      "learning_rate": 7.363386476509038e-06,
+      "loss": 0.9117,
+      "step": 21775
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.149637475153247,
+      "learning_rate": 7.361941993835418e-06,
+      "loss": 0.894,
+      "step": 21780
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.040082047973201,
+      "learning_rate": 7.360497257360707e-06,
+      "loss": 0.9465,
+      "step": 21785
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 15.855224616200738,
+      "learning_rate": 7.3590522672401475e-06,
+      "loss": 0.9296,
+      "step": 21790
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.378131125789015,
+      "learning_rate": 7.357607023629011e-06,
+      "loss": 0.9305,
+      "step": 21795
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.04528733289192,
+      "learning_rate": 7.356161526682595e-06,
+      "loss": 0.9826,
+      "step": 21800
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.635493519426648,
+      "learning_rate": 7.354715776556224e-06,
+      "loss": 0.9241,
+      "step": 21805
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.598379898325025,
+      "learning_rate": 7.35326977340525e-06,
+      "loss": 0.9403,
+      "step": 21810
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.225927086509728,
+      "learning_rate": 7.351823517385054e-06,
+      "loss": 0.8884,
+      "step": 21815
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.162349727177647,
+      "learning_rate": 7.350377008651042e-06,
+      "loss": 0.9734,
+      "step": 21820
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 26.460339598551055,
+      "learning_rate": 7.348930247358646e-06,
+      "loss": 0.9369,
+      "step": 21825
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.16657239966183,
+      "learning_rate": 7.347483233663329e-06,
+      "loss": 0.9436,
+      "step": 21830
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.147835912037744,
+      "learning_rate": 7.346035967720579e-06,
+      "loss": 0.898,
+      "step": 21835
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 25.073306251006585,
+      "learning_rate": 7.34458844968591e-06,
+      "loss": 0.9495,
+      "step": 21840
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.828376469408942,
+      "learning_rate": 7.343140679714865e-06,
+      "loss": 0.8966,
+      "step": 21845
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.967412542946501,
+      "learning_rate": 7.341692657963012e-06,
+      "loss": 0.9045,
+      "step": 21850
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 15.15952437296701,
+      "learning_rate": 7.340244384585948e-06,
+      "loss": 0.9121,
+      "step": 21855
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.247090804048587,
+      "learning_rate": 7.338795859739297e-06,
+      "loss": 0.9536,
+      "step": 21860
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.935108018325918,
+      "learning_rate": 7.33734708357871e-06,
+      "loss": 0.9342,
+      "step": 21865
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.413797889212892,
+      "learning_rate": 7.335898056259862e-06,
+      "loss": 0.9729,
+      "step": 21870
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.6783359022760145,
+      "learning_rate": 7.33444877793846e-06,
+      "loss": 0.935,
+      "step": 21875
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 24.12750081090191,
+      "learning_rate": 7.332999248770233e-06,
+      "loss": 0.9092,
+      "step": 21880
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.755652262046134,
+      "learning_rate": 7.331549468910941e-06,
+      "loss": 0.9301,
+      "step": 21885
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.591485909480699,
+      "learning_rate": 7.3300994385163684e-06,
+      "loss": 0.8947,
+      "step": 21890
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.388719757046221,
+      "learning_rate": 7.328649157742328e-06,
+      "loss": 0.8905,
+      "step": 21895
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.494169298269728,
+      "learning_rate": 7.327198626744658e-06,
+      "loss": 0.8947,
+      "step": 21900
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 51.140332101414835,
+      "learning_rate": 7.325747845679225e-06,
+      "loss": 0.9236,
+      "step": 21905
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 20.862028523405275,
+      "learning_rate": 7.324296814701921e-06,
+      "loss": 0.9417,
+      "step": 21910
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.58923384462542,
+      "learning_rate": 7.322845533968668e-06,
+      "loss": 0.9265,
+      "step": 21915
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.91100161901468,
+      "learning_rate": 7.321394003635409e-06,
+      "loss": 0.9087,
+      "step": 21920
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 5.736212626380114,
+      "learning_rate": 7.319942223858121e-06,
+      "loss": 0.9458,
+      "step": 21925
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 22.091160883422937,
+      "learning_rate": 7.318490194792803e-06,
+      "loss": 0.9176,
+      "step": 21930
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.153667035667461,
+      "learning_rate": 7.317037916595479e-06,
+      "loss": 0.9271,
+      "step": 21935
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.627680599228054,
+      "learning_rate": 7.315585389422205e-06,
+      "loss": 0.9498,
+      "step": 21940
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.114959405659302,
+      "learning_rate": 7.3141326134290624e-06,
+      "loss": 0.9397,
+      "step": 21945
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.562798526521103,
+      "learning_rate": 7.312679588772158e-06,
+      "loss": 0.9259,
+      "step": 21950
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.942286974927281,
+      "learning_rate": 7.3112263156076236e-06,
+      "loss": 0.9455,
+      "step": 21955
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.377704135118599,
+      "learning_rate": 7.309772794091623e-06,
+      "loss": 0.8948,
+      "step": 21960
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.064833454227902,
+      "learning_rate": 7.30831902438034e-06,
+      "loss": 0.8722,
+      "step": 21965
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 5.788425899394792,
+      "learning_rate": 7.3068650066299915e-06,
+      "loss": 0.9434,
+      "step": 21970
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.477282345638749,
+      "learning_rate": 7.305410740996818e-06,
+      "loss": 0.8903,
+      "step": 21975
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.534544233389502,
+      "learning_rate": 7.303956227637086e-06,
+      "loss": 0.9083,
+      "step": 21980
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.62849064233974,
+      "learning_rate": 7.302501466707087e-06,
+      "loss": 0.8949,
+      "step": 21985
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 5.411750978777702,
+      "learning_rate": 7.301046458363143e-06,
+      "loss": 0.9314,
+      "step": 21990
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.274626813199493,
+      "learning_rate": 7.299591202761604e-06,
+      "loss": 0.9111,
+      "step": 21995
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.468538903210115,
+      "learning_rate": 7.298135700058842e-06,
+      "loss": 0.9599,
+      "step": 22000
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.67525963744242,
+      "learning_rate": 7.296679950411255e-06,
+      "loss": 0.9702,
+      "step": 22005
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 52.370173525848614,
+      "learning_rate": 7.295223953975272e-06,
+      "loss": 0.9188,
+      "step": 22010
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 27.721089084373194,
+      "learning_rate": 7.293767710907347e-06,
+      "loss": 0.936,
+      "step": 22015
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.564599315104779,
+      "learning_rate": 7.292311221363956e-06,
+      "loss": 0.9012,
+      "step": 22020
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.409185560916445,
+      "learning_rate": 7.2908544855016086e-06,
+      "loss": 0.9962,
+      "step": 22025
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.684631487459809,
+      "learning_rate": 7.289397503476836e-06,
+      "loss": 0.9282,
+      "step": 22030
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.845526471512144,
+      "learning_rate": 7.287940275446197e-06,
+      "loss": 0.9362,
+      "step": 22035
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.776968972980009,
+      "learning_rate": 7.286482801566279e-06,
+      "loss": 0.9185,
+      "step": 22040
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.349950934724852,
+      "learning_rate": 7.285025081993693e-06,
+      "loss": 0.9342,
+      "step": 22045
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.598907784480035,
+      "learning_rate": 7.283567116885076e-06,
+      "loss": 0.9633,
+      "step": 22050
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.723900511725269,
+      "learning_rate": 7.282108906397096e-06,
+      "loss": 0.9109,
+      "step": 22055
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.507360648128066,
+      "learning_rate": 7.2806504506864415e-06,
+      "loss": 0.9383,
+      "step": 22060
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.991026950664851,
+      "learning_rate": 7.2791917499098304e-06,
+      "loss": 0.9638,
+      "step": 22065
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.61456900114781,
+      "learning_rate": 7.277732804224008e-06,
+      "loss": 0.9348,
+      "step": 22070
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.136769181931975,
+      "learning_rate": 7.2762736137857425e-06,
+      "loss": 0.9043,
+      "step": 22075
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.535947743066793,
+      "learning_rate": 7.274814178751832e-06,
+      "loss": 0.9371,
+      "step": 22080
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.149668446261899,
+      "learning_rate": 7.273354499279097e-06,
+      "loss": 0.9109,
+      "step": 22085
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.259569954106862,
+      "learning_rate": 7.271894575524389e-06,
+      "loss": 0.9372,
+      "step": 22090
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.520044363602685,
+      "learning_rate": 7.270434407644582e-06,
+      "loss": 0.9342,
+      "step": 22095
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.69510894085174,
+      "learning_rate": 7.2689739957965786e-06,
+      "loss": 0.9227,
+      "step": 22100
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.798343423538765,
+      "learning_rate": 7.267513340137304e-06,
+      "loss": 0.9684,
+      "step": 22105
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.145059397665772,
+      "learning_rate": 7.266052440823715e-06,
+      "loss": 0.9256,
+      "step": 22110
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.415302947740033,
+      "learning_rate": 7.26459129801279e-06,
+      "loss": 0.9039,
+      "step": 22115
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.714704079424068,
+      "learning_rate": 7.263129911861535e-06,
+      "loss": 0.9035,
+      "step": 22120
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.01531219978725,
+      "learning_rate": 7.2616682825269835e-06,
+      "loss": 0.9721,
+      "step": 22125
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.33725618236222,
+      "learning_rate": 7.260206410166194e-06,
+      "loss": 0.8885,
+      "step": 22130
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.944860265352075,
+      "learning_rate": 7.2587442949362505e-06,
+      "loss": 0.9016,
+      "step": 22135
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.607350250519633,
+      "learning_rate": 7.257281936994264e-06,
+      "loss": 0.9153,
+      "step": 22140
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.831291594951576,
+      "learning_rate": 7.255819336497371e-06,
+      "loss": 0.9105,
+      "step": 22145
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.073914295879671,
+      "learning_rate": 7.254356493602734e-06,
+      "loss": 0.9161,
+      "step": 22150
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.058757566653028,
+      "learning_rate": 7.252893408467544e-06,
+      "loss": 0.9368,
+      "step": 22155
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.284533039182037,
+      "learning_rate": 7.251430081249013e-06,
+      "loss": 0.9216,
+      "step": 22160
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.434267380137317,
+      "learning_rate": 7.249966512104384e-06,
+      "loss": 0.9177,
+      "step": 22165
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.465102217859059,
+      "learning_rate": 7.2485027011909234e-06,
+      "loss": 1.0202,
+      "step": 22170
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.6467907833389726,
+      "learning_rate": 7.247038648665924e-06,
+      "loss": 0.8691,
+      "step": 22175
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.7271965196415815,
+      "learning_rate": 7.245574354686705e-06,
+      "loss": 0.9186,
+      "step": 22180
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.83492658218162,
+      "learning_rate": 7.244109819410612e-06,
+      "loss": 0.8958,
+      "step": 22185
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.424020275522027,
+      "learning_rate": 7.2426450429950134e-06,
+      "loss": 0.9232,
+      "step": 22190
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 26.204666743484392,
+      "learning_rate": 7.241180025597307e-06,
+      "loss": 0.9251,
+      "step": 22195
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.926425544551613,
+      "learning_rate": 7.2397147673749165e-06,
+      "loss": 0.947,
+      "step": 22200
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.067536442785645,
+      "learning_rate": 7.2382492684852914e-06,
+      "loss": 0.9836,
+      "step": 22205
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.695775362583674,
+      "learning_rate": 7.236783529085903e-06,
+      "loss": 0.9223,
+      "step": 22210
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.453321780226416,
+      "learning_rate": 7.235317549334253e-06,
+      "loss": 0.9266,
+      "step": 22215
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.275123053036545,
+      "learning_rate": 7.233851329387865e-06,
+      "loss": 0.9022,
+      "step": 22220
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.7271395119769375,
+      "learning_rate": 7.232384869404294e-06,
+      "loss": 0.9248,
+      "step": 22225
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.428463664840372,
+      "learning_rate": 7.230918169541117e-06,
+      "loss": 0.8752,
+      "step": 22230
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.946358985156184,
+      "learning_rate": 7.229451229955937e-06,
+      "loss": 0.8365,
+      "step": 22235
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 42.88190560959908,
+      "learning_rate": 7.227984050806385e-06,
+      "loss": 0.9866,
+      "step": 22240
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 21.488661025559217,
+      "learning_rate": 7.226516632250111e-06,
+      "loss": 0.9286,
+      "step": 22245
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.548351512433426,
+      "learning_rate": 7.225048974444799e-06,
+      "loss": 0.9127,
+      "step": 22250
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.534061759613512,
+      "learning_rate": 7.223581077548155e-06,
+      "loss": 0.8978,
+      "step": 22255
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.266938124903895,
+      "learning_rate": 7.222112941717911e-06,
+      "loss": 0.955,
+      "step": 22260
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.432951922879372,
+      "learning_rate": 7.220644567111826e-06,
+      "loss": 0.929,
+      "step": 22265
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.41367274447374,
+      "learning_rate": 7.2191759538876784e-06,
+      "loss": 0.9076,
+      "step": 22270
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.087754794968259,
+      "learning_rate": 7.217707102203283e-06,
+      "loss": 0.935,
+      "step": 22275
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.714156003685,
+      "learning_rate": 7.216238012216471e-06,
+      "loss": 0.9085,
+      "step": 22280
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 44.57032979041386,
+      "learning_rate": 7.214768684085104e-06,
+      "loss": 0.9322,
+      "step": 22285
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.515869012725076,
+      "learning_rate": 7.213299117967068e-06,
+      "loss": 0.949,
+      "step": 22290
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.425743156187252,
+      "learning_rate": 7.211829314020273e-06,
+      "loss": 0.9036,
+      "step": 22295
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 27.190723723781804,
+      "learning_rate": 7.2103592724026575e-06,
+      "loss": 0.8982,
+      "step": 22300
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.355329808517583,
+      "learning_rate": 7.208888993272182e-06,
+      "loss": 0.9073,
+      "step": 22305
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.013564120365336,
+      "learning_rate": 7.207418476786836e-06,
+      "loss": 0.9493,
+      "step": 22310
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.906256703208929,
+      "learning_rate": 7.205947723104632e-06,
+      "loss": 0.9332,
+      "step": 22315
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 24.562258183054485,
+      "learning_rate": 7.2044767323836095e-06,
+      "loss": 0.9374,
+      "step": 22320
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.557564362257907,
+      "learning_rate": 7.203005504781835e-06,
+      "loss": 0.9434,
+      "step": 22325
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.335231646004091,
+      "learning_rate": 7.201534040457394e-06,
+      "loss": 0.8839,
+      "step": 22330
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.212476331170947,
+      "learning_rate": 7.2000623395684055e-06,
+      "loss": 0.9261,
+      "step": 22335
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.275979280902316,
+      "learning_rate": 7.198590402273009e-06,
+      "loss": 0.9194,
+      "step": 22340
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.3114264949476,
+      "learning_rate": 7.1971182287293705e-06,
+      "loss": 0.894,
+      "step": 22345
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.174337369339983,
+      "learning_rate": 7.195645819095682e-06,
+      "loss": 0.9127,
+      "step": 22350
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.262321560933757,
+      "learning_rate": 7.194173173530159e-06,
+      "loss": 0.9321,
+      "step": 22355
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.37327235806034,
+      "learning_rate": 7.192700292191046e-06,
+      "loss": 0.9063,
+      "step": 22360
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.575711988987514,
+      "learning_rate": 7.191227175236609e-06,
+      "loss": 0.891,
+      "step": 22365
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.208225346556525,
+      "learning_rate": 7.189753822825142e-06,
+      "loss": 0.8968,
+      "step": 22370
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.264383245038347,
+      "learning_rate": 7.188280235114962e-06,
+      "loss": 0.9163,
+      "step": 22375
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.749216252748451,
+      "learning_rate": 7.1868064122644135e-06,
+      "loss": 0.9285,
+      "step": 22380
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.920590199400951,
+      "learning_rate": 7.185332354431864e-06,
+      "loss": 0.9474,
+      "step": 22385
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.82044116577742,
+      "learning_rate": 7.18385806177571e-06,
+      "loss": 0.9428,
+      "step": 22390
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.50893858938991,
+      "learning_rate": 7.182383534454368e-06,
+      "loss": 0.9068,
+      "step": 22395
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.595875930238892,
+      "learning_rate": 7.180908772626284e-06,
+      "loss": 0.9012,
+      "step": 22400
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.60964831714443,
+      "learning_rate": 7.179433776449927e-06,
+      "loss": 0.9271,
+      "step": 22405
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.662641977688752,
+      "learning_rate": 7.177958546083792e-06,
+      "loss": 0.9508,
+      "step": 22410
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.434300757203431,
+      "learning_rate": 7.1764830816864e-06,
+      "loss": 0.9888,
+      "step": 22415
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.975524181675334,
+      "learning_rate": 7.175007383416295e-06,
+      "loss": 1.0058,
+      "step": 22420
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.26329533226023,
+      "learning_rate": 7.173531451432047e-06,
+      "loss": 0.9366,
+      "step": 22425
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.718118615736259,
+      "learning_rate": 7.1720552858922525e-06,
+      "loss": 0.8634,
+      "step": 22430
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.28238513878075,
+      "learning_rate": 7.170578886955531e-06,
+      "loss": 0.9186,
+      "step": 22435
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.053497009922532,
+      "learning_rate": 7.16910225478053e-06,
+      "loss": 0.9129,
+      "step": 22440
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.21849780760886,
+      "learning_rate": 7.167625389525917e-06,
+      "loss": 0.9374,
+      "step": 22445
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 30.521970668345666,
+      "learning_rate": 7.16614829135039e-06,
+      "loss": 0.8727,
+      "step": 22450
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 4.898277938310586,
+      "learning_rate": 7.164670960412668e-06,
+      "loss": 0.9136,
+      "step": 22455
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 20.87439964665969,
+      "learning_rate": 7.163193396871499e-06,
+      "loss": 0.8712,
+      "step": 22460
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 35.99702937196072,
+      "learning_rate": 7.1617156008856545e-06,
+      "loss": 0.9381,
+      "step": 22465
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.976611763911491,
+      "learning_rate": 7.160237572613927e-06,
+      "loss": 0.9553,
+      "step": 22470
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 21.086741200334796,
+      "learning_rate": 7.158759312215138e-06,
+      "loss": 0.9335,
+      "step": 22475
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 49.651339677398546,
+      "learning_rate": 7.157280819848133e-06,
+      "loss": 0.9425,
+      "step": 22480
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.191181905501226,
+      "learning_rate": 7.155802095671784e-06,
+      "loss": 0.9171,
+      "step": 22485
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 55.63138015363943,
+      "learning_rate": 7.154323139844985e-06,
+      "loss": 0.9206,
+      "step": 22490
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.696177285242326,
+      "learning_rate": 7.152843952526658e-06,
+      "loss": 0.9754,
+      "step": 22495
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 24.476771225750387,
+      "learning_rate": 7.151364533875746e-06,
+      "loss": 0.8878,
+      "step": 22500
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.538184329075,
+      "learning_rate": 7.149884884051221e-06,
+      "loss": 0.9269,
+      "step": 22505
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 43.2767332850137,
+      "learning_rate": 7.1484050032120755e-06,
+      "loss": 0.9051,
+      "step": 22510
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.320841922432187,
+      "learning_rate": 7.146924891517332e-06,
+      "loss": 0.9053,
+      "step": 22515
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.649112120286738,
+      "learning_rate": 7.1454445491260325e-06,
+      "loss": 0.9733,
+      "step": 22520
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.267161596705898,
+      "learning_rate": 7.14396397619725e-06,
+      "loss": 0.9231,
+      "step": 22525
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.930588218952579,
+      "learning_rate": 7.142483172890076e-06,
+      "loss": 0.9248,
+      "step": 22530
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.6200028199088905,
+      "learning_rate": 7.141002139363627e-06,
+      "loss": 0.8843,
+      "step": 22535
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.556827139306353,
+      "learning_rate": 7.139520875777051e-06,
+      "loss": 0.9483,
+      "step": 22540
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.165119676902172,
+      "learning_rate": 7.138039382289513e-06,
+      "loss": 0.8952,
+      "step": 22545
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.303111649125197,
+      "learning_rate": 7.136557659060211e-06,
+      "loss": 0.9237,
+      "step": 22550
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.704846927534164,
+      "learning_rate": 7.135075706248355e-06,
+      "loss": 0.9604,
+      "step": 22555
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.77677861571753,
+      "learning_rate": 7.133593524013194e-06,
+      "loss": 0.9121,
+      "step": 22560
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.5412843873595525,
+      "learning_rate": 7.132111112513992e-06,
+      "loss": 0.912,
+      "step": 22565
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.670663112543473,
+      "learning_rate": 7.130628471910041e-06,
+      "loss": 0.9618,
+      "step": 22570
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.585732050084596,
+      "learning_rate": 7.129145602360657e-06,
+      "loss": 0.9062,
+      "step": 22575
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.103107598510329,
+      "learning_rate": 7.127662504025183e-06,
+      "loss": 0.9101,
+      "step": 22580
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.01943233725791,
+      "learning_rate": 7.126179177062981e-06,
+      "loss": 0.9261,
+      "step": 22585
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.997821653017817,
+      "learning_rate": 7.124695621633444e-06,
+      "loss": 0.9328,
+      "step": 22590
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.141342936339158,
+      "learning_rate": 7.123211837895985e-06,
+      "loss": 0.9232,
+      "step": 22595
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.72750955764383,
+      "learning_rate": 7.121727826010043e-06,
+      "loss": 0.9172,
+      "step": 22600
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.058597414414063,
+      "learning_rate": 7.1202435861350835e-06,
+      "loss": 0.9407,
+      "step": 22605
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.845098908601672,
+      "learning_rate": 7.118759118430594e-06,
+      "loss": 0.9316,
+      "step": 22610
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.768034101918666,
+      "learning_rate": 7.117274423056085e-06,
+      "loss": 0.9312,
+      "step": 22615
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.769306039901317,
+      "learning_rate": 7.115789500171095e-06,
+      "loss": 0.9556,
+      "step": 22620
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.0519093999596185,
+      "learning_rate": 7.1143043499351864e-06,
+      "loss": 0.8963,
+      "step": 22625
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.907905854675806,
+      "learning_rate": 7.112818972507943e-06,
+      "loss": 0.9511,
+      "step": 22630
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.598332059571938,
+      "learning_rate": 7.111333368048977e-06,
+      "loss": 0.9201,
+      "step": 22635
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.369370519283578,
+      "learning_rate": 7.109847536717922e-06,
+      "loss": 0.881,
+      "step": 22640
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.71384453100486,
+      "learning_rate": 7.10836147867444e-06,
+      "loss": 0.869,
+      "step": 22645
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.675248917546769,
+      "learning_rate": 7.106875194078209e-06,
+      "loss": 0.9442,
+      "step": 22650
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.930285510478638,
+      "learning_rate": 7.105388683088942e-06,
+      "loss": 0.9294,
+      "step": 22655
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.24951190329833,
+      "learning_rate": 7.1039019458663695e-06,
+      "loss": 0.9477,
+      "step": 22660
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.348142857270698,
+      "learning_rate": 7.1024149825702475e-06,
+      "loss": 0.8899,
+      "step": 22665
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.67447699654083,
+      "learning_rate": 7.100927793360356e-06,
+      "loss": 0.9256,
+      "step": 22670
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.115533987088446,
+      "learning_rate": 7.099440378396502e-06,
+      "loss": 0.9057,
+      "step": 22675
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.366326363607728,
+      "learning_rate": 7.097952737838511e-06,
+      "loss": 0.9024,
+      "step": 22680
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.404456823283693,
+      "learning_rate": 7.0964648718462426e-06,
+      "loss": 0.9132,
+      "step": 22685
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 6.792306925925805,
+      "learning_rate": 7.09497678057957e-06,
+      "loss": 0.8965,
+      "step": 22690
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.340332273780737,
+      "learning_rate": 7.093488464198397e-06,
+      "loss": 0.9729,
+      "step": 22695
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.601090418333475,
+      "learning_rate": 7.091999922862649e-06,
+      "loss": 0.9433,
+      "step": 22700
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.122373273003937,
+      "learning_rate": 7.090511156732277e-06,
+      "loss": 0.9546,
+      "step": 22705
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 6.473023619409803,
+      "learning_rate": 7.089022165967254e-06,
+      "loss": 0.906,
+      "step": 22710
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.409357024355955,
+      "learning_rate": 7.087532950727582e-06,
+      "loss": 0.8626,
+      "step": 22715
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 25.322002531201004,
+      "learning_rate": 7.08604351117328e-06,
+      "loss": 0.9287,
+      "step": 22720
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 23.513069694962137,
+      "learning_rate": 7.084553847464395e-06,
+      "loss": 0.885,
+      "step": 22725
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 24.602574775502212,
+      "learning_rate": 7.083063959761001e-06,
+      "loss": 0.9411,
+      "step": 22730
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 38.43463450946847,
+      "learning_rate": 7.08157384822319e-06,
+      "loss": 0.9622,
+      "step": 22735
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.837521503763455,
+      "learning_rate": 7.0800835130110834e-06,
+      "loss": 0.9552,
+      "step": 22740
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.27927981728115,
+      "learning_rate": 7.0785929542848215e-06,
+      "loss": 0.9136,
+      "step": 22745
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.121110550348236,
+      "learning_rate": 7.0771021722045765e-06,
+      "loss": 0.8846,
+      "step": 22750
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 37.53688475301775,
+      "learning_rate": 7.075611166930534e-06,
+      "loss": 0.9124,
+      "step": 22755
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.175021177370883,
+      "learning_rate": 7.074119938622912e-06,
+      "loss": 0.9364,
+      "step": 22760
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 20.262302868378036,
+      "learning_rate": 7.072628487441948e-06,
+      "loss": 0.9637,
+      "step": 22765
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.775707461774358,
+      "learning_rate": 7.071136813547905e-06,
+      "loss": 0.9108,
+      "step": 22770
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.753441719972129,
+      "learning_rate": 7.069644917101071e-06,
+      "loss": 0.8669,
+      "step": 22775
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.98038321878649,
+      "learning_rate": 7.068152798261758e-06,
+      "loss": 0.943,
+      "step": 22780
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.688091548978663,
+      "learning_rate": 7.0666604571903e-06,
+      "loss": 0.9335,
+      "step": 22785
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.089918928021166,
+      "learning_rate": 7.065167894047054e-06,
+      "loss": 0.9484,
+      "step": 22790
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.118865353599768,
+      "learning_rate": 7.063675108992403e-06,
+      "loss": 0.9033,
+      "step": 22795
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 16.963601352405306,
+      "learning_rate": 7.062182102186753e-06,
+      "loss": 0.9264,
+      "step": 22800
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.1220859970529915,
+      "learning_rate": 7.0606888737905365e-06,
+      "loss": 0.8986,
+      "step": 22805
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.487584795061593,
+      "learning_rate": 7.0591954239642065e-06,
+      "loss": 0.9296,
+      "step": 22810
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.99782311670588,
+      "learning_rate": 7.057701752868238e-06,
+      "loss": 0.8505,
+      "step": 22815
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.579620852800494,
+      "learning_rate": 7.056207860663136e-06,
+      "loss": 0.8783,
+      "step": 22820
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.296017321736796,
+      "learning_rate": 7.0547137475094255e-06,
+      "loss": 0.9371,
+      "step": 22825
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.960629404843116,
+      "learning_rate": 7.053219413567653e-06,
+      "loss": 0.9411,
+      "step": 22830
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 18.384737208117784,
+      "learning_rate": 7.051724858998395e-06,
+      "loss": 0.9514,
+      "step": 22835
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.022686875452175,
+      "learning_rate": 7.050230083962245e-06,
+      "loss": 0.879,
+      "step": 22840
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.3519032214357996,
+      "learning_rate": 7.048735088619825e-06,
+      "loss": 0.8871,
+      "step": 22845
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 21.28589690201938,
+      "learning_rate": 7.047239873131777e-06,
+      "loss": 0.8918,
+      "step": 22850
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.823643019813286,
+      "learning_rate": 7.045744437658769e-06,
+      "loss": 0.9036,
+      "step": 22855
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.923494877479909,
+      "learning_rate": 7.044248782361494e-06,
+      "loss": 0.8931,
+      "step": 22860
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.833656736894615,
+      "learning_rate": 7.042752907400663e-06,
+      "loss": 0.8937,
+      "step": 22865
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.326339595268569,
+      "learning_rate": 7.041256812937019e-06,
+      "loss": 0.9264,
+      "step": 22870
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.170266178527214,
+      "learning_rate": 7.0397604991313205e-06,
+      "loss": 0.8944,
+      "step": 22875
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.544711546545316,
+      "learning_rate": 7.038263966144354e-06,
+      "loss": 0.8782,
+      "step": 22880
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.8314676607732245,
+      "learning_rate": 7.0367672141369305e-06,
+      "loss": 0.9347,
+      "step": 22885
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.92429482550247,
+      "learning_rate": 7.035270243269879e-06,
+      "loss": 0.9339,
+      "step": 22890
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.94715772583652,
+      "learning_rate": 7.03377305370406e-06,
+      "loss": 0.9287,
+      "step": 22895
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.271521713882415,
+      "learning_rate": 7.032275645600348e-06,
+      "loss": 0.8781,
+      "step": 22900
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.115409480836378,
+      "learning_rate": 7.030778019119651e-06,
+      "loss": 0.8808,
+      "step": 22905
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.607515014167464,
+      "learning_rate": 7.02928017442289e-06,
+      "loss": 0.9299,
+      "step": 22910
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.112423378133695,
+      "learning_rate": 7.027782111671021e-06,
+      "loss": 0.9115,
+      "step": 22915
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.803563609203383,
+      "learning_rate": 7.026283831025014e-06,
+      "loss": 0.931,
+      "step": 22920
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.6766259783063875,
+      "learning_rate": 7.024785332645868e-06,
+      "loss": 0.9264,
+      "step": 22925
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.894090944961063,
+      "learning_rate": 7.023286616694601e-06,
+      "loss": 0.8574,
+      "step": 22930
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.4690923806329605,
+      "learning_rate": 7.021787683332258e-06,
+      "loss": 0.8889,
+      "step": 22935
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.252728078940986,
+      "learning_rate": 7.020288532719904e-06,
+      "loss": 0.8472,
+      "step": 22940
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 26.9965026541785,
+      "learning_rate": 7.018789165018632e-06,
+      "loss": 0.9592,
+      "step": 22945
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.968393944045484,
+      "learning_rate": 7.017289580389556e-06,
+      "loss": 0.9045,
+      "step": 22950
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 26.179560119848876,
+      "learning_rate": 7.015789778993809e-06,
+      "loss": 0.9276,
+      "step": 22955
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.937481508419991,
+      "learning_rate": 7.014289760992557e-06,
+      "loss": 0.9017,
+      "step": 22960
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.451652369480728,
+      "learning_rate": 7.012789526546977e-06,
+      "loss": 0.9887,
+      "step": 22965
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.803117245864987,
+      "learning_rate": 7.011289075818282e-06,
+      "loss": 0.9478,
+      "step": 22970
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.2883053599456105,
+      "learning_rate": 7.009788408967697e-06,
+      "loss": 0.9335,
+      "step": 22975
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 23.497441395526586,
+      "learning_rate": 7.00828752615648e-06,
+      "loss": 0.8702,
+      "step": 22980
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.855532617397643,
+      "learning_rate": 7.006786427545904e-06,
+      "loss": 0.9082,
+      "step": 22985
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 5.048931952110115,
+      "learning_rate": 7.00528511329727e-06,
+      "loss": 0.9059,
+      "step": 22990
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.66372780157301,
+      "learning_rate": 7.003783583571899e-06,
+      "loss": 0.9761,
+      "step": 22995
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.552353020347688,
+      "learning_rate": 7.00228183853114e-06,
+      "loss": 0.9035,
+      "step": 23000
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 24.534899451912995,
+      "learning_rate": 7.00077987833636e-06,
+      "loss": 0.9455,
+      "step": 23005
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.467100650492112,
+      "learning_rate": 6.9992777031489535e-06,
+      "loss": 0.9418,
+      "step": 23010
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.698976620601943,
+      "learning_rate": 6.997775313130333e-06,
+      "loss": 0.8872,
+      "step": 23015
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.832633983636459,
+      "learning_rate": 6.996272708441938e-06,
+      "loss": 0.9282,
+      "step": 23020
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.352439488233575,
+      "learning_rate": 6.994769889245231e-06,
+      "loss": 0.8995,
+      "step": 23025
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.48147182731211,
+      "learning_rate": 6.993266855701695e-06,
+      "loss": 0.892,
+      "step": 23030
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 18.849754403821247,
+      "learning_rate": 6.9917636079728406e-06,
+      "loss": 0.9528,
+      "step": 23035
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.651359787938244,
+      "learning_rate": 6.990260146220194e-06,
+      "loss": 0.8351,
+      "step": 23040
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.370217410537741,
+      "learning_rate": 6.9887564706053115e-06,
+      "loss": 0.9426,
+      "step": 23045
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.70394133087629,
+      "learning_rate": 6.987252581289768e-06,
+      "loss": 0.9055,
+      "step": 23050
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 27.32885678822093,
+      "learning_rate": 6.9857484784351655e-06,
+      "loss": 0.8945,
+      "step": 23055
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.621572619727342,
+      "learning_rate": 6.984244162203125e-06,
+      "loss": 0.8696,
+      "step": 23060
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.562334679762717,
+      "learning_rate": 6.982739632755293e-06,
+      "loss": 0.938,
+      "step": 23065
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.061149977987767,
+      "learning_rate": 6.981234890253337e-06,
+      "loss": 0.9163,
+      "step": 23070
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.873704539780729,
+      "learning_rate": 6.979729934858948e-06,
+      "loss": 0.8843,
+      "step": 23075
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 5.7028884439878444,
+      "learning_rate": 6.978224766733839e-06,
+      "loss": 0.8695,
+      "step": 23080
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 5.88299343636864,
+      "learning_rate": 6.976719386039749e-06,
+      "loss": 0.9143,
+      "step": 23085
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.677385200318884,
+      "learning_rate": 6.975213792938437e-06,
+      "loss": 0.9459,
+      "step": 23090
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.283570981343427,
+      "learning_rate": 6.973707987591685e-06,
+      "loss": 0.8779,
+      "step": 23095
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 5.7902067354862,
+      "learning_rate": 6.9722019701613e-06,
+      "loss": 0.9108,
+      "step": 23100
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.374454764423398,
+      "learning_rate": 6.970695740809109e-06,
+      "loss": 0.918,
+      "step": 23105
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 40.44961824942354,
+      "learning_rate": 6.969189299696964e-06,
+      "loss": 0.939,
+      "step": 23110
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 25.486534692469263,
+      "learning_rate": 6.967682646986737e-06,
+      "loss": 0.8774,
+      "step": 23115
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.439612198161807,
+      "learning_rate": 6.966175782840326e-06,
+      "loss": 0.8995,
+      "step": 23120
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.447008413712133,
+      "learning_rate": 6.96466870741965e-06,
+      "loss": 0.9588,
+      "step": 23125
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.018192043437732,
+      "learning_rate": 6.96316142088665e-06,
+      "loss": 0.8905,
+      "step": 23130
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.739310783834362,
+      "learning_rate": 6.961653923403292e-06,
+      "loss": 0.8888,
+      "step": 23135
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.560477693119511,
+      "learning_rate": 6.960146215131561e-06,
+      "loss": 0.9398,
+      "step": 23140
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.404104210076437,
+      "learning_rate": 6.95863829623347e-06,
+      "loss": 0.9036,
+      "step": 23145
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.404842343876951,
+      "learning_rate": 6.95713016687105e-06,
+      "loss": 0.9098,
+      "step": 23150
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.287855837477224,
+      "learning_rate": 6.9556218272063555e-06,
+      "loss": 0.8953,
+      "step": 23155
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.665724258424726,
+      "learning_rate": 6.9541132774014655e-06,
+      "loss": 0.939,
+      "step": 23160
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.411905802724334,
+      "learning_rate": 6.952604517618478e-06,
+      "loss": 0.8758,
+      "step": 23165
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.087428296554743,
+      "learning_rate": 6.95109554801952e-06,
+      "loss": 0.8978,
+      "step": 23170
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.981740311835065,
+      "learning_rate": 6.949586368766732e-06,
+      "loss": 0.9031,
+      "step": 23175
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.693184545791876,
+      "learning_rate": 6.948076980022286e-06,
+      "loss": 0.8938,
+      "step": 23180
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.721662877574692,
+      "learning_rate": 6.94656738194837e-06,
+      "loss": 0.882,
+      "step": 23185
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 19.52912853443083,
+      "learning_rate": 6.945057574707199e-06,
+      "loss": 0.9026,
+      "step": 23190
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 41.80736345299809,
+      "learning_rate": 6.943547558461007e-06,
+      "loss": 0.8934,
+      "step": 23195
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.204201458994314,
+      "learning_rate": 6.942037333372051e-06,
+      "loss": 0.9153,
+      "step": 23200
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.818834373576317,
+      "learning_rate": 6.940526899602616e-06,
+      "loss": 0.9235,
+      "step": 23205
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.72989382073166,
+      "learning_rate": 6.939016257314999e-06,
+      "loss": 0.8852,
+      "step": 23210
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.944495771991438,
+      "learning_rate": 6.937505406671529e-06,
+      "loss": 0.932,
+      "step": 23215
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.85102067979384,
+      "learning_rate": 6.93599434783455e-06,
+      "loss": 0.9483,
+      "step": 23220
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.551513722779058,
+      "learning_rate": 6.934483080966436e-06,
+      "loss": 0.8869,
+      "step": 23225
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.127174041000742,
+      "learning_rate": 6.932971606229577e-06,
+      "loss": 0.9309,
+      "step": 23230
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.26645524750772,
+      "learning_rate": 6.931459923786388e-06,
+      "loss": 0.9358,
+      "step": 23235
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.405720641063699,
+      "learning_rate": 6.929948033799306e-06,
+      "loss": 0.9149,
+      "step": 23240
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.41260620788048,
+      "learning_rate": 6.928435936430789e-06,
+      "loss": 0.9381,
+      "step": 23245
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.027273959987731,
+      "learning_rate": 6.926923631843321e-06,
+      "loss": 0.9639,
+      "step": 23250
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.816868218385879,
+      "learning_rate": 6.925411120199405e-06,
+      "loss": 0.8901,
+      "step": 23255
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.012856734903804,
+      "learning_rate": 6.923898401661565e-06,
+      "loss": 0.8995,
+      "step": 23260
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.684332794187576,
+      "learning_rate": 6.9223854763923535e-06,
+      "loss": 0.895,
+      "step": 23265
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.935038240368268,
+      "learning_rate": 6.920872344554337e-06,
+      "loss": 0.8969,
+      "step": 23270
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.931894701975166,
+      "learning_rate": 6.919359006310109e-06,
+      "loss": 0.9714,
+      "step": 23275
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.630957657303016,
+      "learning_rate": 6.917845461822286e-06,
+      "loss": 0.9567,
+      "step": 23280
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.424482875536688,
+      "learning_rate": 6.916331711253502e-06,
+      "loss": 0.9343,
+      "step": 23285
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.557035393427002,
+      "learning_rate": 6.914817754766421e-06,
+      "loss": 0.915,
+      "step": 23290
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.670466170417317,
+      "learning_rate": 6.913303592523721e-06,
+      "loss": 0.8685,
+      "step": 23295
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.83666483915853,
+      "learning_rate": 6.9117892246881045e-06,
+      "loss": 0.8855,
+      "step": 23300
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.782540122639701,
+      "learning_rate": 6.9102746514222995e-06,
+      "loss": 0.9422,
+      "step": 23305
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.961461722029677,
+      "learning_rate": 6.9087598728890525e-06,
+      "loss": 0.9122,
+      "step": 23310
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.241198898764172,
+      "learning_rate": 6.907244889251132e-06,
+      "loss": 0.9071,
+      "step": 23315
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.891236683405234,
+      "learning_rate": 6.905729700671332e-06,
+      "loss": 0.8945,
+      "step": 23320
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.243215176915836,
+      "learning_rate": 6.904214307312465e-06,
+      "loss": 0.9009,
+      "step": 23325
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.069419232486801,
+      "learning_rate": 6.902698709337367e-06,
+      "loss": 0.93,
+      "step": 23330
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.738426661549367,
+      "learning_rate": 6.901182906908894e-06,
+      "loss": 0.896,
+      "step": 23335
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.792474467053207,
+      "learning_rate": 6.8996669001899285e-06,
+      "loss": 0.8695,
+      "step": 23340
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.660481984924338,
+      "learning_rate": 6.898150689343371e-06,
+      "loss": 0.9335,
+      "step": 23345
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.224931352082289,
+      "learning_rate": 6.896634274532143e-06,
+      "loss": 0.9077,
+      "step": 23350
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.88243518031466,
+      "learning_rate": 6.895117655919194e-06,
+      "loss": 0.9098,
+      "step": 23355
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.746063824332506,
+      "learning_rate": 6.893600833667488e-06,
+      "loss": 0.8758,
+      "step": 23360
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.598247819320011,
+      "learning_rate": 6.892083807940014e-06,
+      "loss": 0.9587,
+      "step": 23365
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.9956490645739855,
+      "learning_rate": 6.890566578899785e-06,
+      "loss": 0.9347,
+      "step": 23370
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.814640034413959,
+      "learning_rate": 6.8890491467098346e-06,
+      "loss": 0.8782,
+      "step": 23375
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.075841787547497,
+      "learning_rate": 6.887531511533217e-06,
+      "loss": 0.9052,
+      "step": 23380
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.795366425007409,
+      "learning_rate": 6.886013673533005e-06,
+      "loss": 0.8573,
+      "step": 23385
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.231542374109674,
+      "learning_rate": 6.8844956328723035e-06,
+      "loss": 0.9334,
+      "step": 23390
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 23.192404416044255,
+      "learning_rate": 6.8829773897142275e-06,
+      "loss": 0.907,
+      "step": 23395
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.393423054625753,
+      "learning_rate": 6.88145894422192e-06,
+      "loss": 0.9124,
+      "step": 23400
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.43558683335424,
+      "learning_rate": 6.879940296558547e-06,
+      "loss": 0.8925,
+      "step": 23405
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.044044488230632,
+      "learning_rate": 6.878421446887293e-06,
+      "loss": 0.9474,
+      "step": 23410
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.177135488545063,
+      "learning_rate": 6.8769023953713635e-06,
+      "loss": 0.9493,
+      "step": 23415
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.354104539621598,
+      "learning_rate": 6.875383142173987e-06,
+      "loss": 0.9003,
+      "step": 23420
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 18.2051141430026,
+      "learning_rate": 6.873863687458416e-06,
+      "loss": 0.8885,
+      "step": 23425
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.121902676775598,
+      "learning_rate": 6.872344031387922e-06,
+      "loss": 0.9174,
+      "step": 23430
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.5423910468005415,
+      "learning_rate": 6.8708241741258e-06,
+      "loss": 0.865,
+      "step": 23435
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.002572637031899,
+      "learning_rate": 6.869304115835363e-06,
+      "loss": 0.907,
+      "step": 23440
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 25.44987537229217,
+      "learning_rate": 6.867783856679947e-06,
+      "loss": 0.9445,
+      "step": 23445
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 16.73885257888077,
+      "learning_rate": 6.866263396822914e-06,
+      "loss": 0.9328,
+      "step": 23450
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.086310910129086,
+      "learning_rate": 6.864742736427643e-06,
+      "loss": 0.8993,
+      "step": 23455
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.437841497147433,
+      "learning_rate": 6.863221875657534e-06,
+      "loss": 0.9174,
+      "step": 23460
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.703303964336197,
+      "learning_rate": 6.861700814676012e-06,
+      "loss": 0.8807,
+      "step": 23465
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.506898203069278,
+      "learning_rate": 6.860179553646521e-06,
+      "loss": 0.9135,
+      "step": 23470
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 28.302665131505723,
+      "learning_rate": 6.858658092732528e-06,
+      "loss": 0.892,
+      "step": 23475
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.875087356710619,
+      "learning_rate": 6.85713643209752e-06,
+      "loss": 0.9282,
+      "step": 23480
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.691872392237835,
+      "learning_rate": 6.855614571905005e-06,
+      "loss": 0.9049,
+      "step": 23485
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.514972680015523,
+      "learning_rate": 6.854092512318515e-06,
+      "loss": 0.8664,
+      "step": 23490
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.619277007034693,
+      "learning_rate": 6.852570253501603e-06,
+      "loss": 0.9035,
+      "step": 23495
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.62043661242379,
+      "learning_rate": 6.851047795617841e-06,
+      "loss": 0.8775,
+      "step": 23500
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.446867241351837,
+      "learning_rate": 6.849525138830824e-06,
+      "loss": 0.865,
+      "step": 23505
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.438148316920962,
+      "learning_rate": 6.848002283304167e-06,
+      "loss": 0.9548,
+      "step": 23510
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.359072627366345,
+      "learning_rate": 6.84647922920151e-06,
+      "loss": 0.9274,
+      "step": 23515
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.989361218067277,
+      "learning_rate": 6.844955976686512e-06,
+      "loss": 0.9021,
+      "step": 23520
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.230364599716836,
+      "learning_rate": 6.84343252592285e-06,
+      "loss": 0.8948,
+      "step": 23525
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.98758940404333,
+      "learning_rate": 6.841908877074227e-06,
+      "loss": 0.9016,
+      "step": 23530
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 27.255757499231205,
+      "learning_rate": 6.840385030304368e-06,
+      "loss": 0.8917,
+      "step": 23535
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.143384001163476,
+      "learning_rate": 6.838860985777014e-06,
+      "loss": 0.9043,
+      "step": 23540
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.476389444237618,
+      "learning_rate": 6.837336743655931e-06,
+      "loss": 0.9477,
+      "step": 23545
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.800654232457841,
+      "learning_rate": 6.835812304104907e-06,
+      "loss": 0.8902,
+      "step": 23550
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.544692458361583,
+      "learning_rate": 6.834287667287749e-06,
+      "loss": 0.9201,
+      "step": 23555
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.319286052191164,
+      "learning_rate": 6.832762833368285e-06,
+      "loss": 0.928,
+      "step": 23560
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.0828580472573535,
+      "learning_rate": 6.831237802510367e-06,
+      "loss": 0.9598,
+      "step": 23565
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.409833697299415,
+      "learning_rate": 6.829712574877865e-06,
+      "loss": 0.9373,
+      "step": 23570
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.706241323475199,
+      "learning_rate": 6.828187150634671e-06,
+      "loss": 0.9079,
+      "step": 23575
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.403233187821114,
+      "learning_rate": 6.8266615299447e-06,
+      "loss": 0.8927,
+      "step": 23580
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.851444257744309,
+      "learning_rate": 6.8251357129718865e-06,
+      "loss": 0.8721,
+      "step": 23585
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.0422070223834,
+      "learning_rate": 6.823609699880185e-06,
+      "loss": 0.9054,
+      "step": 23590
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.247800976842763,
+      "learning_rate": 6.822083490833572e-06,
+      "loss": 0.8728,
+      "step": 23595
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 55.45705994144435,
+      "learning_rate": 6.820557085996047e-06,
+      "loss": 0.9218,
+      "step": 23600
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 17.18753143594485,
+      "learning_rate": 6.819030485531629e-06,
+      "loss": 0.9119,
+      "step": 23605
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.866445006459287,
+      "learning_rate": 6.817503689604358e-06,
+      "loss": 0.882,
+      "step": 23610
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 25.334152410579883,
+      "learning_rate": 6.815976698378293e-06,
+      "loss": 0.8768,
+      "step": 23615
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.220031067608439,
+      "learning_rate": 6.814449512017517e-06,
+      "loss": 0.8812,
+      "step": 23620
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.619646587220572,
+      "learning_rate": 6.812922130686134e-06,
+      "loss": 0.9145,
+      "step": 23625
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.900802967511703,
+      "learning_rate": 6.8113945545482675e-06,
+      "loss": 0.8857,
+      "step": 23630
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.076090197548098,
+      "learning_rate": 6.809866783768061e-06,
+      "loss": 0.9291,
+      "step": 23635
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.410453301923219,
+      "learning_rate": 6.808338818509681e-06,
+      "loss": 0.9378,
+      "step": 23640
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.998593463902603,
+      "learning_rate": 6.806810658937314e-06,
+      "loss": 0.9021,
+      "step": 23645
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.890474846801583,
+      "learning_rate": 6.805282305215167e-06,
+      "loss": 0.9017,
+      "step": 23650
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.055329875681373,
+      "learning_rate": 6.803753757507471e-06,
+      "loss": 0.8884,
+      "step": 23655
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.75440217956922,
+      "learning_rate": 6.802225015978472e-06,
+      "loss": 0.8596,
+      "step": 23660
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.427630934113202,
+      "learning_rate": 6.800696080792442e-06,
+      "loss": 0.8856,
+      "step": 23665
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.886267425253317,
+      "learning_rate": 6.79916695211367e-06,
+      "loss": 0.9021,
+      "step": 23670
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.015285032522612,
+      "learning_rate": 6.797637630106468e-06,
+      "loss": 0.937,
+      "step": 23675
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.529029572300798,
+      "learning_rate": 6.796108114935171e-06,
+      "loss": 0.9225,
+      "step": 23680
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.32900836607524,
+      "learning_rate": 6.7945784067641295e-06,
+      "loss": 0.8797,
+      "step": 23685
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.687494127214395,
+      "learning_rate": 6.793048505757718e-06,
+      "loss": 0.9021,
+      "step": 23690
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.758744075524426,
+      "learning_rate": 6.791518412080331e-06,
+      "loss": 0.9482,
+      "step": 23695
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.61880923103712,
+      "learning_rate": 6.789988125896384e-06,
+      "loss": 0.9058,
+      "step": 23700
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 14.022768199244858,
+      "learning_rate": 6.788457647370313e-06,
+      "loss": 0.8848,
+      "step": 23705
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.060705404701003,
+      "learning_rate": 6.786926976666575e-06,
+      "loss": 0.8892,
+      "step": 23710
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.479713213199258,
+      "learning_rate": 6.785396113949647e-06,
+      "loss": 0.9402,
+      "step": 23715
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.928263873223764,
+      "learning_rate": 6.783865059384029e-06,
+      "loss": 0.8923,
+      "step": 23720
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.566092359449536,
+      "learning_rate": 6.782333813134234e-06,
+      "loss": 0.9144,
+      "step": 23725
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.25578266659643,
+      "learning_rate": 6.780802375364806e-06,
+      "loss": 0.9016,
+      "step": 23730
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.726431794649871,
+      "learning_rate": 6.779270746240303e-06,
+      "loss": 0.9566,
+      "step": 23735
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.915722491732196,
+      "learning_rate": 6.777738925925306e-06,
+      "loss": 0.8913,
+      "step": 23740
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.003518216733056,
+      "learning_rate": 6.776206914584418e-06,
+      "loss": 0.8772,
+      "step": 23745
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.632009317832925,
+      "learning_rate": 6.774674712382256e-06,
+      "loss": 0.9134,
+      "step": 23750
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.188689870427255,
+      "learning_rate": 6.773142319483465e-06,
+      "loss": 0.9276,
+      "step": 23755
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.402133030778085,
+      "learning_rate": 6.771609736052704e-06,
+      "loss": 0.9102,
+      "step": 23760
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.9249351411386435,
+      "learning_rate": 6.770076962254661e-06,
+      "loss": 0.9158,
+      "step": 23765
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.750023153531884,
+      "learning_rate": 6.768543998254035e-06,
+      "loss": 0.8807,
+      "step": 23770
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 4.909869868765495,
+      "learning_rate": 6.767010844215551e-06,
+      "loss": 0.884,
+      "step": 23775
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.326120300976482,
+      "learning_rate": 6.765477500303954e-06,
+      "loss": 0.8567,
+      "step": 23780
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.578031638349115,
+      "learning_rate": 6.763943966684008e-06,
+      "loss": 0.8706,
+      "step": 23785
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.261813353698697,
+      "learning_rate": 6.7624102435204965e-06,
+      "loss": 0.8838,
+      "step": 23790
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 29.741810587824805,
+      "learning_rate": 6.760876330978228e-06,
+      "loss": 0.8639,
+      "step": 23795
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 16.270226875512073,
+      "learning_rate": 6.7593422292220255e-06,
+      "loss": 0.9144,
+      "step": 23800
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.448881661424943,
+      "learning_rate": 6.7578079384167365e-06,
+      "loss": 0.896,
+      "step": 23805
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.202110394989795,
+      "learning_rate": 6.756273458727226e-06,
+      "loss": 0.916,
+      "step": 23810
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.903833267320922,
+      "learning_rate": 6.754738790318383e-06,
+      "loss": 0.8663,
+      "step": 23815
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.391739423170614,
+      "learning_rate": 6.75320393335511e-06,
+      "loss": 0.8905,
+      "step": 23820
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.077050977707632,
+      "learning_rate": 6.7516688880023385e-06,
+      "loss": 0.9332,
+      "step": 23825
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 23.743628173305318,
+      "learning_rate": 6.750133654425013e-06,
+      "loss": 0.9161,
+      "step": 23830
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 16.745657377829414,
+      "learning_rate": 6.748598232788104e-06,
+      "loss": 0.8957,
+      "step": 23835
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.446130902764844,
+      "learning_rate": 6.747062623256597e-06,
+      "loss": 0.8879,
+      "step": 23840
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 23.66458250447321,
+      "learning_rate": 6.745526825995501e-06,
+      "loss": 0.9368,
+      "step": 23845
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.361233466436759,
+      "learning_rate": 6.743990841169844e-06,
+      "loss": 0.8773,
+      "step": 23850
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.162267733768624,
+      "learning_rate": 6.7424546689446735e-06,
+      "loss": 0.9123,
+      "step": 23855
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.6656794383814075,
+      "learning_rate": 6.7409183094850606e-06,
+      "loss": 0.8381,
+      "step": 23860
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.420367992270999,
+      "learning_rate": 6.739381762956091e-06,
+      "loss": 0.8862,
+      "step": 23865
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.007625253825814,
+      "learning_rate": 6.737845029522873e-06,
+      "loss": 0.9083,
+      "step": 23870
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.152991556675145,
+      "learning_rate": 6.7363081093505385e-06,
+      "loss": 0.948,
+      "step": 23875
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.763319672444526,
+      "learning_rate": 6.734771002604234e-06,
+      "loss": 0.9215,
+      "step": 23880
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.940980897857443,
+      "learning_rate": 6.73323370944913e-06,
+      "loss": 0.8981,
+      "step": 23885
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.008914292982135,
+      "learning_rate": 6.731696230050415e-06,
+      "loss": 0.9224,
+      "step": 23890
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 23.381809517188863,
+      "learning_rate": 6.730158564573297e-06,
+      "loss": 0.9244,
+      "step": 23895
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.774575802257527,
+      "learning_rate": 6.728620713183005e-06,
+      "loss": 0.9301,
+      "step": 23900
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.993235258233463,
+      "learning_rate": 6.727082676044789e-06,
+      "loss": 0.8784,
+      "step": 23905
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 19.364974308253316,
+      "learning_rate": 6.725544453323917e-06,
+      "loss": 0.8971,
+      "step": 23910
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.348752077592053,
+      "learning_rate": 6.724006045185678e-06,
+      "loss": 0.8947,
+      "step": 23915
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.945222014315064,
+      "learning_rate": 6.72246745179538e-06,
+      "loss": 0.9303,
+      "step": 23920
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.256884730233514,
+      "learning_rate": 6.720928673318354e-06,
+      "loss": 0.908,
+      "step": 23925
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.2846890661713175,
+      "learning_rate": 6.719389709919946e-06,
+      "loss": 0.8654,
+      "step": 23930
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.431304333951894,
+      "learning_rate": 6.717850561765526e-06,
+      "loss": 0.8463,
+      "step": 23935
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.647786409774287,
+      "learning_rate": 6.716311229020482e-06,
+      "loss": 0.9063,
+      "step": 23940
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.275085440626508,
+      "learning_rate": 6.7147717118502195e-06,
+      "loss": 0.8848,
+      "step": 23945
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.451789215212424,
+      "learning_rate": 6.713232010420172e-06,
+      "loss": 0.9511,
+      "step": 23950
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.311045463061686,
+      "learning_rate": 6.7116921248957825e-06,
+      "loss": 0.9176,
+      "step": 23955
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.385579122070995,
+      "learning_rate": 6.71015205544252e-06,
+      "loss": 0.9255,
+      "step": 23960
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.113695843819294,
+      "learning_rate": 6.70861180222587e-06,
+      "loss": 0.9171,
+      "step": 23965
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.7170783559669385,
+      "learning_rate": 6.707071365411342e-06,
+      "loss": 0.861,
+      "step": 23970
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.540916617313485,
+      "learning_rate": 6.705530745164462e-06,
+      "loss": 0.9204,
+      "step": 23975
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.6984347449119905,
+      "learning_rate": 6.703989941650778e-06,
+      "loss": 0.923,
+      "step": 23980
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.954047441243636,
+      "learning_rate": 6.702448955035854e-06,
+      "loss": 0.9062,
+      "step": 23985
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.854859766330289,
+      "learning_rate": 6.700907785485276e-06,
+      "loss": 0.9246,
+      "step": 23990
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.4202023328370075,
+      "learning_rate": 6.699366433164648e-06,
+      "loss": 0.8483,
+      "step": 23995
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.144621933045308,
+      "learning_rate": 6.6978248982395985e-06,
+      "loss": 0.9012,
+      "step": 24000
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.063579652154793,
+      "learning_rate": 6.69628318087577e-06,
+      "loss": 0.8871,
+      "step": 24005
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.171947794665156,
+      "learning_rate": 6.694741281238828e-06,
+      "loss": 0.8746,
+      "step": 24010
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.753600198946728,
+      "learning_rate": 6.693199199494455e-06,
+      "loss": 0.8552,
+      "step": 24015
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.824389911709043,
+      "learning_rate": 6.691656935808356e-06,
+      "loss": 0.9,
+      "step": 24020
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.070175980848593,
+      "learning_rate": 6.690114490346253e-06,
+      "loss": 0.9461,
+      "step": 24025
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.3289723176474615,
+      "learning_rate": 6.6885718632738905e-06,
+      "loss": 0.8524,
+      "step": 24030
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.3145039574700155,
+      "learning_rate": 6.687029054757028e-06,
+      "loss": 0.9205,
+      "step": 24035
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.822699493157175,
+      "learning_rate": 6.6854860649614485e-06,
+      "loss": 0.9234,
+      "step": 24040
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.996966757584824,
+      "learning_rate": 6.683942894052953e-06,
+      "loss": 0.9018,
+      "step": 24045
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.993873202858208,
+      "learning_rate": 6.6823995421973595e-06,
+      "loss": 0.875,
+      "step": 24050
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.882760470552306,
+      "learning_rate": 6.680856009560512e-06,
+      "loss": 0.9592,
+      "step": 24055
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.374507719422299,
+      "learning_rate": 6.679312296308267e-06,
+      "loss": 0.8681,
+      "step": 24060
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.972104971967129,
+      "learning_rate": 6.677768402606506e-06,
+      "loss": 0.8594,
+      "step": 24065
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.961894260257484,
+      "learning_rate": 6.676224328621125e-06,
+      "loss": 0.9086,
+      "step": 24070
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.207913770094201,
+      "learning_rate": 6.674680074518042e-06,
+      "loss": 0.9557,
+      "step": 24075
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.48564843157255,
+      "learning_rate": 6.673135640463193e-06,
+      "loss": 0.9001,
+      "step": 24080
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.833650036343723,
+      "learning_rate": 6.671591026622537e-06,
+      "loss": 0.9063,
+      "step": 24085
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.149536944530058,
+      "learning_rate": 6.670046233162047e-06,
+      "loss": 0.8651,
+      "step": 24090
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 34.02702911959087,
+      "learning_rate": 6.6685012602477195e-06,
+      "loss": 0.8629,
+      "step": 24095
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.067272681334696,
+      "learning_rate": 6.666956108045567e-06,
+      "loss": 0.9668,
+      "step": 24100
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.837599270410714,
+      "learning_rate": 6.665410776721624e-06,
+      "loss": 0.8588,
+      "step": 24105
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.400501710036545,
+      "learning_rate": 6.6638652664419435e-06,
+      "loss": 0.8645,
+      "step": 24110
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.535531093312578,
+      "learning_rate": 6.662319577372595e-06,
+      "loss": 0.8897,
+      "step": 24115
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.668187540853762,
+      "learning_rate": 6.660773709679675e-06,
+      "loss": 0.9188,
+      "step": 24120
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.22530715349334,
+      "learning_rate": 6.659227663529289e-06,
+      "loss": 0.9295,
+      "step": 24125
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.4353611915676865,
+      "learning_rate": 6.657681439087568e-06,
+      "loss": 0.9431,
+      "step": 24130
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.163732371466336,
+      "learning_rate": 6.656135036520659e-06,
+      "loss": 0.9214,
+      "step": 24135
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.221908914196966,
+      "learning_rate": 6.6545884559947326e-06,
+      "loss": 0.9004,
+      "step": 24140
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 16.785413527142993,
+      "learning_rate": 6.6530416976759726e-06,
+      "loss": 0.903,
+      "step": 24145
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.236098941288745,
+      "learning_rate": 6.651494761730588e-06,
+      "loss": 0.8998,
+      "step": 24150
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.807280501531137,
+      "learning_rate": 6.649947648324804e-06,
+      "loss": 0.8643,
+      "step": 24155
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.15111923004061,
+      "learning_rate": 6.648400357624861e-06,
+      "loss": 0.9288,
+      "step": 24160
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.499821742039364,
+      "learning_rate": 6.646852889797026e-06,
+      "loss": 0.8844,
+      "step": 24165
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.0951989170665595,
+      "learning_rate": 6.645305245007579e-06,
+      "loss": 0.877,
+      "step": 24170
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.205097681339211,
+      "learning_rate": 6.643757423422823e-06,
+      "loss": 0.9341,
+      "step": 24175
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.533925349318908,
+      "learning_rate": 6.6422094252090785e-06,
+      "loss": 0.8331,
+      "step": 24180
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.514997555381436,
+      "learning_rate": 6.640661250532682e-06,
+      "loss": 0.8939,
+      "step": 24185
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.497389223312976,
+      "learning_rate": 6.639112899559993e-06,
+      "loss": 0.9044,
+      "step": 24190
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.689805575524457,
+      "learning_rate": 6.637564372457389e-06,
+      "loss": 0.8848,
+      "step": 24195
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.161155418796742,
+      "learning_rate": 6.636015669391267e-06,
+      "loss": 0.8491,
+      "step": 24200
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.322713922986848,
+      "learning_rate": 6.634466790528041e-06,
+      "loss": 0.9094,
+      "step": 24205
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.73281249455527,
+      "learning_rate": 6.632917736034144e-06,
+      "loss": 0.8755,
+      "step": 24210
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.709596528897627,
+      "learning_rate": 6.63136850607603e-06,
+      "loss": 0.8604,
+      "step": 24215
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.4034785590760555,
+      "learning_rate": 6.629819100820171e-06,
+      "loss": 0.8708,
+      "step": 24220
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.1557035973551235,
+      "learning_rate": 6.628269520433056e-06,
+      "loss": 0.9246,
+      "step": 24225
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.652696342299771,
+      "learning_rate": 6.626719765081196e-06,
+      "loss": 0.8771,
+      "step": 24230
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.037612278057251,
+      "learning_rate": 6.6251698349311155e-06,
+      "loss": 0.9395,
+      "step": 24235
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.725450029748061,
+      "learning_rate": 6.623619730149366e-06,
+      "loss": 0.9113,
+      "step": 24240
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.9854179888099,
+      "learning_rate": 6.622069450902511e-06,
+      "loss": 0.976,
+      "step": 24245
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.286664327455405,
+      "learning_rate": 6.6205189973571326e-06,
+      "loss": 0.8414,
+      "step": 24250
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.455908245073016,
+      "learning_rate": 6.6189683696798375e-06,
+      "loss": 0.8579,
+      "step": 24255
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 26.273909501813236,
+      "learning_rate": 6.617417568037246e-06,
+      "loss": 0.9301,
+      "step": 24260
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.677031404711934,
+      "learning_rate": 6.615866592596e-06,
+      "loss": 0.8775,
+      "step": 24265
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.576371235769397,
+      "learning_rate": 6.614315443522756e-06,
+      "loss": 0.8928,
+      "step": 24270
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.643525558880776,
+      "learning_rate": 6.612764120984192e-06,
+      "loss": 0.8931,
+      "step": 24275
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.599883259262377,
+      "learning_rate": 6.6112126251470075e-06,
+      "loss": 0.8728,
+      "step": 24280
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.8797725768296605,
+      "learning_rate": 6.609660956177915e-06,
+      "loss": 0.877,
+      "step": 24285
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.531567409636851,
+      "learning_rate": 6.6081091142436475e-06,
+      "loss": 0.8984,
+      "step": 24290
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.849773411092357,
+      "learning_rate": 6.60655709951096e-06,
+      "loss": 0.8277,
+      "step": 24295
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.6100089507358035,
+      "learning_rate": 6.60500491214662e-06,
+      "loss": 0.8839,
+      "step": 24300
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.848983897832184,
+      "learning_rate": 6.60345255231742e-06,
+      "loss": 0.9342,
+      "step": 24305
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.013188110826276,
+      "learning_rate": 6.601900020190167e-06,
+      "loss": 0.8687,
+      "step": 24310
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.455077724905622,
+      "learning_rate": 6.600347315931687e-06,
+      "loss": 0.8886,
+      "step": 24315
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.825880326441571,
+      "learning_rate": 6.598794439708825e-06,
+      "loss": 0.9145,
+      "step": 24320
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.456776446291894,
+      "learning_rate": 6.597241391688445e-06,
+      "loss": 0.9086,
+      "step": 24325
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.479615448910597,
+      "learning_rate": 6.595688172037426e-06,
+      "loss": 0.8931,
+      "step": 24330
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.65624270643566,
+      "learning_rate": 6.594134780922672e-06,
+      "loss": 0.8784,
+      "step": 24335
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.385421035507843,
+      "learning_rate": 6.5925812185111e-06,
+      "loss": 0.854,
+      "step": 24340
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.20468556438624,
+      "learning_rate": 6.591027484969649e-06,
+      "loss": 0.9047,
+      "step": 24345
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.081358205395726,
+      "learning_rate": 6.589473580465273e-06,
+      "loss": 0.8766,
+      "step": 24350
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.60755075072945,
+      "learning_rate": 6.587919505164944e-06,
+      "loss": 0.8444,
+      "step": 24355
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.975445326040262,
+      "learning_rate": 6.586365259235658e-06,
+      "loss": 0.9368,
+      "step": 24360
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.440889850561119,
+      "learning_rate": 6.584810842844423e-06,
+      "loss": 0.8926,
+      "step": 24365
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.97226779666037,
+      "learning_rate": 6.583256256158268e-06,
+      "loss": 0.9052,
+      "step": 24370
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.614871942084818,
+      "learning_rate": 6.581701499344241e-06,
+      "loss": 0.891,
+      "step": 24375
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.98225230714363,
+      "learning_rate": 6.5801465725694084e-06,
+      "loss": 0.8997,
+      "step": 24380
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.141527249054406,
+      "learning_rate": 6.57859147600085e-06,
+      "loss": 0.9336,
+      "step": 24385
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 26.978978481650447,
+      "learning_rate": 6.577036209805674e-06,
+      "loss": 0.9623,
+      "step": 24390
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.0085285157054145,
+      "learning_rate": 6.575480774150995e-06,
+      "loss": 0.883,
+      "step": 24395
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.973297400687372,
+      "learning_rate": 6.573925169203955e-06,
+      "loss": 0.8907,
+      "step": 24400
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.12874098597241,
+      "learning_rate": 6.57236939513171e-06,
+      "loss": 0.8681,
+      "step": 24405
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 20.489405514091995,
+      "learning_rate": 6.570813452101433e-06,
+      "loss": 0.9485,
+      "step": 24410
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.825511689833132,
+      "learning_rate": 6.5692573402803185e-06,
+      "loss": 0.9311,
+      "step": 24415
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.207204710047804,
+      "learning_rate": 6.567701059835576e-06,
+      "loss": 0.9101,
+      "step": 24420
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.565623632795624,
+      "learning_rate": 6.566144610934435e-06,
+      "loss": 0.8604,
+      "step": 24425
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.191331791124046,
+      "learning_rate": 6.564587993744146e-06,
+      "loss": 0.9112,
+      "step": 24430
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.719323186471554,
+      "learning_rate": 6.563031208431972e-06,
+      "loss": 0.8985,
+      "step": 24435
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.472543929164538,
+      "learning_rate": 6.561474255165195e-06,
+      "loss": 0.8866,
+      "step": 24440
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.278997033725792,
+      "learning_rate": 6.559917134111119e-06,
+      "loss": 0.8752,
+      "step": 24445
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.931142163657672,
+      "learning_rate": 6.558359845437062e-06,
+      "loss": 0.8647,
+      "step": 24450
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.655121288301975,
+      "learning_rate": 6.556802389310363e-06,
+      "loss": 0.8521,
+      "step": 24455
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.483630269448815,
+      "learning_rate": 6.555244765898375e-06,
+      "loss": 0.9128,
+      "step": 24460
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.528404552133885,
+      "learning_rate": 6.553686975368474e-06,
+      "loss": 0.9096,
+      "step": 24465
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.20347045163404,
+      "learning_rate": 6.5521290178880506e-06,
+      "loss": 0.8642,
+      "step": 24470
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.485085670272495,
+      "learning_rate": 6.550570893624514e-06,
+      "loss": 0.9377,
+      "step": 24475
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.887696281800459,
+      "learning_rate": 6.549012602745291e-06,
+      "loss": 0.8747,
+      "step": 24480
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.994828999328062,
+      "learning_rate": 6.547454145417829e-06,
+      "loss": 0.9385,
+      "step": 24485
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.334231853516249,
+      "learning_rate": 6.545895521809589e-06,
+      "loss": 0.8794,
+      "step": 24490
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.268624783916795,
+      "learning_rate": 6.5443367320880525e-06,
+      "loss": 0.8619,
+      "step": 24495
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.783650959083377,
+      "learning_rate": 6.542777776420718e-06,
+      "loss": 0.8789,
+      "step": 24500
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.9292372237481,
+      "learning_rate": 6.541218654975104e-06,
+      "loss": 0.8944,
+      "step": 24505
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.230987610753485,
+      "learning_rate": 6.539659367918742e-06,
+      "loss": 0.8336,
+      "step": 24510
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.402426235708,
+      "learning_rate": 6.538099915419186e-06,
+      "loss": 0.9086,
+      "step": 24515
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.521747168511222,
+      "learning_rate": 6.536540297644006e-06,
+      "loss": 0.907,
+      "step": 24520
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.288832044319982,
+      "learning_rate": 6.53498051476079e-06,
+      "loss": 0.8607,
+      "step": 24525
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.257944223613798,
+      "learning_rate": 6.533420566937144e-06,
+      "loss": 0.8738,
+      "step": 24530
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.697763669074525,
+      "learning_rate": 6.53186045434069e-06,
+      "loss": 0.8742,
+      "step": 24535
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.057755241535447,
+      "learning_rate": 6.530300177139069e-06,
+      "loss": 0.9117,
+      "step": 24540
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.195785411490776,
+      "learning_rate": 6.528739735499941e-06,
+      "loss": 0.9315,
+      "step": 24545
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.077626964672405,
+      "learning_rate": 6.5271791295909825e-06,
+      "loss": 0.8673,
+      "step": 24550
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.225630531251211,
+      "learning_rate": 6.525618359579886e-06,
+      "loss": 0.9096,
+      "step": 24555
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.130510366921712,
+      "learning_rate": 6.5240574256343616e-06,
+      "loss": 0.8678,
+      "step": 24560
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.621493811126335,
+      "learning_rate": 6.5224963279221425e-06,
+      "loss": 0.9051,
+      "step": 24565
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.722220389242948,
+      "learning_rate": 6.520935066610974e-06,
+      "loss": 0.8903,
+      "step": 24570
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.155998115487919,
+      "learning_rate": 6.519373641868618e-06,
+      "loss": 0.8764,
+      "step": 24575
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.048569421418783,
+      "learning_rate": 6.517812053862862e-06,
+      "loss": 0.9247,
+      "step": 24580
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.756624372537695,
+      "learning_rate": 6.5162503027615e-06,
+      "loss": 0.9014,
+      "step": 24585
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.737979349085914,
+      "learning_rate": 6.514688388732352e-06,
+      "loss": 0.8604,
+      "step": 24590
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.919735464378718,
+      "learning_rate": 6.513126311943252e-06,
+      "loss": 0.9148,
+      "step": 24595
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.890362198732355,
+      "learning_rate": 6.511564072562051e-06,
+      "loss": 0.8921,
+      "step": 24600
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.487482611124358,
+      "learning_rate": 6.51000167075662e-06,
+      "loss": 0.8393,
+      "step": 24605
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.096843095711902,
+      "learning_rate": 6.508439106694845e-06,
+      "loss": 0.8853,
+      "step": 24610
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.893865245494559,
+      "learning_rate": 6.50687638054463e-06,
+      "loss": 0.9308,
+      "step": 24615
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.374757157768002,
+      "learning_rate": 6.505313492473899e-06,
+      "loss": 0.8687,
+      "step": 24620
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.920382017173019,
+      "learning_rate": 6.503750442650589e-06,
+      "loss": 0.8968,
+      "step": 24625
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.83421800476811,
+      "learning_rate": 6.502187231242657e-06,
+      "loss": 0.8701,
+      "step": 24630
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.689279051469565,
+      "learning_rate": 6.500623858418079e-06,
+      "loss": 0.9134,
+      "step": 24635
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.225474214139165,
+      "learning_rate": 6.499060324344844e-06,
+      "loss": 0.8801,
+      "step": 24640
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.285216527163399,
+      "learning_rate": 6.497496629190961e-06,
+      "loss": 0.9135,
+      "step": 24645
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.869406471728855,
+      "learning_rate": 6.4959327731244555e-06,
+      "loss": 0.8699,
+      "step": 24650
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.684139317173707,
+      "learning_rate": 6.494368756313371e-06,
+      "loss": 0.8906,
+      "step": 24655
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.71579311627586,
+      "learning_rate": 6.492804578925769e-06,
+      "loss": 0.893,
+      "step": 24660
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.936805310285867,
+      "learning_rate": 6.4912402411297286e-06,
+      "loss": 0.8631,
+      "step": 24665
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.05884857433253,
+      "learning_rate": 6.489675743093341e-06,
+      "loss": 0.9146,
+      "step": 24670
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.3067112291437715,
+      "learning_rate": 6.4881110849847204e-06,
+      "loss": 0.8826,
+      "step": 24675
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.880415774068949,
+      "learning_rate": 6.486546266971998e-06,
+      "loss": 0.7971,
+      "step": 24680
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.686128962403247,
+      "learning_rate": 6.484981289223317e-06,
+      "loss": 0.842,
+      "step": 24685
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.92598098721158,
+      "learning_rate": 6.483416151906844e-06,
+      "loss": 0.883,
+      "step": 24690
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.630521797861853,
+      "learning_rate": 6.481850855190757e-06,
+      "loss": 0.9244,
+      "step": 24695
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 27.8139426452066,
+      "learning_rate": 6.480285399243256e-06,
+      "loss": 0.8625,
+      "step": 24700
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.430948094362927,
+      "learning_rate": 6.478719784232556e-06,
+      "loss": 0.8972,
+      "step": 24705
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.999048720723598,
+      "learning_rate": 6.477154010326889e-06,
+      "loss": 0.9009,
+      "step": 24710
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.965381645274808,
+      "learning_rate": 6.475588077694505e-06,
+      "loss": 0.8493,
+      "step": 24715
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.18368555625259,
+      "learning_rate": 6.474021986503671e-06,
+      "loss": 0.8346,
+      "step": 24720
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.818410445079008,
+      "learning_rate": 6.472455736922667e-06,
+      "loss": 0.8637,
+      "step": 24725
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.831499519101065,
+      "learning_rate": 6.4708893291197975e-06,
+      "loss": 0.9107,
+      "step": 24730
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.697058660581076,
+      "learning_rate": 6.469322763263376e-06,
+      "loss": 0.9177,
+      "step": 24735
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.019008254570577,
+      "learning_rate": 6.46775603952174e-06,
+      "loss": 0.9133,
+      "step": 24740
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.298237758721369,
+      "learning_rate": 6.466189158063239e-06,
+      "loss": 0.9255,
+      "step": 24745
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.8914488766991715,
+      "learning_rate": 6.464622119056244e-06,
+      "loss": 0.9161,
+      "step": 24750
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.43374631683808,
+      "learning_rate": 6.463054922669137e-06,
+      "loss": 0.9035,
+      "step": 24755
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.132517055749973,
+      "learning_rate": 6.4614875690703215e-06,
+      "loss": 0.8667,
+      "step": 24760
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.49444207450362,
+      "learning_rate": 6.459920058428217e-06,
+      "loss": 0.8921,
+      "step": 24765
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.045145716078492,
+      "learning_rate": 6.45835239091126e-06,
+      "loss": 0.8992,
+      "step": 24770
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.961172024303564,
+      "learning_rate": 6.456784566687902e-06,
+      "loss": 0.8938,
+      "step": 24775
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.233519988942753,
+      "learning_rate": 6.455216585926613e-06,
+      "loss": 0.8617,
+      "step": 24780
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.23949624083405,
+      "learning_rate": 6.45364844879588e-06,
+      "loss": 0.8431,
+      "step": 24785
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.396077134941041,
+      "learning_rate": 6.452080155464203e-06,
+      "loss": 0.8919,
+      "step": 24790
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.396749198929252,
+      "learning_rate": 6.450511706100107e-06,
+      "loss": 0.8904,
+      "step": 24795
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.877715855336982,
+      "learning_rate": 6.448943100872126e-06,
+      "loss": 0.8361,
+      "step": 24800
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.822409634067037,
+      "learning_rate": 6.4473743399488155e-06,
+      "loss": 0.9148,
+      "step": 24805
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.7102776655145,
+      "learning_rate": 6.445805423498743e-06,
+      "loss": 0.9055,
+      "step": 24810
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.858371842544252,
+      "learning_rate": 6.444236351690497e-06,
+      "loss": 0.914,
+      "step": 24815
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.962813924582823,
+      "learning_rate": 6.442667124692682e-06,
+      "loss": 0.8621,
+      "step": 24820
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.891505022894111,
+      "learning_rate": 6.441097742673916e-06,
+      "loss": 0.8471,
+      "step": 24825
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.47898487299598,
+      "learning_rate": 6.439528205802839e-06,
+      "loss": 0.8807,
+      "step": 24830
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.065318273903309,
+      "learning_rate": 6.437958514248103e-06,
+      "loss": 0.8413,
+      "step": 24835
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.29291027692286,
+      "learning_rate": 6.436388668178377e-06,
+      "loss": 0.9162,
+      "step": 24840
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.91946410906538,
+      "learning_rate": 6.434818667762351e-06,
+      "loss": 0.8399,
+      "step": 24845
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.824563437322766,
+      "learning_rate": 6.433248513168726e-06,
+      "loss": 0.9025,
+      "step": 24850
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.914602587605255,
+      "learning_rate": 6.431678204566223e-06,
+      "loss": 0.8224,
+      "step": 24855
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 19.121135709261793,
+      "learning_rate": 6.430107742123578e-06,
+      "loss": 0.8606,
+      "step": 24860
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.500416970879522,
+      "learning_rate": 6.428537126009547e-06,
+      "loss": 0.8898,
+      "step": 24865
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.45941980315844,
+      "learning_rate": 6.4269663563928945e-06,
+      "loss": 0.8737,
+      "step": 24870
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.411222889141725,
+      "learning_rate": 6.42539543344241e-06,
+      "loss": 0.8824,
+      "step": 24875
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.010538313701555,
+      "learning_rate": 6.423824357326894e-06,
+      "loss": 0.912,
+      "step": 24880
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.049221208392482,
+      "learning_rate": 6.422253128215168e-06,
+      "loss": 0.874,
+      "step": 24885
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.948141376148575,
+      "learning_rate": 6.420681746276067e-06,
+      "loss": 0.8519,
+      "step": 24890
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.051185069525754,
+      "learning_rate": 6.419110211678443e-06,
+      "loss": 0.8837,
+      "step": 24895
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.6009875393177175,
+      "learning_rate": 6.4175385245911625e-06,
+      "loss": 0.8951,
+      "step": 24900
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.436140089614032,
+      "learning_rate": 6.415966685183113e-06,
+      "loss": 0.9,
+      "step": 24905
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.148240214992114,
+      "learning_rate": 6.414394693623193e-06,
+      "loss": 0.8465,
+      "step": 24910
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.605172598122554,
+      "learning_rate": 6.4128225500803196e-06,
+      "loss": 0.8897,
+      "step": 24915
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.289992920383536,
+      "learning_rate": 6.4112502547234315e-06,
+      "loss": 0.8595,
+      "step": 24920
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.744894114120008,
+      "learning_rate": 6.409677807721472e-06,
+      "loss": 0.9081,
+      "step": 24925
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.459954436924413,
+      "learning_rate": 6.408105209243411e-06,
+      "loss": 0.898,
+      "step": 24930
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.015891739297883,
+      "learning_rate": 6.406532459458233e-06,
+      "loss": 0.8679,
+      "step": 24935
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.921865963329948,
+      "learning_rate": 6.404959558534932e-06,
+      "loss": 0.8989,
+      "step": 24940
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.06684515241933,
+      "learning_rate": 6.403386506642527e-06,
+      "loss": 0.8739,
+      "step": 24945
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.934994574831989,
+      "learning_rate": 6.40181330395005e-06,
+      "loss": 0.8825,
+      "step": 24950
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.9766846337708746,
+      "learning_rate": 6.400239950626545e-06,
+      "loss": 0.8815,
+      "step": 24955
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.282948521741755,
+      "learning_rate": 6.398666446841077e-06,
+      "loss": 0.8781,
+      "step": 24960
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.211274590780548,
+      "learning_rate": 6.397092792762728e-06,
+      "loss": 0.8597,
+      "step": 24965
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.644971253752303,
+      "learning_rate": 6.395518988560592e-06,
+      "loss": 0.8741,
+      "step": 24970
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.434603215972421,
+      "learning_rate": 6.393945034403781e-06,
+      "loss": 0.874,
+      "step": 24975
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.6981290913609834,
+      "learning_rate": 6.392370930461425e-06,
+      "loss": 0.8454,
+      "step": 24980
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.097345226606922,
+      "learning_rate": 6.390796676902668e-06,
+      "loss": 0.8954,
+      "step": 24985
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.765114001394074,
+      "learning_rate": 6.3892222738966695e-06,
+      "loss": 0.8748,
+      "step": 24990
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.81137370437781,
+      "learning_rate": 6.387647721612605e-06,
+      "loss": 0.8775,
+      "step": 24995
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.5015110343649205,
+      "learning_rate": 6.38607302021967e-06,
+      "loss": 0.8673,
+      "step": 25000
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.9944624280897525,
+      "learning_rate": 6.384498169887073e-06,
+      "loss": 0.8924,
+      "step": 25005
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.728359410450134,
+      "learning_rate": 6.382923170784037e-06,
+      "loss": 0.8606,
+      "step": 25010
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.327083871082298,
+      "learning_rate": 6.381348023079801e-06,
+      "loss": 0.8697,
+      "step": 25015
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.699955445201512,
+      "learning_rate": 6.379772726943624e-06,
+      "loss": 0.904,
+      "step": 25020
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 36.29050066769943,
+      "learning_rate": 6.378197282544779e-06,
+      "loss": 0.8883,
+      "step": 25025
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.3506843223871,
+      "learning_rate": 6.376621690052553e-06,
+      "loss": 0.9365,
+      "step": 25030
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 13.72574921360742,
+      "learning_rate": 6.375045949636253e-06,
+      "loss": 0.8563,
+      "step": 25035
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.781050704121032,
+      "learning_rate": 6.3734700614651945e-06,
+      "loss": 0.9129,
+      "step": 25040
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.732855766505102,
+      "learning_rate": 6.371894025708718e-06,
+      "loss": 0.8889,
+      "step": 25045
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.536775001681233,
+      "learning_rate": 6.370317842536173e-06,
+      "loss": 0.9323,
+      "step": 25050
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.937566040867825,
+      "learning_rate": 6.368741512116927e-06,
+      "loss": 0.915,
+      "step": 25055
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.695035820102787,
+      "learning_rate": 6.367165034620366e-06,
+      "loss": 0.8801,
+      "step": 25060
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.506692219149493,
+      "learning_rate": 6.365588410215888e-06,
+      "loss": 0.8918,
+      "step": 25065
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.933956555174458,
+      "learning_rate": 6.364011639072909e-06,
+      "loss": 0.8454,
+      "step": 25070
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.953535565751432,
+      "learning_rate": 6.362434721360859e-06,
+      "loss": 0.9239,
+      "step": 25075
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 37.87355257476908,
+      "learning_rate": 6.3608576572491865e-06,
+      "loss": 0.9462,
+      "step": 25080
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 20.470777727115546,
+      "learning_rate": 6.3592804469073525e-06,
+      "loss": 0.9016,
+      "step": 25085
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.541677099796816,
+      "learning_rate": 6.357703090504835e-06,
+      "loss": 1.0141,
+      "step": 25090
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.172382287645066,
+      "learning_rate": 6.356125588211128e-06,
+      "loss": 0.8746,
+      "step": 25095
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.655241483180802,
+      "learning_rate": 6.354547940195743e-06,
+      "loss": 0.9102,
+      "step": 25100
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 21.650089145484905,
+      "learning_rate": 6.352970146628203e-06,
+      "loss": 0.8752,
+      "step": 25105
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 38.9052328367733,
+      "learning_rate": 6.351392207678049e-06,
+      "loss": 0.9005,
+      "step": 25110
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 25.1361538119838,
+      "learning_rate": 6.349814123514837e-06,
+      "loss": 0.9217,
+      "step": 25115
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.728783466528613,
+      "learning_rate": 6.3482358943081426e-06,
+      "loss": 0.9287,
+      "step": 25120
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 25.17885962726306,
+      "learning_rate": 6.34665752022755e-06,
+      "loss": 0.8961,
+      "step": 25125
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 28.414541659143694,
+      "learning_rate": 6.3450790014426645e-06,
+      "loss": 0.9085,
+      "step": 25130
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.030625309670791,
+      "learning_rate": 6.343500338123103e-06,
+      "loss": 0.8608,
+      "step": 25135
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 25.042553651334984,
+      "learning_rate": 6.341921530438501e-06,
+      "loss": 0.8593,
+      "step": 25140
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 50.62621748569941,
+      "learning_rate": 6.340342578558508e-06,
+      "loss": 0.9097,
+      "step": 25145
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 27.80881585341356,
+      "learning_rate": 6.338763482652791e-06,
+      "loss": 0.9368,
+      "step": 25150
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 61.609395311695614,
+      "learning_rate": 6.3371842428910264e-06,
+      "loss": 0.8344,
+      "step": 25155
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.816544428104855,
+      "learning_rate": 6.335604859442917e-06,
+      "loss": 0.8538,
+      "step": 25160
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.890584371153086,
+      "learning_rate": 6.3340253324781686e-06,
+      "loss": 0.9208,
+      "step": 25165
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.0517005574234455,
+      "learning_rate": 6.332445662166511e-06,
+      "loss": 0.8755,
+      "step": 25170
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.851264428215037,
+      "learning_rate": 6.330865848677688e-06,
+      "loss": 0.8904,
+      "step": 25175
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.452579228286874,
+      "learning_rate": 6.3292858921814545e-06,
+      "loss": 0.8392,
+      "step": 25180
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.108128960824086,
+      "learning_rate": 6.327705792847586e-06,
+      "loss": 0.8849,
+      "step": 25185
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.0810901540929905,
+      "learning_rate": 6.32612555084587e-06,
+      "loss": 0.9223,
+      "step": 25190
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.215293047065598,
+      "learning_rate": 6.324545166346111e-06,
+      "loss": 0.9109,
+      "step": 25195
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.738112141873573,
+      "learning_rate": 6.322964639518129e-06,
+      "loss": 0.85,
+      "step": 25200
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.278897458302731,
+      "learning_rate": 6.321383970531757e-06,
+      "loss": 0.9094,
+      "step": 25205
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.796606996809201,
+      "learning_rate": 6.3198031595568456e-06,
+      "loss": 0.8618,
+      "step": 25210
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.086265816029416,
+      "learning_rate": 6.318222206763261e-06,
+      "loss": 0.8958,
+      "step": 25215
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.688934732495671,
+      "learning_rate": 6.3166411123208805e-06,
+      "loss": 0.886,
+      "step": 25220
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.798233746830581,
+      "learning_rate": 6.315059876399603e-06,
+      "loss": 0.8873,
+      "step": 25225
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.26875537498929,
+      "learning_rate": 6.313478499169338e-06,
+      "loss": 0.8936,
+      "step": 25230
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.271522126023936,
+      "learning_rate": 6.311896980800012e-06,
+      "loss": 0.8911,
+      "step": 25235
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.11105347007813,
+      "learning_rate": 6.310315321461564e-06,
+      "loss": 0.8793,
+      "step": 25240
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.537558613075709,
+      "learning_rate": 6.308733521323953e-06,
+      "loss": 0.8907,
+      "step": 25245
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.966864864219315,
+      "learning_rate": 6.307151580557148e-06,
+      "loss": 0.8813,
+      "step": 25250
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.0511101707481165,
+      "learning_rate": 6.305569499331138e-06,
+      "loss": 0.8534,
+      "step": 25255
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.625865022087279,
+      "learning_rate": 6.303987277815923e-06,
+      "loss": 0.8478,
+      "step": 25260
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.251346315040928,
+      "learning_rate": 6.302404916181521e-06,
+      "loss": 0.8918,
+      "step": 25265
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.968406670522626,
+      "learning_rate": 6.300822414597962e-06,
+      "loss": 0.9049,
+      "step": 25270
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.945161424095243,
+      "learning_rate": 6.299239773235293e-06,
+      "loss": 0.9005,
+      "step": 25275
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.92724483468996,
+      "learning_rate": 6.297656992263577e-06,
+      "loss": 0.8903,
+      "step": 25280
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.52748808648185,
+      "learning_rate": 6.29607407185289e-06,
+      "loss": 0.8991,
+      "step": 25285
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.440892827504172,
+      "learning_rate": 6.294491012173325e-06,
+      "loss": 0.8745,
+      "step": 25290
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.138088117123807,
+      "learning_rate": 6.2929078133949865e-06,
+      "loss": 0.8244,
+      "step": 25295
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.072088311293955,
+      "learning_rate": 6.291324475687997e-06,
+      "loss": 0.8408,
+      "step": 25300
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.80878787504207,
+      "learning_rate": 6.289740999222495e-06,
+      "loss": 0.8961,
+      "step": 25305
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.100257753467226,
+      "learning_rate": 6.28815738416863e-06,
+      "loss": 0.8426,
+      "step": 25310
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.961299535659332,
+      "learning_rate": 6.286573630696569e-06,
+      "loss": 0.8381,
+      "step": 25315
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.865925335649451,
+      "learning_rate": 6.284989738976492e-06,
+      "loss": 0.8848,
+      "step": 25320
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 22.30800108665002,
+      "learning_rate": 6.283405709178598e-06,
+      "loss": 0.8981,
+      "step": 25325
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.862355353675248,
+      "learning_rate": 6.2818215414730965e-06,
+      "loss": 0.8677,
+      "step": 25330
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.434766157425761,
+      "learning_rate": 6.280237236030213e-06,
+      "loss": 0.8778,
+      "step": 25335
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.9510276868521865,
+      "learning_rate": 6.278652793020187e-06,
+      "loss": 0.8422,
+      "step": 25340
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.428288294984716,
+      "learning_rate": 6.277068212613276e-06,
+      "loss": 0.8578,
+      "step": 25345
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.324861746825624,
+      "learning_rate": 6.27548349497975e-06,
+      "loss": 0.8391,
+      "step": 25350
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.275352353676312,
+      "learning_rate": 6.273898640289894e-06,
+      "loss": 0.8343,
+      "step": 25355
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.770407383546127,
+      "learning_rate": 6.272313648714006e-06,
+      "loss": 0.8451,
+      "step": 25360
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.403999619241585,
+      "learning_rate": 6.270728520422402e-06,
+      "loss": 0.8631,
+      "step": 25365
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 17.52725588838729,
+      "learning_rate": 6.26914325558541e-06,
+      "loss": 0.8681,
+      "step": 25370
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 22.16613945415085,
+      "learning_rate": 6.2675578543733735e-06,
+      "loss": 0.8837,
+      "step": 25375
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.09014252592451,
+      "learning_rate": 6.2659723169566515e-06,
+      "loss": 0.9024,
+      "step": 25380
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.956239606675524,
+      "learning_rate": 6.2643866435056165e-06,
+      "loss": 0.9503,
+      "step": 25385
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.7753885835698435,
+      "learning_rate": 6.262800834190657e-06,
+      "loss": 0.8486,
+      "step": 25390
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.343432226257339,
+      "learning_rate": 6.261214889182174e-06,
+      "loss": 0.8774,
+      "step": 25395
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.291199180118161,
+      "learning_rate": 6.259628808650586e-06,
+      "loss": 0.9142,
+      "step": 25400
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.1642766099559285,
+      "learning_rate": 6.258042592766325e-06,
+      "loss": 0.8584,
+      "step": 25405
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.878353611957847,
+      "learning_rate": 6.256456241699833e-06,
+      "loss": 0.8818,
+      "step": 25410
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.138105161693613,
+      "learning_rate": 6.254869755621573e-06,
+      "loss": 0.8619,
+      "step": 25415
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.734859138661207,
+      "learning_rate": 6.253283134702019e-06,
+      "loss": 0.8822,
+      "step": 25420
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.454157811035056,
+      "learning_rate": 6.251696379111662e-06,
+      "loss": 0.8738,
+      "step": 25425
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.731280489534711,
+      "learning_rate": 6.250109489021004e-06,
+      "loss": 0.8902,
+      "step": 25430
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.863324415171886,
+      "learning_rate": 6.248522464600565e-06,
+      "loss": 0.9407,
+      "step": 25435
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.242334946554647,
+      "learning_rate": 6.246935306020877e-06,
+      "loss": 0.8491,
+      "step": 25440
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.69778428285142,
+      "learning_rate": 6.245348013452487e-06,
+      "loss": 0.8757,
+      "step": 25445
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.971461688685569,
+      "learning_rate": 6.2437605870659565e-06,
+      "loss": 0.9059,
+      "step": 25450
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.889903360165224,
+      "learning_rate": 6.242173027031862e-06,
+      "loss": 0.8762,
+      "step": 25455
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.757845037341863,
+      "learning_rate": 6.240585333520793e-06,
+      "loss": 0.8612,
+      "step": 25460
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.4613410407274925,
+      "learning_rate": 6.238997506703357e-06,
+      "loss": 0.8858,
+      "step": 25465
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.133444905801013,
+      "learning_rate": 6.237409546750169e-06,
+      "loss": 0.8537,
+      "step": 25470
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 25.661710255151302,
+      "learning_rate": 6.235821453831864e-06,
+      "loss": 0.8836,
+      "step": 25475
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.957654960348783,
+      "learning_rate": 6.23423322811909e-06,
+      "loss": 0.895,
+      "step": 25480
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.80550971294075,
+      "learning_rate": 6.232644869782508e-06,
+      "loss": 0.86,
+      "step": 25485
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.606331000684494,
+      "learning_rate": 6.231056378992797e-06,
+      "loss": 0.9075,
+      "step": 25490
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.300816701891632,
+      "learning_rate": 6.229467755920643e-06,
+      "loss": 0.8679,
+      "step": 25495
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.333988336117034,
+      "learning_rate": 6.227879000736753e-06,
+      "loss": 0.9036,
+      "step": 25500
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.192784693661378,
+      "learning_rate": 6.226290113611845e-06,
+      "loss": 0.9143,
+      "step": 25505
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.473003754287832,
+      "learning_rate": 6.224701094716652e-06,
+      "loss": 0.8714,
+      "step": 25510
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.326087216702297,
+      "learning_rate": 6.2231119442219225e-06,
+      "loss": 0.8601,
+      "step": 25515
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.333098860162012,
+      "learning_rate": 6.2215226622984155e-06,
+      "loss": 0.8855,
+      "step": 25520
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.376095325289636,
+      "learning_rate": 6.219933249116908e-06,
+      "loss": 0.8789,
+      "step": 25525
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.533668475310641,
+      "learning_rate": 6.218343704848189e-06,
+      "loss": 0.9266,
+      "step": 25530
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.932136890663288,
+      "learning_rate": 6.216754029663061e-06,
+      "loss": 0.8689,
+      "step": 25535
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.184289943538942,
+      "learning_rate": 6.2151642237323436e-06,
+      "loss": 0.9176,
+      "step": 25540
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.054911218312706,
+      "learning_rate": 6.2135742872268665e-06,
+      "loss": 0.8163,
+      "step": 25545
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.531494243211934,
+      "learning_rate": 6.211984220317479e-06,
+      "loss": 0.8368,
+      "step": 25550
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.439367060763998,
+      "learning_rate": 6.210394023175036e-06,
+      "loss": 0.895,
+      "step": 25555
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.1587734485685965,
+      "learning_rate": 6.208803695970414e-06,
+      "loss": 0.8906,
+      "step": 25560
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.0237307675450555,
+      "learning_rate": 6.207213238874498e-06,
+      "loss": 0.8784,
+      "step": 25565
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.224329296527005,
+      "learning_rate": 6.205622652058192e-06,
+      "loss": 0.8613,
+      "step": 25570
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 15.720986066416545,
+      "learning_rate": 6.2040319356924136e-06,
+      "loss": 0.856,
+      "step": 25575
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.276578756297132,
+      "learning_rate": 6.202441089948088e-06,
+      "loss": 0.8646,
+      "step": 25580
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.281021723958332,
+      "learning_rate": 6.200850114996162e-06,
+      "loss": 0.8785,
+      "step": 25585
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.392830282437864,
+      "learning_rate": 6.19925901100759e-06,
+      "loss": 0.8752,
+      "step": 25590
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.762018802190973,
+      "learning_rate": 6.197667778153345e-06,
+      "loss": 0.8433,
+      "step": 25595
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.028238494354232,
+      "learning_rate": 6.196076416604412e-06,
+      "loss": 0.8203,
+      "step": 25600
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.853377348294338,
+      "learning_rate": 6.19448492653179e-06,
+      "loss": 0.8125,
+      "step": 25605
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.415364521392845,
+      "learning_rate": 6.192893308106488e-06,
+      "loss": 0.8794,
+      "step": 25610
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.918289053201691,
+      "learning_rate": 6.191301561499537e-06,
+      "loss": 0.9117,
+      "step": 25615
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 17.442732371636694,
+      "learning_rate": 6.1897096868819764e-06,
+      "loss": 0.9287,
+      "step": 25620
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.7851664084420396,
+      "learning_rate": 6.188117684424859e-06,
+      "loss": 0.847,
+      "step": 25625
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.601353429277399,
+      "learning_rate": 6.186525554299252e-06,
+      "loss": 0.8426,
+      "step": 25630
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.7269618683439685,
+      "learning_rate": 6.18493329667624e-06,
+      "loss": 0.8715,
+      "step": 25635
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.231706552876182,
+      "learning_rate": 6.183340911726914e-06,
+      "loss": 0.8436,
+      "step": 25640
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.4517877216033455,
+      "learning_rate": 6.181748399622385e-06,
+      "loss": 0.9128,
+      "step": 25645
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.578331933065757,
+      "learning_rate": 6.180155760533773e-06,
+      "loss": 0.814,
+      "step": 25650
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 19.07324712853925,
+      "learning_rate": 6.178562994632219e-06,
+      "loss": 0.8264,
+      "step": 25655
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 17.345564566748667,
+      "learning_rate": 6.176970102088869e-06,
+      "loss": 0.8393,
+      "step": 25660
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.14160681854394,
+      "learning_rate": 6.175377083074886e-06,
+      "loss": 0.9265,
+      "step": 25665
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 19.381696915187014,
+      "learning_rate": 6.1737839377614485e-06,
+      "loss": 0.8998,
+      "step": 25670
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.339583608213842,
+      "learning_rate": 6.172190666319747e-06,
+      "loss": 0.8295,
+      "step": 25675
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.249527397611181,
+      "learning_rate": 6.170597268920986e-06,
+      "loss": 0.8858,
+      "step": 25680
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.136525711987327,
+      "learning_rate": 6.169003745736381e-06,
+      "loss": 0.8748,
+      "step": 25685
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.6186409880445645,
+      "learning_rate": 6.1674100969371655e-06,
+      "loss": 0.8481,
+      "step": 25690
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.968110246671394,
+      "learning_rate": 6.165816322694582e-06,
+      "loss": 0.8588,
+      "step": 25695
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.981580135910331,
+      "learning_rate": 6.1642224231798885e-06,
+      "loss": 0.8405,
+      "step": 25700
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.7395863009387424,
+      "learning_rate": 6.162628398564358e-06,
+      "loss": 0.9008,
+      "step": 25705
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.99837936229072,
+      "learning_rate": 6.161034249019275e-06,
+      "loss": 0.8874,
+      "step": 25710
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.393926376479731,
+      "learning_rate": 6.1594399747159396e-06,
+      "loss": 0.8948,
+      "step": 25715
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.96398760200713,
+      "learning_rate": 6.15784557582566e-06,
+      "loss": 0.8327,
+      "step": 25720
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.733698169791028,
+      "learning_rate": 6.1562510525197645e-06,
+      "loss": 0.8984,
+      "step": 25725
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.229570441270813,
+      "learning_rate": 6.15465640496959e-06,
+      "loss": 0.8901,
+      "step": 25730
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.39025469930247,
+      "learning_rate": 6.1530616333464895e-06,
+      "loss": 0.9036,
+      "step": 25735
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.330411512117964,
+      "learning_rate": 6.151466737821827e-06,
+      "loss": 0.8556,
+      "step": 25740
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.757175274270205,
+      "learning_rate": 6.149871718566982e-06,
+      "loss": 0.9078,
+      "step": 25745
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.541428998012914,
+      "learning_rate": 6.1482765757533455e-06,
+      "loss": 0.8349,
+      "step": 25750
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.20281935323804,
+      "learning_rate": 6.146681309552324e-06,
+      "loss": 0.8623,
+      "step": 25755
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.684748046413062,
+      "learning_rate": 6.145085920135334e-06,
+      "loss": 0.8665,
+      "step": 25760
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.782285328980653,
+      "learning_rate": 6.143490407673809e-06,
+      "loss": 0.8454,
+      "step": 25765
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.793694387236938,
+      "learning_rate": 6.141894772339193e-06,
+      "loss": 0.8941,
+      "step": 25770
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.392990744236515,
+      "learning_rate": 6.140299014302945e-06,
+      "loss": 0.8868,
+      "step": 25775
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.538279445165557,
+      "learning_rate": 6.1387031337365346e-06,
+      "loss": 0.9097,
+      "step": 25780
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 37.88139639419802,
+      "learning_rate": 6.137107130811447e-06,
+      "loss": 0.8684,
+      "step": 25785
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.763698477540671,
+      "learning_rate": 6.13551100569918e-06,
+      "loss": 0.8093,
+      "step": 25790
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.76037090425412,
+      "learning_rate": 6.1339147585712435e-06,
+      "loss": 0.878,
+      "step": 25795
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.012779298640561,
+      "learning_rate": 6.132318389599161e-06,
+      "loss": 0.8586,
+      "step": 25800
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.446305715107773,
+      "learning_rate": 6.130721898954472e-06,
+      "loss": 0.8847,
+      "step": 25805
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.20015550247254,
+      "learning_rate": 6.129125286808723e-06,
+      "loss": 0.9221,
+      "step": 25810
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.865817013173404,
+      "learning_rate": 6.12752855333348e-06,
+      "loss": 0.8899,
+      "step": 25815
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.086374757712554,
+      "learning_rate": 6.125931698700318e-06,
+      "loss": 0.8559,
+      "step": 25820
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.124075536029831,
+      "learning_rate": 6.124334723080826e-06,
+      "loss": 0.8383,
+      "step": 25825
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.621989313643415,
+      "learning_rate": 6.122737626646605e-06,
+      "loss": 0.826,
+      "step": 25830
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.507963313798742,
+      "learning_rate": 6.121140409569273e-06,
+      "loss": 0.8725,
+      "step": 25835
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.085110510275198,
+      "learning_rate": 6.119543072020453e-06,
+      "loss": 0.8643,
+      "step": 25840
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.8194383645461665,
+      "learning_rate": 6.117945614171789e-06,
+      "loss": 0.9087,
+      "step": 25845
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.125330358039872,
+      "learning_rate": 6.1163480361949366e-06,
+      "loss": 0.8626,
+      "step": 25850
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.824130877205306,
+      "learning_rate": 6.114750338261562e-06,
+      "loss": 0.8789,
+      "step": 25855
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.466364782295777,
+      "learning_rate": 6.113152520543342e-06,
+      "loss": 0.8464,
+      "step": 25860
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 27.772788330067467,
+      "learning_rate": 6.111554583211972e-06,
+      "loss": 0.8352,
+      "step": 25865
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.724817065075063,
+      "learning_rate": 6.109956526439154e-06,
+      "loss": 0.894,
+      "step": 25870
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.741257206597169,
+      "learning_rate": 6.108358350396612e-06,
+      "loss": 0.8662,
+      "step": 25875
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.590538278157433,
+      "learning_rate": 6.106760055256071e-06,
+      "loss": 0.848,
+      "step": 25880
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.634961584039799,
+      "learning_rate": 6.1051616411892785e-06,
+      "loss": 0.8735,
+      "step": 25885
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.136378425841045,
+      "learning_rate": 6.10356310836799e-06,
+      "loss": 0.8728,
+      "step": 25890
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.0965211451313905,
+      "learning_rate": 6.101964456963976e-06,
+      "loss": 0.9167,
+      "step": 25895
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 19.23678514989524,
+      "learning_rate": 6.100365687149017e-06,
+      "loss": 0.8554,
+      "step": 25900
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.77477074084548,
+      "learning_rate": 6.098766799094909e-06,
+      "loss": 0.8947,
+      "step": 25905
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.899352467156238,
+      "learning_rate": 6.097167792973458e-06,
+      "loss": 0.8689,
+      "step": 25910
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.032953600140125,
+      "learning_rate": 6.0955686689564865e-06,
+      "loss": 0.8986,
+      "step": 25915
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.290130421515645,
+      "learning_rate": 6.093969427215827e-06,
+      "loss": 0.8044,
+      "step": 25920
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 53.66729131657105,
+      "learning_rate": 6.092370067923323e-06,
+      "loss": 0.8699,
+      "step": 25925
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 32.50415566669549,
+      "learning_rate": 6.090770591250835e-06,
+      "loss": 0.8465,
+      "step": 25930
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.308652352161277,
+      "learning_rate": 6.089170997370233e-06,
+      "loss": 0.8476,
+      "step": 25935
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.15751096832169,
+      "learning_rate": 6.0875712864534e-06,
+      "loss": 0.8467,
+      "step": 25940
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.654747038364158,
+      "learning_rate": 6.085971458672233e-06,
+      "loss": 0.8663,
+      "step": 25945
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.02574199828502,
+      "learning_rate": 6.084371514198641e-06,
+      "loss": 0.9138,
+      "step": 25950
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 22.236670726774232,
+      "learning_rate": 6.0827714532045425e-06,
+      "loss": 0.8634,
+      "step": 25955
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 29.715118450946402,
+      "learning_rate": 6.081171275861873e-06,
+      "loss": 0.8551,
+      "step": 25960
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.779398096974892,
+      "learning_rate": 6.0795709823425784e-06,
+      "loss": 0.8719,
+      "step": 25965
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.369736599047194,
+      "learning_rate": 6.077970572818618e-06,
+      "loss": 0.9093,
+      "step": 25970
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.121210471723412,
+      "learning_rate": 6.076370047461961e-06,
+      "loss": 0.8434,
+      "step": 25975
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.907771106995737,
+      "learning_rate": 6.074769406444592e-06,
+      "loss": 0.8584,
+      "step": 25980
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.36401161040538,
+      "learning_rate": 6.073168649938508e-06,
+      "loss": 0.8572,
+      "step": 25985
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.44906106688398,
+      "learning_rate": 6.071567778115717e-06,
+      "loss": 0.8313,
+      "step": 25990
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.738736637561908,
+      "learning_rate": 6.069966791148237e-06,
+      "loss": 0.8937,
+      "step": 25995
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 24.246733754892638,
+      "learning_rate": 6.068365689208105e-06,
+      "loss": 0.9566,
+      "step": 26000
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.41108631179515,
+      "learning_rate": 6.066764472467364e-06,
+      "loss": 0.8699,
+      "step": 26005
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.306770724715992,
+      "learning_rate": 6.065163141098072e-06,
+      "loss": 0.8423,
+      "step": 26010
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.272191908306718,
+      "learning_rate": 6.063561695272299e-06,
+      "loss": 0.8599,
+      "step": 26015
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.055455827171254,
+      "learning_rate": 6.061960135162129e-06,
+      "loss": 0.8612,
+      "step": 26020
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.359373771983753,
+      "learning_rate": 6.060358460939655e-06,
+      "loss": 0.8299,
+      "step": 26025
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.059908628353393,
+      "learning_rate": 6.058756672776984e-06,
+      "loss": 0.8333,
+      "step": 26030
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.1028732743827,
+      "learning_rate": 6.057154770846239e-06,
+      "loss": 0.8876,
+      "step": 26035
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.869030784604681,
+      "learning_rate": 6.055552755319545e-06,
+      "loss": 0.8168,
+      "step": 26040
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.2862331595314,
+      "learning_rate": 6.053950626369051e-06,
+      "loss": 0.8742,
+      "step": 26045
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.788579628932328,
+      "learning_rate": 6.0523483841669085e-06,
+      "loss": 0.8712,
+      "step": 26050
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.030578341239467,
+      "learning_rate": 6.050746028885289e-06,
+      "loss": 0.9296,
+      "step": 26055
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.468570397588891,
+      "learning_rate": 6.049143560696374e-06,
+      "loss": 0.8109,
+      "step": 26060
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.8571517956563435,
+      "learning_rate": 6.04754097977235e-06,
+      "loss": 0.8524,
+      "step": 26065
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.333615100612435,
+      "learning_rate": 6.045938286285424e-06,
+      "loss": 0.9171,
+      "step": 26070
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.157846780627792,
+      "learning_rate": 6.044335480407815e-06,
+      "loss": 0.8542,
+      "step": 26075
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.195776752013943,
+      "learning_rate": 6.042732562311749e-06,
+      "loss": 0.8581,
+      "step": 26080
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.325092536158999,
+      "learning_rate": 6.041129532169466e-06,
+      "loss": 0.8757,
+      "step": 26085
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.254949243153465,
+      "learning_rate": 6.039526390153222e-06,
+      "loss": 0.8842,
+      "step": 26090
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.11847454594186,
+      "learning_rate": 6.037923136435279e-06,
+      "loss": 0.8913,
+      "step": 26095
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 18.89783607734687,
+      "learning_rate": 6.036319771187914e-06,
+      "loss": 0.8695,
+      "step": 26100
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.36171291397209,
+      "learning_rate": 6.034716294583416e-06,
+      "loss": 0.916,
+      "step": 26105
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.259638650383265,
+      "learning_rate": 6.033112706794085e-06,
+      "loss": 0.8317,
+      "step": 26110
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 18.1173258846941,
+      "learning_rate": 6.031509007992235e-06,
+      "loss": 0.841,
+      "step": 26115
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.863621515780087,
+      "learning_rate": 6.029905198350189e-06,
+      "loss": 0.8602,
+      "step": 26120
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.7387268398581375,
+      "learning_rate": 6.028301278040285e-06,
+      "loss": 0.8569,
+      "step": 26125
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.180831099641892,
+      "learning_rate": 6.02669724723487e-06,
+      "loss": 0.8721,
+      "step": 26130
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.1811762275798,
+      "learning_rate": 6.025093106106305e-06,
+      "loss": 0.8745,
+      "step": 26135
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.04758399463794,
+      "learning_rate": 6.0234888548269596e-06,
+      "loss": 0.8167,
+      "step": 26140
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.843301870825521,
+      "learning_rate": 6.0218844935692225e-06,
+      "loss": 0.8116,
+      "step": 26145
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.742408571969738,
+      "learning_rate": 6.020280022505487e-06,
+      "loss": 0.8639,
+      "step": 26150
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.883430245060165,
+      "learning_rate": 6.018675441808158e-06,
+      "loss": 0.8507,
+      "step": 26155
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.442113503186823,
+      "learning_rate": 6.017070751649657e-06,
+      "loss": 0.8663,
+      "step": 26160
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.87351851293303,
+      "learning_rate": 6.015465952202415e-06,
+      "loss": 0.8642,
+      "step": 26165
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.980669194673036,
+      "learning_rate": 6.013861043638877e-06,
+      "loss": 0.8512,
+      "step": 26170
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.7478501276795,
+      "learning_rate": 6.012256026131495e-06,
+      "loss": 0.7938,
+      "step": 26175
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.796242570720157,
+      "learning_rate": 6.010650899852733e-06,
+      "loss": 0.8971,
+      "step": 26180
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.763599465635087,
+      "learning_rate": 6.009045664975074e-06,
+      "loss": 0.8306,
+      "step": 26185
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.694555344096171,
+      "learning_rate": 6.007440321671005e-06,
+      "loss": 0.8465,
+      "step": 26190
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.829992844317758,
+      "learning_rate": 6.005834870113025e-06,
+      "loss": 0.8198,
+      "step": 26195
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.949152588181757,
+      "learning_rate": 6.0042293104736505e-06,
+      "loss": 0.8348,
+      "step": 26200
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.863277357566641,
+      "learning_rate": 6.002623642925405e-06,
+      "loss": 0.8549,
+      "step": 26205
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.242921882279838,
+      "learning_rate": 6.001017867640824e-06,
+      "loss": 0.8744,
+      "step": 26210
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.58829460970268,
+      "learning_rate": 5.999411984792455e-06,
+      "loss": 0.8297,
+      "step": 26215
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.809968365257946,
+      "learning_rate": 5.9978059945528575e-06,
+      "loss": 0.8587,
+      "step": 26220
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.15059994130675,
+      "learning_rate": 5.996199897094602e-06,
+      "loss": 0.876,
+      "step": 26225
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.983579632267533,
+      "learning_rate": 5.994593692590272e-06,
+      "loss": 0.8651,
+      "step": 26230
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 23.86432844905842,
+      "learning_rate": 5.9929873812124595e-06,
+      "loss": 0.8433,
+      "step": 26235
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 20.35663173936695,
+      "learning_rate": 5.991380963133771e-06,
+      "loss": 0.8552,
+      "step": 26240
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.216820046992261,
+      "learning_rate": 5.989774438526822e-06,
+      "loss": 0.8551,
+      "step": 26245
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.447418647454864,
+      "learning_rate": 5.988167807564241e-06,
+      "loss": 0.8571,
+      "step": 26250
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.416570965490965,
+      "learning_rate": 5.98656107041867e-06,
+      "loss": 0.8286,
+      "step": 26255
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.353861550378891,
+      "learning_rate": 5.984954227262756e-06,
+      "loss": 0.8744,
+      "step": 26260
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.267075615125807,
+      "learning_rate": 5.983347278269164e-06,
+      "loss": 0.859,
+      "step": 26265
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.84684410680608,
+      "learning_rate": 5.981740223610566e-06,
+      "loss": 0.8676,
+      "step": 26270
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.334338786403093,
+      "learning_rate": 5.98013306345965e-06,
+      "loss": 0.8853,
+      "step": 26275
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.6244967967702,
+      "learning_rate": 5.97852579798911e-06,
+      "loss": 0.8575,
+      "step": 26280
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.397749346676092,
+      "learning_rate": 5.976918427371654e-06,
+      "loss": 0.843,
+      "step": 26285
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.880693302654921,
+      "learning_rate": 5.975310951780002e-06,
+      "loss": 0.8969,
+      "step": 26290
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.217466712483155,
+      "learning_rate": 5.9737033713868846e-06,
+      "loss": 0.8426,
+      "step": 26295
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.225582960100946,
+      "learning_rate": 5.9720956863650405e-06,
+      "loss": 0.8841,
+      "step": 26300
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.776909165116245,
+      "learning_rate": 5.970487896887226e-06,
+      "loss": 0.8247,
+      "step": 26305
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.292319378014117,
+      "learning_rate": 5.968880003126203e-06,
+      "loss": 0.8048,
+      "step": 26310
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.6203860612520415,
+      "learning_rate": 5.96727200525475e-06,
+      "loss": 0.8799,
+      "step": 26315
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.8572292503413985,
+      "learning_rate": 5.96566390344565e-06,
+      "loss": 0.8938,
+      "step": 26320
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.877072267245683,
+      "learning_rate": 5.964055697871701e-06,
+      "loss": 0.8483,
+      "step": 26325
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.612855111178123,
+      "learning_rate": 5.962447388705713e-06,
+      "loss": 0.8452,
+      "step": 26330
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.170553941416647,
+      "learning_rate": 5.960838976120505e-06,
+      "loss": 0.8697,
+      "step": 26335
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.846951533533244,
+      "learning_rate": 5.95923046028891e-06,
+      "loss": 0.8521,
+      "step": 26340
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.58267276833323,
+      "learning_rate": 5.957621841383767e-06,
+      "loss": 0.9076,
+      "step": 26345
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.529739733706275,
+      "learning_rate": 5.9560131195779315e-06,
+      "loss": 0.8344,
+      "step": 26350
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.242639917064167,
+      "learning_rate": 5.954404295044267e-06,
+      "loss": 0.8724,
+      "step": 26355
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.338437814497742,
+      "learning_rate": 5.952795367955649e-06,
+      "loss": 0.822,
+      "step": 26360
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.543310735724762,
+      "learning_rate": 5.951186338484964e-06,
+      "loss": 0.8386,
+      "step": 26365
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.260121446580111,
+      "learning_rate": 5.9495772068051085e-06,
+      "loss": 0.9259,
+      "step": 26370
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.784040658635956,
+      "learning_rate": 5.947967973088992e-06,
+      "loss": 0.8829,
+      "step": 26375
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.65086281036451,
+      "learning_rate": 5.946358637509533e-06,
+      "loss": 0.882,
+      "step": 26380
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.808652705047955,
+      "learning_rate": 5.944749200239661e-06,
+      "loss": 0.8612,
+      "step": 26385
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.622312197747917,
+      "learning_rate": 5.943139661452319e-06,
+      "loss": 0.8575,
+      "step": 26390
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.163361460036183,
+      "learning_rate": 5.941530021320456e-06,
+      "loss": 0.8847,
+      "step": 26395
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.766205106282325,
+      "learning_rate": 5.939920280017038e-06,
+      "loss": 0.8476,
+      "step": 26400
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 5.365436882565832,
+      "learning_rate": 5.938310437715038e-06,
+      "loss": 0.868,
+      "step": 26405
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.429499550649396,
+      "learning_rate": 5.936700494587441e-06,
+      "loss": 0.8616,
+      "step": 26410
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.52896133360857,
+      "learning_rate": 5.935090450807242e-06,
+      "loss": 0.8435,
+      "step": 26415
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.2322821234426815,
+      "learning_rate": 5.933480306547445e-06,
+      "loss": 0.8653,
+      "step": 26420
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.154628254643407,
+      "learning_rate": 5.931870061981071e-06,
+      "loss": 0.8732,
+      "step": 26425
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 5.227823427154511,
+      "learning_rate": 5.930259717281146e-06,
+      "loss": 0.8819,
+      "step": 26430
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.9315025295618895,
+      "learning_rate": 5.928649272620709e-06,
+      "loss": 0.8692,
+      "step": 26435
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 5.54624631506354,
+      "learning_rate": 5.927038728172809e-06,
+      "loss": 0.8403,
+      "step": 26440
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.683362399276405,
+      "learning_rate": 5.925428084110506e-06,
+      "loss": 0.8477,
+      "step": 26445
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.459554397267144,
+      "learning_rate": 5.923817340606871e-06,
+      "loss": 0.8812,
+      "step": 26450
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.7061135470042155,
+      "learning_rate": 5.922206497834986e-06,
+      "loss": 0.854,
+      "step": 26455
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 5.2193204632723065,
+      "learning_rate": 5.920595555967944e-06,
+      "loss": 0.8174,
+      "step": 26460
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.599737835293299,
+      "learning_rate": 5.918984515178845e-06,
+      "loss": 0.8598,
+      "step": 26465
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.635118514467898,
+      "learning_rate": 5.917373375640805e-06,
+      "loss": 0.8705,
+      "step": 26470
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.687349928354102,
+      "learning_rate": 5.915762137526947e-06,
+      "loss": 0.8505,
+      "step": 26475
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.885049284565121,
+      "learning_rate": 5.914150801010406e-06,
+      "loss": 0.9069,
+      "step": 26480
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.873364750571962,
+      "learning_rate": 5.912539366264325e-06,
+      "loss": 0.9065,
+      "step": 26485
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.753382422178554,
+      "learning_rate": 5.9109278334618634e-06,
+      "loss": 0.8216,
+      "step": 26490
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.780003970057351,
+      "learning_rate": 5.909316202776185e-06,
+      "loss": 0.8263,
+      "step": 26495
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.34209097689926,
+      "learning_rate": 5.907704474380469e-06,
+      "loss": 0.8689,
+      "step": 26500
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.97635300998426,
+      "learning_rate": 5.9060926484479e-06,
+      "loss": 0.8586,
+      "step": 26505
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.604465766189382,
+      "learning_rate": 5.904480725151677e-06,
+      "loss": 0.8824,
+      "step": 26510
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.891633120947663,
+      "learning_rate": 5.902868704665008e-06,
+      "loss": 0.8252,
+      "step": 26515
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.571455838359508,
+      "learning_rate": 5.901256587161114e-06,
+      "loss": 0.8808,
+      "step": 26520
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.274679742798048,
+      "learning_rate": 5.899644372813221e-06,
+      "loss": 0.8805,
+      "step": 26525
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.93466942095143,
+      "learning_rate": 5.898032061794569e-06,
+      "loss": 0.8652,
+      "step": 26530
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.48594329453608,
+      "learning_rate": 5.896419654278409e-06,
+      "loss": 0.9081,
+      "step": 26535
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.709402291802655,
+      "learning_rate": 5.894807150438e-06,
+      "loss": 0.8297,
+      "step": 26540
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.842407234736854,
+      "learning_rate": 5.893194550446616e-06,
+      "loss": 0.8525,
+      "step": 26545
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.981793728229583,
+      "learning_rate": 5.891581854477535e-06,
+      "loss": 0.8435,
+      "step": 26550
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.866464491317057,
+      "learning_rate": 5.889969062704047e-06,
+      "loss": 0.8778,
+      "step": 26555
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.931442225747622,
+      "learning_rate": 5.888356175299457e-06,
+      "loss": 0.8712,
+      "step": 26560
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 31.89315978023349,
+      "learning_rate": 5.886743192437076e-06,
+      "loss": 0.8997,
+      "step": 26565
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.0658673776465,
+      "learning_rate": 5.885130114290224e-06,
+      "loss": 0.8914,
+      "step": 26570
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.887782399973648,
+      "learning_rate": 5.883516941032235e-06,
+      "loss": 0.9017,
+      "step": 26575
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 26.208381565763595,
+      "learning_rate": 5.881903672836452e-06,
+      "loss": 0.8956,
+      "step": 26580
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.179136283971342,
+      "learning_rate": 5.8802903098762265e-06,
+      "loss": 0.8898,
+      "step": 26585
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.364618616373024,
+      "learning_rate": 5.878676852324922e-06,
+      "loss": 0.8734,
+      "step": 26590
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.833551058429732,
+      "learning_rate": 5.877063300355913e-06,
+      "loss": 0.8445,
+      "step": 26595
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.814225852311926,
+      "learning_rate": 5.875449654142581e-06,
+      "loss": 0.831,
+      "step": 26600
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.366540745200195,
+      "learning_rate": 5.873835913858321e-06,
+      "loss": 0.8436,
+      "step": 26605
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.764322985703324,
+      "learning_rate": 5.872222079676535e-06,
+      "loss": 0.8883,
+      "step": 26610
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.471545870336112,
+      "learning_rate": 5.870608151770637e-06,
+      "loss": 0.8157,
+      "step": 26615
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.889854059013036,
+      "learning_rate": 5.86899413031405e-06,
+      "loss": 0.8686,
+      "step": 26620
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.941368536110251,
+      "learning_rate": 5.867380015480208e-06,
+      "loss": 0.8934,
+      "step": 26625
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 14.955809733667413,
+      "learning_rate": 5.865765807442556e-06,
+      "loss": 0.8227,
+      "step": 26630
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.9779549529841205,
+      "learning_rate": 5.864151506374549e-06,
+      "loss": 0.8634,
+      "step": 26635
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.1885930870661845,
+      "learning_rate": 5.862537112449647e-06,
+      "loss": 0.859,
+      "step": 26640
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.758706652338646,
+      "learning_rate": 5.860922625841325e-06,
+      "loss": 0.9062,
+      "step": 26645
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.423297224186076,
+      "learning_rate": 5.859308046723069e-06,
+      "loss": 0.8836,
+      "step": 26650
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 20.812369359611463,
+      "learning_rate": 5.857693375268371e-06,
+      "loss": 0.8479,
+      "step": 26655
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.979318288189456,
+      "learning_rate": 5.856078611650734e-06,
+      "loss": 0.8498,
+      "step": 26660
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.343566488232046,
+      "learning_rate": 5.85446375604367e-06,
+      "loss": 0.8483,
+      "step": 26665
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.815669932575958,
+      "learning_rate": 5.852848808620707e-06,
+      "loss": 0.8853,
+      "step": 26670
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.294733499794159,
+      "learning_rate": 5.851233769555374e-06,
+      "loss": 0.8478,
+      "step": 26675
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.788783041888735,
+      "learning_rate": 5.849618639021216e-06,
+      "loss": 0.8931,
+      "step": 26680
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.2797218962180015,
+      "learning_rate": 5.848003417191787e-06,
+      "loss": 0.8236,
+      "step": 26685
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.4128678818500875,
+      "learning_rate": 5.846388104240648e-06,
+      "loss": 0.8344,
+      "step": 26690
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.388001643542782,
+      "learning_rate": 5.844772700341371e-06,
+      "loss": 0.8416,
+      "step": 26695
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.3996933906459805,
+      "learning_rate": 5.843157205667539e-06,
+      "loss": 0.8516,
+      "step": 26700
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.619516343738756,
+      "learning_rate": 5.841541620392744e-06,
+      "loss": 0.8907,
+      "step": 26705
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.568510711522257,
+      "learning_rate": 5.839925944690588e-06,
+      "loss": 0.8166,
+      "step": 26710
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.934216784304126,
+      "learning_rate": 5.838310178734683e-06,
+      "loss": 0.8277,
+      "step": 26715
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.345659943397673,
+      "learning_rate": 5.836694322698648e-06,
+      "loss": 0.8328,
+      "step": 26720
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.141023314856025,
+      "learning_rate": 5.835078376756117e-06,
+      "loss": 0.8593,
+      "step": 26725
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.783991886082312,
+      "learning_rate": 5.833462341080729e-06,
+      "loss": 0.8385,
+      "step": 26730
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.255356798585868,
+      "learning_rate": 5.831846215846133e-06,
+      "loss": 0.8574,
+      "step": 26735
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.131233336217069,
+      "learning_rate": 5.8302300012259905e-06,
+      "loss": 0.8325,
+      "step": 26740
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.702782139324231,
+      "learning_rate": 5.828613697393971e-06,
+      "loss": 0.9062,
+      "step": 26745
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.040898225230506,
+      "learning_rate": 5.826997304523752e-06,
+      "loss": 0.852,
+      "step": 26750
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 20.11270605309605,
+      "learning_rate": 5.825380822789023e-06,
+      "loss": 0.8876,
+      "step": 26755
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.134634927485443,
+      "learning_rate": 5.823764252363481e-06,
+      "loss": 0.8443,
+      "step": 26760
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.645463287262561,
+      "learning_rate": 5.822147593420835e-06,
+      "loss": 0.8999,
+      "step": 26765
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.396966521136354,
+      "learning_rate": 5.820530846134802e-06,
+      "loss": 0.833,
+      "step": 26770
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.078969221794201,
+      "learning_rate": 5.818914010679109e-06,
+      "loss": 0.8709,
+      "step": 26775
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.447435177422953,
+      "learning_rate": 5.8172970872274905e-06,
+      "loss": 0.8583,
+      "step": 26780
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.762526281484488,
+      "learning_rate": 5.815680075953692e-06,
+      "loss": 0.8843,
+      "step": 26785
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.341490315359552,
+      "learning_rate": 5.814062977031471e-06,
+      "loss": 0.87,
+      "step": 26790
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.570139263823231,
+      "learning_rate": 5.81244579063459e-06,
+      "loss": 0.8592,
+      "step": 26795
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.324585695361593,
+      "learning_rate": 5.810828516936823e-06,
+      "loss": 0.8609,
+      "step": 26800
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.2786761122367,
+      "learning_rate": 5.809211156111954e-06,
+      "loss": 0.7837,
+      "step": 26805
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.115853743290486,
+      "learning_rate": 5.807593708333774e-06,
+      "loss": 0.865,
+      "step": 26810
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.533442451737326,
+      "learning_rate": 5.805976173776087e-06,
+      "loss": 0.809,
+      "step": 26815
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.912855762982474,
+      "learning_rate": 5.804358552612702e-06,
+      "loss": 0.8432,
+      "step": 26820
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.80822625181859,
+      "learning_rate": 5.802740845017441e-06,
+      "loss": 0.8475,
+      "step": 26825
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.190074610071065,
+      "learning_rate": 5.801123051164132e-06,
+      "loss": 0.8487,
+      "step": 26830
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.120069966431346,
+      "learning_rate": 5.799505171226619e-06,
+      "loss": 0.8375,
+      "step": 26835
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.389772520084062,
+      "learning_rate": 5.797887205378745e-06,
+      "loss": 0.8214,
+      "step": 26840
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.410071650198306,
+      "learning_rate": 5.79626915379437e-06,
+      "loss": 0.8389,
+      "step": 26845
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.534543475819742,
+      "learning_rate": 5.794651016647359e-06,
+      "loss": 0.8018,
+      "step": 26850
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.047415150480368,
+      "learning_rate": 5.7930327941115886e-06,
+      "loss": 0.8607,
+      "step": 26855
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.63189317376628,
+      "learning_rate": 5.7914144863609465e-06,
+      "loss": 0.8406,
+      "step": 26860
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.846280295668944,
+      "learning_rate": 5.789796093569324e-06,
+      "loss": 0.8312,
+      "step": 26865
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.664091825460446,
+      "learning_rate": 5.788177615910626e-06,
+      "loss": 0.8888,
+      "step": 26870
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.129029328346334,
+      "learning_rate": 5.786559053558765e-06,
+      "loss": 0.8246,
+      "step": 26875
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.73827927659292,
+      "learning_rate": 5.784940406687662e-06,
+      "loss": 0.8094,
+      "step": 26880
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 21.90043747894924,
+      "learning_rate": 5.783321675471248e-06,
+      "loss": 0.8565,
+      "step": 26885
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 31.415125837571356,
+      "learning_rate": 5.781702860083463e-06,
+      "loss": 0.824,
+      "step": 26890
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 40.69209896384807,
+      "learning_rate": 5.780083960698255e-06,
+      "loss": 0.8858,
+      "step": 26895
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.292146418174342,
+      "learning_rate": 5.778464977489583e-06,
+      "loss": 0.8729,
+      "step": 26900
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.991862160933817,
+      "learning_rate": 5.776845910631413e-06,
+      "loss": 0.8523,
+      "step": 26905
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.537098363159829,
+      "learning_rate": 5.775226760297723e-06,
+      "loss": 0.8714,
+      "step": 26910
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 72.04608074423221,
+      "learning_rate": 5.773607526662495e-06,
+      "loss": 0.889,
+      "step": 26915
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 29.135941371981282,
+      "learning_rate": 5.771988209899726e-06,
+      "loss": 0.8272,
+      "step": 26920
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.55994572101125,
+      "learning_rate": 5.770368810183416e-06,
+      "loss": 0.8662,
+      "step": 26925
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 24.86800422217252,
+      "learning_rate": 5.768749327687576e-06,
+      "loss": 0.8341,
+      "step": 26930
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.44367611703542,
+      "learning_rate": 5.7671297625862285e-06,
+      "loss": 0.8864,
+      "step": 26935
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 27.140857210256808,
+      "learning_rate": 5.765510115053403e-06,
+      "loss": 0.867,
+      "step": 26940
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 25.503062108739485,
+      "learning_rate": 5.763890385263137e-06,
+      "loss": 0.8602,
+      "step": 26945
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.262505144514137,
+      "learning_rate": 5.7622705733894765e-06,
+      "loss": 0.8382,
+      "step": 26950
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.544102102282014,
+      "learning_rate": 5.760650679606481e-06,
+      "loss": 0.8326,
+      "step": 26955
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.16640864225168,
+      "learning_rate": 5.759030704088212e-06,
+      "loss": 0.8495,
+      "step": 26960
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 31.16127289154043,
+      "learning_rate": 5.757410647008743e-06,
+      "loss": 0.8607,
+      "step": 26965
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 5.524574888705631,
+      "learning_rate": 5.7557905085421586e-06,
+      "loss": 0.8961,
+      "step": 26970
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 22.84678258694053,
+      "learning_rate": 5.7541702888625494e-06,
+      "loss": 0.8712,
+      "step": 26975
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 27.847945253156105,
+      "learning_rate": 5.752549988144013e-06,
+      "loss": 0.8503,
+      "step": 26980
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.34699505928167,
+      "learning_rate": 5.750929606560659e-06,
+      "loss": 0.8432,
+      "step": 26985
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.853432723443728,
+      "learning_rate": 5.749309144286603e-06,
+      "loss": 0.8297,
+      "step": 26990
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.400375145732713,
+      "learning_rate": 5.747688601495974e-06,
+      "loss": 0.8697,
+      "step": 26995
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 5.871987907957095,
+      "learning_rate": 5.746067978362906e-06,
+      "loss": 0.8809,
+      "step": 27000
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.997440121192785,
+      "learning_rate": 5.744447275061541e-06,
+      "loss": 0.8242,
+      "step": 27005
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 5.611445145556261,
+      "learning_rate": 5.74282649176603e-06,
+      "loss": 0.8658,
+      "step": 27010
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 23.036934755302443,
+      "learning_rate": 5.741205628650534e-06,
+      "loss": 0.8667,
+      "step": 27015
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.898614904548687,
+      "learning_rate": 5.739584685889222e-06,
+      "loss": 0.863,
+      "step": 27020
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 5.778958091052699,
+      "learning_rate": 5.737963663656272e-06,
+      "loss": 0.8631,
+      "step": 27025
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 5.452358480653361,
+      "learning_rate": 5.736342562125869e-06,
+      "loss": 0.8225,
+      "step": 27030
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.5718974147484,
+      "learning_rate": 5.734721381472208e-06,
+      "loss": 0.8896,
+      "step": 27035
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.39509962960557,
+      "learning_rate": 5.733100121869491e-06,
+      "loss": 0.8663,
+      "step": 27040
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.338306012900134,
+      "learning_rate": 5.731478783491933e-06,
+      "loss": 0.8702,
+      "step": 27045
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.794404135191542,
+      "learning_rate": 5.729857366513749e-06,
+      "loss": 0.8503,
+      "step": 27050
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.670373140811577,
+      "learning_rate": 5.728235871109171e-06,
+      "loss": 0.9143,
+      "step": 27055
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 32.70669935782094,
+      "learning_rate": 5.726614297452435e-06,
+      "loss": 0.8292,
+      "step": 27060
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.953620508293519,
+      "learning_rate": 5.7249926457177864e-06,
+      "loss": 0.8404,
+      "step": 27065
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.67039148261923,
+      "learning_rate": 5.723370916079477e-06,
+      "loss": 0.8678,
+      "step": 27070
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.365671781112125,
+      "learning_rate": 5.721749108711771e-06,
+      "loss": 0.8381,
+      "step": 27075
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 23.26522805711699,
+      "learning_rate": 5.720127223788937e-06,
+      "loss": 0.8151,
+      "step": 27080
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.859929883109832,
+      "learning_rate": 5.718505261485254e-06,
+      "loss": 0.8431,
+      "step": 27085
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.799837914430625,
+      "learning_rate": 5.716883221975012e-06,
+      "loss": 0.8729,
+      "step": 27090
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.821493804212123,
+      "learning_rate": 5.7152611054325035e-06,
+      "loss": 0.8486,
+      "step": 27095
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.809258458222593,
+      "learning_rate": 5.713638912032031e-06,
+      "loss": 0.8258,
+      "step": 27100
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.7600763162085995,
+      "learning_rate": 5.71201664194791e-06,
+      "loss": 0.8208,
+      "step": 27105
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.140165397678468,
+      "learning_rate": 5.710394295354457e-06,
+      "loss": 0.7833,
+      "step": 27110
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.888013156962577,
+      "learning_rate": 5.708771872426001e-06,
+      "loss": 0.8541,
+      "step": 27115
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.210882855080766,
+      "learning_rate": 5.707149373336882e-06,
+      "loss": 0.8656,
+      "step": 27120
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.53341906596725,
+      "learning_rate": 5.705526798261439e-06,
+      "loss": 0.8182,
+      "step": 27125
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.030618587221534,
+      "learning_rate": 5.703904147374029e-06,
+      "loss": 0.8436,
+      "step": 27130
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.83816126088371,
+      "learning_rate": 5.702281420849012e-06,
+      "loss": 0.8318,
+      "step": 27135
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.291796271100154,
+      "learning_rate": 5.700658618860756e-06,
+      "loss": 0.804,
+      "step": 27140
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.409684993814613,
+      "learning_rate": 5.69903574158364e-06,
+      "loss": 0.8522,
+      "step": 27145
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.419367765427134,
+      "learning_rate": 5.697412789192049e-06,
+      "loss": 0.8312,
+      "step": 27150
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.929834417340285,
+      "learning_rate": 5.695789761860376e-06,
+      "loss": 0.8549,
+      "step": 27155
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.511636447472515,
+      "learning_rate": 5.694166659763021e-06,
+      "loss": 0.8601,
+      "step": 27160
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.3262169619225785,
+      "learning_rate": 5.692543483074395e-06,
+      "loss": 0.8706,
+      "step": 27165
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.21830397641139,
+      "learning_rate": 5.690920231968916e-06,
+      "loss": 0.8675,
+      "step": 27170
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.60583307079588,
+      "learning_rate": 5.689296906621009e-06,
+      "loss": 0.8583,
+      "step": 27175
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.1778738588941176,
+      "learning_rate": 5.687673507205106e-06,
+      "loss": 0.8124,
+      "step": 27180
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.144574801402935,
+      "learning_rate": 5.68605003389565e-06,
+      "loss": 0.8301,
+      "step": 27185
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.262813440679303,
+      "learning_rate": 5.684426486867091e-06,
+      "loss": 0.8332,
+      "step": 27190
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.062909804526193,
+      "learning_rate": 5.682802866293885e-06,
+      "loss": 0.8136,
+      "step": 27195
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.476692240168745,
+      "learning_rate": 5.681179172350498e-06,
+      "loss": 0.805,
+      "step": 27200
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.778244180909876,
+      "learning_rate": 5.679555405211403e-06,
+      "loss": 0.8624,
+      "step": 27205
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.187187229792764,
+      "learning_rate": 5.67793156505108e-06,
+      "loss": 0.8377,
+      "step": 27210
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.7618080121242485,
+      "learning_rate": 5.676307652044017e-06,
+      "loss": 0.8537,
+      "step": 27215
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.379226001157689,
+      "learning_rate": 5.674683666364713e-06,
+      "loss": 0.84,
+      "step": 27220
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.9890669802333285,
+      "learning_rate": 5.673059608187672e-06,
+      "loss": 0.8068,
+      "step": 27225
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.043482550445178,
+      "learning_rate": 5.671435477687406e-06,
+      "loss": 0.8391,
+      "step": 27230
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.341320791877013,
+      "learning_rate": 5.669811275038434e-06,
+      "loss": 0.8516,
+      "step": 27235
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.902154855316724,
+      "learning_rate": 5.668187000415284e-06,
+      "loss": 0.8329,
+      "step": 27240
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.595010163035232,
+      "learning_rate": 5.666562653992492e-06,
+      "loss": 0.8757,
+      "step": 27245
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 18.225129141628653,
+      "learning_rate": 5.6649382359446015e-06,
+      "loss": 0.8014,
+      "step": 27250
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.458756381956273,
+      "learning_rate": 5.663313746446163e-06,
+      "loss": 0.8402,
+      "step": 27255
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.787930171534921,
+      "learning_rate": 5.661689185671735e-06,
+      "loss": 0.7883,
+      "step": 27260
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.914455180278969,
+      "learning_rate": 5.660064553795883e-06,
+      "loss": 0.904,
+      "step": 27265
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.551594467432754,
+      "learning_rate": 5.658439850993183e-06,
+      "loss": 0.8098,
+      "step": 27270
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.065101015780171,
+      "learning_rate": 5.6568150774382135e-06,
+      "loss": 0.8523,
+      "step": 27275
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.162843789091219,
+      "learning_rate": 5.655190233305566e-06,
+      "loss": 0.8694,
+      "step": 27280
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 19.2239011034022,
+      "learning_rate": 5.653565318769836e-06,
+      "loss": 0.8375,
+      "step": 27285
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.116686191897343,
+      "learning_rate": 5.65194033400563e-06,
+      "loss": 0.794,
+      "step": 27290
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 21.334769537034536,
+      "learning_rate": 5.650315279187556e-06,
+      "loss": 0.8177,
+      "step": 27295
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.080425864614117,
+      "learning_rate": 5.648690154490235e-06,
+      "loss": 0.8346,
+      "step": 27300
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.872567644534107,
+      "learning_rate": 5.647064960088294e-06,
+      "loss": 0.8222,
+      "step": 27305
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.4852150094843,
+      "learning_rate": 5.645439696156368e-06,
+      "loss": 0.8425,
+      "step": 27310
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.094028907139457,
+      "learning_rate": 5.643814362869098e-06,
+      "loss": 0.7855,
+      "step": 27315
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.267161992812545,
+      "learning_rate": 5.642188960401133e-06,
+      "loss": 0.7879,
+      "step": 27320
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.6381639561117,
+      "learning_rate": 5.640563488927131e-06,
+      "loss": 0.848,
+      "step": 27325
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.090066368644928,
+      "learning_rate": 5.638937948621755e-06,
+      "loss": 0.8077,
+      "step": 27330
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 18.086032406154715,
+      "learning_rate": 5.637312339659675e-06,
+      "loss": 0.861,
+      "step": 27335
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.460124892521248,
+      "learning_rate": 5.635686662215572e-06,
+      "loss": 0.8224,
+      "step": 27340
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.710627853746999,
+      "learning_rate": 5.6340609164641335e-06,
+      "loss": 0.8473,
+      "step": 27345
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.55769853529202,
+      "learning_rate": 5.632435102580049e-06,
+      "loss": 0.8736,
+      "step": 27350
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.201174022453873,
+      "learning_rate": 5.630809220738022e-06,
+      "loss": 0.8547,
+      "step": 27355
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.843169835655888,
+      "learning_rate": 5.629183271112759e-06,
+      "loss": 0.8214,
+      "step": 27360
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.910826737135694,
+      "learning_rate": 5.627557253878979e-06,
+      "loss": 0.848,
+      "step": 27365
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 18.059057445888836,
+      "learning_rate": 5.625931169211401e-06,
+      "loss": 0.7987,
+      "step": 27370
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.074060846192095,
+      "learning_rate": 5.624305017284759e-06,
+      "loss": 0.8243,
+      "step": 27375
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.729710748597427,
+      "learning_rate": 5.6226787982737855e-06,
+      "loss": 0.8602,
+      "step": 27380
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.756800832903632,
+      "learning_rate": 5.621052512353227e-06,
+      "loss": 0.8395,
+      "step": 27385
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.351292362153368,
+      "learning_rate": 5.619426159697835e-06,
+      "loss": 0.8523,
+      "step": 27390
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.789759867363111,
+      "learning_rate": 5.617799740482369e-06,
+      "loss": 0.8895,
+      "step": 27395
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.234766910628488,
+      "learning_rate": 5.616173254881594e-06,
+      "loss": 0.8221,
+      "step": 27400
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.358802567627612,
+      "learning_rate": 5.614546703070284e-06,
+      "loss": 0.8594,
+      "step": 27405
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.78267701844748,
+      "learning_rate": 5.612920085223218e-06,
+      "loss": 0.8545,
+      "step": 27410
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.088368755960512,
+      "learning_rate": 5.611293401515183e-06,
+      "loss": 0.8499,
+      "step": 27415
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.73630214387602,
+      "learning_rate": 5.609666652120975e-06,
+      "loss": 0.8254,
+      "step": 27420
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.017294780631743,
+      "learning_rate": 5.608039837215395e-06,
+      "loss": 0.8532,
+      "step": 27425
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.662313902311142,
+      "learning_rate": 5.606412956973252e-06,
+      "loss": 0.7892,
+      "step": 27430
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.687695641695749,
+      "learning_rate": 5.604786011569359e-06,
+      "loss": 0.8551,
+      "step": 27435
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 20.408377014543518,
+      "learning_rate": 5.60315900117854e-06,
+      "loss": 0.8325,
+      "step": 27440
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.739644669376226,
+      "learning_rate": 5.601531925975624e-06,
+      "loss": 0.836,
+      "step": 27445
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.245318583892333,
+      "learning_rate": 5.599904786135447e-06,
+      "loss": 0.8567,
+      "step": 27450
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.848560725241006,
+      "learning_rate": 5.598277581832855e-06,
+      "loss": 0.8272,
+      "step": 27455
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.181595541273278,
+      "learning_rate": 5.596650313242697e-06,
+      "loss": 0.8291,
+      "step": 27460
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 22.732225652466553,
+      "learning_rate": 5.595022980539828e-06,
+      "loss": 0.8439,
+      "step": 27465
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.749331682316672,
+      "learning_rate": 5.593395583899114e-06,
+      "loss": 0.8074,
+      "step": 27470
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.6155307899934765,
+      "learning_rate": 5.591768123495425e-06,
+      "loss": 0.8162,
+      "step": 27475
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.38078240445812,
+      "learning_rate": 5.59014059950364e-06,
+      "loss": 0.8112,
+      "step": 27480
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.726742359434211,
+      "learning_rate": 5.588513012098644e-06,
+      "loss": 0.854,
+      "step": 27485
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.522912054408138,
+      "learning_rate": 5.586885361455326e-06,
+      "loss": 0.8676,
+      "step": 27490
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 42.93731286731151,
+      "learning_rate": 5.585257647748587e-06,
+      "loss": 0.8993,
+      "step": 27495
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.749483489627861,
+      "learning_rate": 5.583629871153331e-06,
+      "loss": 0.8451,
+      "step": 27500
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 20.590363748463695,
+      "learning_rate": 5.582002031844469e-06,
+      "loss": 0.8579,
+      "step": 27505
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.337436630851883,
+      "learning_rate": 5.580374129996919e-06,
+      "loss": 0.8373,
+      "step": 27510
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 31.544819676786396,
+      "learning_rate": 5.578746165785609e-06,
+      "loss": 0.8752,
+      "step": 27515
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.415996857447883,
+      "learning_rate": 5.5771181393854705e-06,
+      "loss": 0.8477,
+      "step": 27520
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.86394109230957,
+      "learning_rate": 5.5754900509714396e-06,
+      "loss": 0.8092,
+      "step": 27525
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.3592762405029,
+      "learning_rate": 5.573861900718461e-06,
+      "loss": 0.8656,
+      "step": 27530
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.46238396955788,
+      "learning_rate": 5.5722336888014905e-06,
+      "loss": 0.82,
+      "step": 27535
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.127625863744789,
+      "learning_rate": 5.570605415395485e-06,
+      "loss": 0.8092,
+      "step": 27540
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.667990204026085,
+      "learning_rate": 5.568977080675407e-06,
+      "loss": 0.8503,
+      "step": 27545
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.960084324015313,
+      "learning_rate": 5.567348684816233e-06,
+      "loss": 0.8626,
+      "step": 27550
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.8856049884045785,
+      "learning_rate": 5.5657202279929375e-06,
+      "loss": 0.8334,
+      "step": 27555
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.562954285835208,
+      "learning_rate": 5.564091710380507e-06,
+      "loss": 0.8936,
+      "step": 27560
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.1887514664284975,
+      "learning_rate": 5.562463132153933e-06,
+      "loss": 0.8367,
+      "step": 27565
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.910120053504759,
+      "learning_rate": 5.560834493488213e-06,
+      "loss": 0.8185,
+      "step": 27570
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.007103534982401,
+      "learning_rate": 5.559205794558354e-06,
+      "loss": 0.8084,
+      "step": 27575
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.508089863839682,
+      "learning_rate": 5.557577035539361e-06,
+      "loss": 0.8396,
+      "step": 27580
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.201086783663813,
+      "learning_rate": 5.555948216606255e-06,
+      "loss": 0.877,
+      "step": 27585
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.82826226084147,
+      "learning_rate": 5.55431933793406e-06,
+      "loss": 0.8088,
+      "step": 27590
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.4018673360403024,
+      "learning_rate": 5.552690399697807e-06,
+      "loss": 0.8179,
+      "step": 27595
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.896817301291493,
+      "learning_rate": 5.551061402072529e-06,
+      "loss": 0.8889,
+      "step": 27600
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.8512092962836375,
+      "learning_rate": 5.549432345233274e-06,
+      "loss": 0.886,
+      "step": 27605
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.78467056315316,
+      "learning_rate": 5.5478032293550875e-06,
+      "loss": 0.8348,
+      "step": 27610
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.038521798819803,
+      "learning_rate": 5.546174054613027e-06,
+      "loss": 0.8004,
+      "step": 27615
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.543501419243198,
+      "learning_rate": 5.544544821182153e-06,
+      "loss": 0.802,
+      "step": 27620
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.576978082024347,
+      "learning_rate": 5.542915529237535e-06,
+      "loss": 0.8237,
+      "step": 27625
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.346995386496742,
+      "learning_rate": 5.541286178954247e-06,
+      "loss": 0.8602,
+      "step": 27630
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.233414275924641,
+      "learning_rate": 5.53965677050737e-06,
+      "loss": 0.8565,
+      "step": 27635
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 24.543995130295432,
+      "learning_rate": 5.538027304071992e-06,
+      "loss": 0.837,
+      "step": 27640
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.357417946046084,
+      "learning_rate": 5.536397779823205e-06,
+      "loss": 0.7966,
+      "step": 27645
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.137854842287357,
+      "learning_rate": 5.53476819793611e-06,
+      "loss": 0.825,
+      "step": 27650
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.30063959889956,
+      "learning_rate": 5.533138558585811e-06,
+      "loss": 0.8646,
+      "step": 27655
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.692924139491135,
+      "learning_rate": 5.5315088619474235e-06,
+      "loss": 0.8498,
+      "step": 27660
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.711643769294201,
+      "learning_rate": 5.529879108196061e-06,
+      "loss": 0.8135,
+      "step": 27665
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.025663645641752,
+      "learning_rate": 5.5282492975068495e-06,
+      "loss": 0.8151,
+      "step": 27670
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.062118954198813,
+      "learning_rate": 5.52661943005492e-06,
+      "loss": 0.8185,
+      "step": 27675
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.757415316997683,
+      "learning_rate": 5.524989506015408e-06,
+      "loss": 0.8247,
+      "step": 27680
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.26047128596025,
+      "learning_rate": 5.523359525563457e-06,
+      "loss": 0.8758,
+      "step": 27685
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.441446298587094,
+      "learning_rate": 5.521729488874217e-06,
+      "loss": 0.9443,
+      "step": 27690
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.590963719737849,
+      "learning_rate": 5.520099396122839e-06,
+      "loss": 0.8296,
+      "step": 27695
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.370192771088231,
+      "learning_rate": 5.518469247484486e-06,
+      "loss": 0.867,
+      "step": 27700
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.2407369045623176,
+      "learning_rate": 5.516839043134325e-06,
+      "loss": 0.8568,
+      "step": 27705
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.088367550059989,
+      "learning_rate": 5.515208783247525e-06,
+      "loss": 0.8679,
+      "step": 27710
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.493212415317777,
+      "learning_rate": 5.513578467999271e-06,
+      "loss": 0.8355,
+      "step": 27715
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.280554850900711,
+      "learning_rate": 5.511948097564741e-06,
+      "loss": 0.8707,
+      "step": 27720
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.138374637209298,
+      "learning_rate": 5.510317672119129e-06,
+      "loss": 0.8567,
+      "step": 27725
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.892377205049216,
+      "learning_rate": 5.508687191837633e-06,
+      "loss": 0.8684,
+      "step": 27730
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.286114569552311,
+      "learning_rate": 5.507056656895452e-06,
+      "loss": 0.7954,
+      "step": 27735
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.108608327951508,
+      "learning_rate": 5.5054260674677965e-06,
+      "loss": 0.828,
+      "step": 27740
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.019463904192895,
+      "learning_rate": 5.503795423729879e-06,
+      "loss": 0.8467,
+      "step": 27745
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.911627974968405,
+      "learning_rate": 5.502164725856919e-06,
+      "loss": 0.8575,
+      "step": 27750
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.949764024135887,
+      "learning_rate": 5.500533974024143e-06,
+      "loss": 0.8088,
+      "step": 27755
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.710812954342271,
+      "learning_rate": 5.498903168406783e-06,
+      "loss": 0.8165,
+      "step": 27760
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.615274832979397,
+      "learning_rate": 5.497272309180075e-06,
+      "loss": 0.7984,
+      "step": 27765
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.690133110156747,
+      "learning_rate": 5.4956413965192636e-06,
+      "loss": 0.862,
+      "step": 27770
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.723900262271425,
+      "learning_rate": 5.494010430599595e-06,
+      "loss": 0.8347,
+      "step": 27775
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.791899694774015,
+      "learning_rate": 5.4923794115963265e-06,
+      "loss": 0.8129,
+      "step": 27780
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.223685329637303,
+      "learning_rate": 5.4907483396847185e-06,
+      "loss": 0.8601,
+      "step": 27785
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.8504187217528445,
+      "learning_rate": 5.489117215040034e-06,
+      "loss": 0.8321,
+      "step": 27790
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.702570409384052,
+      "learning_rate": 5.487486037837546e-06,
+      "loss": 0.8708,
+      "step": 27795
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.058235735878204,
+      "learning_rate": 5.485854808252533e-06,
+      "loss": 0.8327,
+      "step": 27800
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.450428454937076,
+      "learning_rate": 5.484223526460276e-06,
+      "loss": 0.8538,
+      "step": 27805
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.907230420761335,
+      "learning_rate": 5.4825921926360644e-06,
+      "loss": 0.852,
+      "step": 27810
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.108923660685045,
+      "learning_rate": 5.48096080695519e-06,
+      "loss": 0.8309,
+      "step": 27815
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.867693572535084,
+      "learning_rate": 5.4793293695929564e-06,
+      "loss": 0.8385,
+      "step": 27820
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.946951013702713,
+      "learning_rate": 5.477697880724665e-06,
+      "loss": 0.7753,
+      "step": 27825
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.816107661414321,
+      "learning_rate": 5.47606634052563e-06,
+      "loss": 0.8015,
+      "step": 27830
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.016088357097027,
+      "learning_rate": 5.474434749171164e-06,
+      "loss": 0.8528,
+      "step": 27835
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 21.306577446607616,
+      "learning_rate": 5.472803106836592e-06,
+      "loss": 0.8113,
+      "step": 27840
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.674021200535829,
+      "learning_rate": 5.471171413697238e-06,
+      "loss": 0.8126,
+      "step": 27845
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.430222254658212,
+      "learning_rate": 5.469539669928437e-06,
+      "loss": 0.8224,
+      "step": 27850
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.143416577397818,
+      "learning_rate": 5.467907875705526e-06,
+      "loss": 0.8392,
+      "step": 27855
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.417623630354982,
+      "learning_rate": 5.466276031203849e-06,
+      "loss": 0.8414,
+      "step": 27860
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.67882115343529,
+      "learning_rate": 5.464644136598753e-06,
+      "loss": 0.8471,
+      "step": 27865
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 26.87099653384102,
+      "learning_rate": 5.463012192065595e-06,
+      "loss": 0.8393,
+      "step": 27870
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.282517645050707,
+      "learning_rate": 5.4613801977797345e-06,
+      "loss": 0.7999,
+      "step": 27875
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 32.54110515129484,
+      "learning_rate": 5.459748153916535e-06,
+      "loss": 0.908,
+      "step": 27880
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.740210452511342,
+      "learning_rate": 5.458116060651366e-06,
+      "loss": 0.8866,
+      "step": 27885
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.81184829771982,
+      "learning_rate": 5.456483918159606e-06,
+      "loss": 0.8226,
+      "step": 27890
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.378253161486748,
+      "learning_rate": 5.4548517266166346e-06,
+      "loss": 0.8511,
+      "step": 27895
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.630480359676842,
+      "learning_rate": 5.453219486197837e-06,
+      "loss": 0.8325,
+      "step": 27900
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.870311708880287,
+      "learning_rate": 5.451587197078605e-06,
+      "loss": 0.8656,
+      "step": 27905
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.967312223873702,
+      "learning_rate": 5.449954859434336e-06,
+      "loss": 0.8518,
+      "step": 27910
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.148821654746222,
+      "learning_rate": 5.448322473440431e-06,
+      "loss": 0.8134,
+      "step": 27915
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.03014567338059,
+      "learning_rate": 5.446690039272299e-06,
+      "loss": 0.8618,
+      "step": 27920
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.297395813566277,
+      "learning_rate": 5.4450575571053496e-06,
+      "loss": 0.8319,
+      "step": 27925
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 34.17516218317393,
+      "learning_rate": 5.443425027115002e-06,
+      "loss": 0.7979,
+      "step": 27930
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 21.500496252608457,
+      "learning_rate": 5.441792449476678e-06,
+      "loss": 0.8697,
+      "step": 27935
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.519097324982825,
+      "learning_rate": 5.440159824365806e-06,
+      "loss": 0.8588,
+      "step": 27940
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.312360802795807,
+      "learning_rate": 5.438527151957819e-06,
+      "loss": 0.8273,
+      "step": 27945
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.24726881885029,
+      "learning_rate": 5.436894432428153e-06,
+      "loss": 0.841,
+      "step": 27950
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.7477453811371015,
+      "learning_rate": 5.435261665952252e-06,
+      "loss": 0.8138,
+      "step": 27955
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.59863456757837,
+      "learning_rate": 5.433628852705565e-06,
+      "loss": 0.8524,
+      "step": 27960
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.281543679950202,
+      "learning_rate": 5.431995992863542e-06,
+      "loss": 0.8141,
+      "step": 27965
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.853037424161569,
+      "learning_rate": 5.430363086601646e-06,
+      "loss": 0.8229,
+      "step": 27970
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.118403261749481,
+      "learning_rate": 5.428730134095336e-06,
+      "loss": 0.8422,
+      "step": 27975
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.127501702032474,
+      "learning_rate": 5.4270971355200795e-06,
+      "loss": 0.8262,
+      "step": 27980
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.403048819659286,
+      "learning_rate": 5.4254640910513525e-06,
+      "loss": 0.842,
+      "step": 27985
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.816020851418337,
+      "learning_rate": 5.42383100086463e-06,
+      "loss": 0.8253,
+      "step": 27990
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.133337369541252,
+      "learning_rate": 5.422197865135397e-06,
+      "loss": 0.9429,
+      "step": 27995
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.782257755899195,
+      "learning_rate": 5.420564684039139e-06,
+      "loss": 0.8021,
+      "step": 28000
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.532147995135568,
+      "learning_rate": 5.4189314577513495e-06,
+      "loss": 0.8192,
+      "step": 28005
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.78199884547465,
+      "learning_rate": 5.417298186447527e-06,
+      "loss": 0.7816,
+      "step": 28010
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.928969019553306,
+      "learning_rate": 5.415664870303172e-06,
+      "loss": 0.845,
+      "step": 28015
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.633092908974432,
+      "learning_rate": 5.414031509493792e-06,
+      "loss": 0.8306,
+      "step": 28020
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.095078882652489,
+      "learning_rate": 5.4123981041949005e-06,
+      "loss": 0.827,
+      "step": 28025
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.938725798644738,
+      "learning_rate": 5.410764654582012e-06,
+      "loss": 0.8444,
+      "step": 28030
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.951926269678413,
+      "learning_rate": 5.409131160830649e-06,
+      "loss": 0.781,
+      "step": 28035
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.150375771619542,
+      "learning_rate": 5.407497623116336e-06,
+      "loss": 0.8226,
+      "step": 28040
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.32529151836401,
+      "learning_rate": 5.405864041614604e-06,
+      "loss": 0.8172,
+      "step": 28045
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.692926834476918,
+      "learning_rate": 5.40423041650099e-06,
+      "loss": 0.839,
+      "step": 28050
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.009573418508833,
+      "learning_rate": 5.402596747951036e-06,
+      "loss": 0.8159,
+      "step": 28055
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.999411297179064,
+      "learning_rate": 5.400963036140283e-06,
+      "loss": 0.8206,
+      "step": 28060
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.888704546367954,
+      "learning_rate": 5.399329281244283e-06,
+      "loss": 0.8445,
+      "step": 28065
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.24314780906232,
+      "learning_rate": 5.397695483438588e-06,
+      "loss": 0.8414,
+      "step": 28070
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.532741570267251,
+      "learning_rate": 5.39606164289876e-06,
+      "loss": 0.8034,
+      "step": 28075
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.245586825990241,
+      "learning_rate": 5.394427759800359e-06,
+      "loss": 0.849,
+      "step": 28080
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.087065643864672,
+      "learning_rate": 5.392793834318954e-06,
+      "loss": 0.8166,
+      "step": 28085
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.660077410121367,
+      "learning_rate": 5.3911598666301206e-06,
+      "loss": 0.8042,
+      "step": 28090
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.129664756007115,
+      "learning_rate": 5.38952585690943e-06,
+      "loss": 0.7934,
+      "step": 28095
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.758722826603849,
+      "learning_rate": 5.387891805332469e-06,
+      "loss": 0.8012,
+      "step": 28100
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.35020321715153,
+      "learning_rate": 5.3862577120748204e-06,
+      "loss": 0.8362,
+      "step": 28105
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.948839984601692,
+      "learning_rate": 5.384623577312077e-06,
+      "loss": 0.7891,
+      "step": 28110
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.983929128502247,
+      "learning_rate": 5.382989401219834e-06,
+      "loss": 0.8467,
+      "step": 28115
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.532246618009143,
+      "learning_rate": 5.381355183973688e-06,
+      "loss": 0.8319,
+      "step": 28120
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.237831590175402,
+      "learning_rate": 5.379720925749243e-06,
+      "loss": 0.8106,
+      "step": 28125
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.704105732080835,
+      "learning_rate": 5.378086626722111e-06,
+      "loss": 0.8412,
+      "step": 28130
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.420320694062168,
+      "learning_rate": 5.376452287067901e-06,
+      "loss": 0.7876,
+      "step": 28135
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.405119387129808,
+      "learning_rate": 5.37481790696223e-06,
+      "loss": 0.8545,
+      "step": 28140
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.403413268390825,
+      "learning_rate": 5.373183486580724e-06,
+      "loss": 0.8172,
+      "step": 28145
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.302438246491855,
+      "learning_rate": 5.371549026099004e-06,
+      "loss": 0.8505,
+      "step": 28150
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.282060502041304,
+      "learning_rate": 5.369914525692701e-06,
+      "loss": 0.8316,
+      "step": 28155
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.323682826733916,
+      "learning_rate": 5.3682799855374505e-06,
+      "loss": 0.8347,
+      "step": 28160
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.227848468789792,
+      "learning_rate": 5.366645405808889e-06,
+      "loss": 0.826,
+      "step": 28165
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.618638011999,
+      "learning_rate": 5.3650107866826605e-06,
+      "loss": 0.8574,
+      "step": 28170
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.267553653194656,
+      "learning_rate": 5.3633761283344145e-06,
+      "loss": 0.8136,
+      "step": 28175
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.54685896451495,
+      "learning_rate": 5.361741430939796e-06,
+      "loss": 0.8709,
+      "step": 28180
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.247908029266732,
+      "learning_rate": 5.360106694674467e-06,
+      "loss": 0.8234,
+      "step": 28185
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.683044212708232,
+      "learning_rate": 5.358471919714083e-06,
+      "loss": 0.8416,
+      "step": 28190
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.054085673351551,
+      "learning_rate": 5.3568371062343095e-06,
+      "loss": 0.8172,
+      "step": 28195
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.030692635069453,
+      "learning_rate": 5.355202254410815e-06,
+      "loss": 0.8039,
+      "step": 28200
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 24.589980772210108,
+      "learning_rate": 5.35356736441927e-06,
+      "loss": 0.8337,
+      "step": 28205
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.812933030426917,
+      "learning_rate": 5.351932436435352e-06,
+      "loss": 0.8023,
+      "step": 28210
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.462925966717622,
+      "learning_rate": 5.350297470634739e-06,
+      "loss": 0.8422,
+      "step": 28215
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.128437956796128,
+      "learning_rate": 5.348662467193117e-06,
+      "loss": 0.8196,
+      "step": 28220
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.437050767796805,
+      "learning_rate": 5.347027426286175e-06,
+      "loss": 0.7703,
+      "step": 28225
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.495761817884926,
+      "learning_rate": 5.3453923480896045e-06,
+      "loss": 0.8727,
+      "step": 28230
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.910424106159469,
+      "learning_rate": 5.343757232779101e-06,
+      "loss": 0.8417,
+      "step": 28235
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.15136226937716,
+      "learning_rate": 5.342122080530366e-06,
+      "loss": 0.7904,
+      "step": 28240
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.786258876532818,
+      "learning_rate": 5.340486891519104e-06,
+      "loss": 0.8356,
+      "step": 28245
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.425838374468625,
+      "learning_rate": 5.338851665921023e-06,
+      "loss": 0.8012,
+      "step": 28250
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.029136058538585,
+      "learning_rate": 5.337216403911835e-06,
+      "loss": 0.8284,
+      "step": 28255
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.539465752975717,
+      "learning_rate": 5.335581105667257e-06,
+      "loss": 0.8912,
+      "step": 28260
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.122478646096978,
+      "learning_rate": 5.333945771363007e-06,
+      "loss": 0.8147,
+      "step": 28265
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.8301488665732215,
+      "learning_rate": 5.3323104011748095e-06,
+      "loss": 0.82,
+      "step": 28270
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.454402230807887,
+      "learning_rate": 5.330674995278392e-06,
+      "loss": 0.8273,
+      "step": 28275
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.442497445804376,
+      "learning_rate": 5.3290395538494884e-06,
+      "loss": 0.774,
+      "step": 28280
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.509011519846112,
+      "learning_rate": 5.327404077063833e-06,
+      "loss": 0.8503,
+      "step": 28285
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.119290469516933,
+      "learning_rate": 5.325768565097165e-06,
+      "loss": 0.8425,
+      "step": 28290
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.794450221973544,
+      "learning_rate": 5.324133018125225e-06,
+      "loss": 0.8286,
+      "step": 28295
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.116707377438561,
+      "learning_rate": 5.322497436323762e-06,
+      "loss": 0.8461,
+      "step": 28300
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.9167396023247125,
+      "learning_rate": 5.320861819868528e-06,
+      "loss": 0.8355,
+      "step": 28305
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.810364930544731,
+      "learning_rate": 5.319226168935274e-06,
+      "loss": 0.8558,
+      "step": 28310
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 18.762813564943713,
+      "learning_rate": 5.317590483699759e-06,
+      "loss": 0.8194,
+      "step": 28315
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.088033894341939,
+      "learning_rate": 5.315954764337745e-06,
+      "loss": 0.8797,
+      "step": 28320
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.968115299170364,
+      "learning_rate": 5.314319011024997e-06,
+      "loss": 0.799,
+      "step": 28325
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 20.125631095807982,
+      "learning_rate": 5.312683223937286e-06,
+      "loss": 0.8644,
+      "step": 28330
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.920482514170674,
+      "learning_rate": 5.311047403250381e-06,
+      "loss": 0.8641,
+      "step": 28335
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.541664732561482,
+      "learning_rate": 5.30941154914006e-06,
+      "loss": 0.8346,
+      "step": 28340
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.105027711919444,
+      "learning_rate": 5.307775661782103e-06,
+      "loss": 0.8226,
+      "step": 28345
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.637906092546576,
+      "learning_rate": 5.306139741352292e-06,
+      "loss": 0.8212,
+      "step": 28350
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.838597684914096,
+      "learning_rate": 5.304503788026415e-06,
+      "loss": 0.8369,
+      "step": 28355
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.083702528464722,
+      "learning_rate": 5.302867801980263e-06,
+      "loss": 0.8499,
+      "step": 28360
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.065812029607843,
+      "learning_rate": 5.3012317833896275e-06,
+      "loss": 0.8237,
+      "step": 28365
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.070149019654948,
+      "learning_rate": 5.299595732430307e-06,
+      "loss": 0.8599,
+      "step": 28370
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.644554902747811,
+      "learning_rate": 5.297959649278105e-06,
+      "loss": 0.8012,
+      "step": 28375
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.663294659461215,
+      "learning_rate": 5.296323534108823e-06,
+      "loss": 0.8384,
+      "step": 28380
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.04330433603099,
+      "learning_rate": 5.294687387098268e-06,
+      "loss": 0.8057,
+      "step": 28385
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.311084241460337,
+      "learning_rate": 5.293051208422254e-06,
+      "loss": 0.8146,
+      "step": 28390
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.021395805213237,
+      "learning_rate": 5.2914149982565945e-06,
+      "loss": 0.8392,
+      "step": 28395
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.8248191892494425,
+      "learning_rate": 5.289778756777107e-06,
+      "loss": 0.8171,
+      "step": 28400
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.4583143913361685,
+      "learning_rate": 5.288142484159614e-06,
+      "loss": 0.8439,
+      "step": 28405
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.20549804743247,
+      "learning_rate": 5.2865061805799364e-06,
+      "loss": 0.7952,
+      "step": 28410
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.586153347066439,
+      "learning_rate": 5.284869846213908e-06,
+      "loss": 0.7838,
+      "step": 28415
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.322082353890254,
+      "learning_rate": 5.283233481237357e-06,
+      "loss": 0.8923,
+      "step": 28420
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.567110742476824,
+      "learning_rate": 5.281597085826118e-06,
+      "loss": 0.773,
+      "step": 28425
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.71610228404068,
+      "learning_rate": 5.2799606601560305e-06,
+      "loss": 0.8532,
+      "step": 28430
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.171136241192343,
+      "learning_rate": 5.278324204402934e-06,
+      "loss": 0.7937,
+      "step": 28435
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.155123385249471,
+      "learning_rate": 5.276687718742672e-06,
+      "loss": 0.8554,
+      "step": 28440
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.064522669111217,
+      "learning_rate": 5.275051203351095e-06,
+      "loss": 0.8011,
+      "step": 28445
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.544786271309227,
+      "learning_rate": 5.2734146584040515e-06,
+      "loss": 0.8425,
+      "step": 28450
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 18.444782487313137,
+      "learning_rate": 5.2717780840773966e-06,
+      "loss": 0.8349,
+      "step": 28455
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.317603540270103,
+      "learning_rate": 5.270141480546986e-06,
+      "loss": 0.8091,
+      "step": 28460
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.180428451338997,
+      "learning_rate": 5.268504847988682e-06,
+      "loss": 0.8081,
+      "step": 28465
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.209484459801911,
+      "learning_rate": 5.266868186578348e-06,
+      "loss": 0.8028,
+      "step": 28470
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.031280521102857,
+      "learning_rate": 5.265231496491849e-06,
+      "loss": 0.8212,
+      "step": 28475
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 21.740918196196784,
+      "learning_rate": 5.263594777905055e-06,
+      "loss": 0.8097,
+      "step": 28480
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.710009659556956,
+      "learning_rate": 5.261958030993839e-06,
+      "loss": 0.8248,
+      "step": 28485
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.624321534245403,
+      "learning_rate": 5.2603212559340766e-06,
+      "loss": 0.8303,
+      "step": 28490
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.159543330024958,
+      "learning_rate": 5.2586844529016455e-06,
+      "loss": 0.8359,
+      "step": 28495
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.356023775654574,
+      "learning_rate": 5.257047622072429e-06,
+      "loss": 0.8248,
+      "step": 28500
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.950501614008669,
+      "learning_rate": 5.25541076362231e-06,
+      "loss": 0.7807,
+      "step": 28505
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.302670648120601,
+      "learning_rate": 5.253773877727177e-06,
+      "loss": 0.8055,
+      "step": 28510
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.127714021436408,
+      "learning_rate": 5.252136964562924e-06,
+      "loss": 0.8283,
+      "step": 28515
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.3778428849436155,
+      "learning_rate": 5.250500024305439e-06,
+      "loss": 0.845,
+      "step": 28520
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.1378916131104955,
+      "learning_rate": 5.248863057130621e-06,
+      "loss": 0.804,
+      "step": 28525
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.645583237418911,
+      "learning_rate": 5.247226063214369e-06,
+      "loss": 0.8339,
+      "step": 28530
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.039580870099732,
+      "learning_rate": 5.245589042732586e-06,
+      "loss": 0.8129,
+      "step": 28535
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.2790213479552435,
+      "learning_rate": 5.243951995861176e-06,
+      "loss": 0.8125,
+      "step": 28540
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.55020039178158,
+      "learning_rate": 5.242314922776048e-06,
+      "loss": 0.8256,
+      "step": 28545
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.6076665561491135,
+      "learning_rate": 5.240677823653112e-06,
+      "loss": 0.8748,
+      "step": 28550
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 24.278825335381025,
+      "learning_rate": 5.23904069866828e-06,
+      "loss": 0.8338,
+      "step": 28555
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.42729182361593,
+      "learning_rate": 5.237403547997471e-06,
+      "loss": 0.8013,
+      "step": 28560
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.773900554676061,
+      "learning_rate": 5.235766371816604e-06,
+      "loss": 0.8201,
+      "step": 28565
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.500273051139277,
+      "learning_rate": 5.2341291703016e-06,
+      "loss": 0.8157,
+      "step": 28570
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 22.024888483130667,
+      "learning_rate": 5.232491943628384e-06,
+      "loss": 0.8235,
+      "step": 28575
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 17.465939287461797,
+      "learning_rate": 5.230854691972881e-06,
+      "loss": 0.8051,
+      "step": 28580
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 28.469470431834168,
+      "learning_rate": 5.229217415511022e-06,
+      "loss": 0.8297,
+      "step": 28585
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 19.61513964826842,
+      "learning_rate": 5.227580114418741e-06,
+      "loss": 0.7875,
+      "step": 28590
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 28.960400754131758,
+      "learning_rate": 5.225942788871971e-06,
+      "loss": 0.9221,
+      "step": 28595
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 34.85775133734265,
+      "learning_rate": 5.224305439046653e-06,
+      "loss": 0.826,
+      "step": 28600
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 32.89610012348651,
+      "learning_rate": 5.222668065118725e-06,
+      "loss": 0.8416,
+      "step": 28605
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 21.755337613991,
+      "learning_rate": 5.221030667264131e-06,
+      "loss": 0.8629,
+      "step": 28610
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 30.940812572866243,
+      "learning_rate": 5.219393245658816e-06,
+      "loss": 0.8116,
+      "step": 28615
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.6509867370032225,
+      "learning_rate": 5.217755800478728e-06,
+      "loss": 0.8112,
+      "step": 28620
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.937239650798151,
+      "learning_rate": 5.216118331899819e-06,
+      "loss": 0.7856,
+      "step": 28625
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 33.073728592417766,
+      "learning_rate": 5.214480840098042e-06,
+      "loss": 0.8371,
+      "step": 28630
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.417096497324796,
+      "learning_rate": 5.21284332524935e-06,
+      "loss": 0.7839,
+      "step": 28635
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 16.509024976662598,
+      "learning_rate": 5.211205787529704e-06,
+      "loss": 0.8257,
+      "step": 28640
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.721006779388342,
+      "learning_rate": 5.209568227115065e-06,
+      "loss": 0.8432,
+      "step": 28645
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.370181041668513,
+      "learning_rate": 5.207930644181393e-06,
+      "loss": 0.8082,
+      "step": 28650
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.423895454065968,
+      "learning_rate": 5.206293038904658e-06,
+      "loss": 0.7768,
+      "step": 28655
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.835741904843163,
+      "learning_rate": 5.204655411460826e-06,
+      "loss": 0.8313,
+      "step": 28660
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.913292766298175,
+      "learning_rate": 5.203017762025866e-06,
+      "loss": 0.8042,
+      "step": 28665
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.02807283367277,
+      "learning_rate": 5.20138009077575e-06,
+      "loss": 0.8219,
+      "step": 28670
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.091339301424439,
+      "learning_rate": 5.199742397886457e-06,
+      "loss": 0.8608,
+      "step": 28675
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.285834745633364,
+      "learning_rate": 5.19810468353396e-06,
+      "loss": 0.8222,
+      "step": 28680
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.516574615927582,
+      "learning_rate": 5.196466947894241e-06,
+      "loss": 0.815,
+      "step": 28685
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.0057631378192236,
+      "learning_rate": 5.194829191143281e-06,
+      "loss": 0.8031,
+      "step": 28690
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.965740632432216,
+      "learning_rate": 5.193191413457065e-06,
+      "loss": 0.8364,
+      "step": 28695
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.045844744196989,
+      "learning_rate": 5.1915536150115785e-06,
+      "loss": 0.8203,
+      "step": 28700
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.274808808539857,
+      "learning_rate": 5.1899157959828105e-06,
+      "loss": 0.8586,
+      "step": 28705
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.203246946244696,
+      "learning_rate": 5.188277956546753e-06,
+      "loss": 0.7896,
+      "step": 28710
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.9627082070989506,
+      "learning_rate": 5.186640096879398e-06,
+      "loss": 0.7943,
+      "step": 28715
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.90119774631027,
+      "learning_rate": 5.185002217156742e-06,
+      "loss": 0.8242,
+      "step": 28720
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.557675697750803,
+      "learning_rate": 5.18336431755478e-06,
+      "loss": 0.8143,
+      "step": 28725
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.881661502349697,
+      "learning_rate": 5.1817263982495115e-06,
+      "loss": 0.8028,
+      "step": 28730
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.085382369715009,
+      "learning_rate": 5.18008845941694e-06,
+      "loss": 0.7911,
+      "step": 28735
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.216198443207247,
+      "learning_rate": 5.17845050123307e-06,
+      "loss": 0.813,
+      "step": 28740
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.685846069694916,
+      "learning_rate": 5.176812523873907e-06,
+      "loss": 0.8364,
+      "step": 28745
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.908023317253362,
+      "learning_rate": 5.175174527515456e-06,
+      "loss": 0.834,
+      "step": 28750
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.344547124321556,
+      "learning_rate": 5.173536512333729e-06,
+      "loss": 0.7877,
+      "step": 28755
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.776375872443901,
+      "learning_rate": 5.171898478504738e-06,
+      "loss": 0.7963,
+      "step": 28760
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.1794716386261195,
+      "learning_rate": 5.170260426204497e-06,
+      "loss": 0.8484,
+      "step": 28765
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.451058636147341,
+      "learning_rate": 5.1686223556090224e-06,
+      "loss": 0.8327,
+      "step": 28770
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.46441247888714,
+      "learning_rate": 5.1669842668943306e-06,
+      "loss": 0.8248,
+      "step": 28775
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.877061812176017,
+      "learning_rate": 5.165346160236443e-06,
+      "loss": 0.7816,
+      "step": 28780
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.741123780733577,
+      "learning_rate": 5.16370803581138e-06,
+      "loss": 0.8224,
+      "step": 28785
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.88265021141814,
+      "learning_rate": 5.162069893795167e-06,
+      "loss": 0.8327,
+      "step": 28790
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.586186895688275,
+      "learning_rate": 5.160431734363828e-06,
+      "loss": 0.8383,
+      "step": 28795
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.773180350430062,
+      "learning_rate": 5.158793557693392e-06,
+      "loss": 0.8317,
+      "step": 28800
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.26473750806853,
+      "learning_rate": 5.157155363959887e-06,
+      "loss": 0.8056,
+      "step": 28805
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.082843396271144,
+      "learning_rate": 5.155517153339344e-06,
+      "loss": 0.8532,
+      "step": 28810
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.019847333823887,
+      "learning_rate": 5.153878926007797e-06,
+      "loss": 0.8124,
+      "step": 28815
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.5005656325108,
+      "learning_rate": 5.15224068214128e-06,
+      "loss": 0.8311,
+      "step": 28820
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.202641021959506,
+      "learning_rate": 5.15060242191583e-06,
+      "loss": 0.8252,
+      "step": 28825
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.522168519484391,
+      "learning_rate": 5.1489641455074846e-06,
+      "loss": 0.7817,
+      "step": 28830
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.19026913316636,
+      "learning_rate": 5.147325853092285e-06,
+      "loss": 0.8029,
+      "step": 28835
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.068565292854005,
+      "learning_rate": 5.145687544846272e-06,
+      "loss": 0.848,
+      "step": 28840
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.18378413557943,
+      "learning_rate": 5.144049220945489e-06,
+      "loss": 0.8461,
+      "step": 28845
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.8838921355026566,
+      "learning_rate": 5.142410881565982e-06,
+      "loss": 0.7872,
+      "step": 28850
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.665106347052056,
+      "learning_rate": 5.140772526883797e-06,
+      "loss": 0.7971,
+      "step": 28855
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.1166735629730065,
+      "learning_rate": 5.139134157074984e-06,
+      "loss": 0.8436,
+      "step": 28860
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.847427139454322,
+      "learning_rate": 5.137495772315591e-06,
+      "loss": 0.8124,
+      "step": 28865
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.837221321593361,
+      "learning_rate": 5.1358573727816685e-06,
+      "loss": 0.8529,
+      "step": 28870
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.201604849290314,
+      "learning_rate": 5.134218958649274e-06,
+      "loss": 0.8133,
+      "step": 28875
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.032899750780574,
+      "learning_rate": 5.132580530094459e-06,
+      "loss": 0.8408,
+      "step": 28880
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.202869799994652,
+      "learning_rate": 5.130942087293283e-06,
+      "loss": 0.8099,
+      "step": 28885
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.6021648021502,
+      "learning_rate": 5.1293036304217995e-06,
+      "loss": 0.8057,
+      "step": 28890
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.195860451037316,
+      "learning_rate": 5.12766515965607e-06,
+      "loss": 0.7707,
+      "step": 28895
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.125250980532645,
+      "learning_rate": 5.126026675172157e-06,
+      "loss": 0.8227,
+      "step": 28900
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.398595748643314,
+      "learning_rate": 5.124388177146121e-06,
+      "loss": 0.7838,
+      "step": 28905
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.700082787818464,
+      "learning_rate": 5.122749665754026e-06,
+      "loss": 0.8456,
+      "step": 28910
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 23.825412981295585,
+      "learning_rate": 5.121111141171938e-06,
+      "loss": 0.8168,
+      "step": 28915
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.916138066122154,
+      "learning_rate": 5.119472603575924e-06,
+      "loss": 0.8153,
+      "step": 28920
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.0793455181871625,
+      "learning_rate": 5.117834053142051e-06,
+      "loss": 0.8472,
+      "step": 28925
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.323977949052801,
+      "learning_rate": 5.116195490046388e-06,
+      "loss": 0.8472,
+      "step": 28930
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.238936639245196,
+      "learning_rate": 5.114556914465009e-06,
+      "loss": 0.7972,
+      "step": 28935
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.243687229405834,
+      "learning_rate": 5.112918326573982e-06,
+      "loss": 0.8409,
+      "step": 28940
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.783046907850406,
+      "learning_rate": 5.111279726549384e-06,
+      "loss": 0.7658,
+      "step": 28945
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.4566288577281625,
+      "learning_rate": 5.109641114567288e-06,
+      "loss": 0.7915,
+      "step": 28950
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.741691260101394,
+      "learning_rate": 5.1080024908037694e-06,
+      "loss": 0.86,
+      "step": 28955
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.270059758269816,
+      "learning_rate": 5.106363855434907e-06,
+      "loss": 0.8123,
+      "step": 28960
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.904700105678346,
+      "learning_rate": 5.104725208636777e-06,
+      "loss": 0.8222,
+      "step": 28965
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.837541181724244,
+      "learning_rate": 5.103086550585463e-06,
+      "loss": 0.8085,
+      "step": 28970
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.179304702646292,
+      "learning_rate": 5.101447881457046e-06,
+      "loss": 0.8227,
+      "step": 28975
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 4.776609548219907,
+      "learning_rate": 5.099809201427603e-06,
+      "loss": 0.7846,
+      "step": 28980
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.730764020627527,
+      "learning_rate": 5.098170510673224e-06,
+      "loss": 0.7935,
+      "step": 28985
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.635731099123605,
+      "learning_rate": 5.096531809369988e-06,
+      "loss": 0.8109,
+      "step": 28990
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.810360155246453,
+      "learning_rate": 5.094893097693983e-06,
+      "loss": 0.8293,
+      "step": 28995
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.460276744279129,
+      "learning_rate": 5.093254375821299e-06,
+      "loss": 0.7785,
+      "step": 29000
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 4.728778861040613,
+      "learning_rate": 5.091615643928016e-06,
+      "loss": 0.789,
+      "step": 29005
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.868446201290173,
+      "learning_rate": 5.08997690219023e-06,
+      "loss": 0.8192,
+      "step": 29010
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 4.8510699300250755,
+      "learning_rate": 5.088338150784029e-06,
+      "loss": 0.7578,
+      "step": 29015
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.562075881914659,
+      "learning_rate": 5.0866993898855045e-06,
+      "loss": 0.8352,
+      "step": 29020
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.143850679057571,
+      "learning_rate": 5.085060619670746e-06,
+      "loss": 0.7261,
+      "step": 29025
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.397323078474872,
+      "learning_rate": 5.083421840315851e-06,
+      "loss": 0.8438,
+      "step": 29030
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.808400027460724,
+      "learning_rate": 5.08178305199691e-06,
+      "loss": 0.8454,
+      "step": 29035
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.881506130414715,
+      "learning_rate": 5.0801442548900195e-06,
+      "loss": 0.8391,
+      "step": 29040
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.679740233421449,
+      "learning_rate": 5.078505449171274e-06,
+      "loss": 0.8092,
+      "step": 29045
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.404697141770395,
+      "learning_rate": 5.076866635016772e-06,
+      "loss": 0.7824,
+      "step": 29050
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.904767650446284,
+      "learning_rate": 5.07522781260261e-06,
+      "loss": 0.8483,
+      "step": 29055
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 25.718605229696344,
+      "learning_rate": 5.073588982104889e-06,
+      "loss": 0.8033,
+      "step": 29060
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.759656116234666,
+      "learning_rate": 5.071950143699707e-06,
+      "loss": 0.801,
+      "step": 29065
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.32674223064685,
+      "learning_rate": 5.070311297563163e-06,
+      "loss": 0.811,
+      "step": 29070
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.494047609200806,
+      "learning_rate": 5.068672443871361e-06,
+      "loss": 0.8213,
+      "step": 29075
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.18498999013527,
+      "learning_rate": 5.067033582800401e-06,
+      "loss": 0.8329,
+      "step": 29080
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.603159807360061,
+      "learning_rate": 5.0653947145263885e-06,
+      "loss": 0.7768,
+      "step": 29085
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.705809007745918,
+      "learning_rate": 5.0637558392254235e-06,
+      "loss": 0.8532,
+      "step": 29090
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.485124137036119,
+      "learning_rate": 5.062116957073613e-06,
+      "loss": 0.7952,
+      "step": 29095
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.993705068357134,
+      "learning_rate": 5.06047806824706e-06,
+      "loss": 0.8336,
+      "step": 29100
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.681963323212381,
+      "learning_rate": 5.058839172921873e-06,
+      "loss": 0.797,
+      "step": 29105
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.111160127137747,
+      "learning_rate": 5.057200271274156e-06,
+      "loss": 0.8228,
+      "step": 29110
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.411693347426782,
+      "learning_rate": 5.05556136348002e-06,
+      "loss": 0.8309,
+      "step": 29115
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.62345410385421,
+      "learning_rate": 5.05392244971557e-06,
+      "loss": 0.8151,
+      "step": 29120
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.933159233687968,
+      "learning_rate": 5.052283530156912e-06,
+      "loss": 0.8178,
+      "step": 29125
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.04589105362451,
+      "learning_rate": 5.050644604980159e-06,
+      "loss": 0.7979,
+      "step": 29130
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.277252822990433,
+      "learning_rate": 5.04900567436142e-06,
+      "loss": 0.7749,
+      "step": 29135
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.322892892107303,
+      "learning_rate": 5.0473667384768054e-06,
+      "loss": 0.7885,
+      "step": 29140
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 20.882763964089886,
+      "learning_rate": 5.045727797502426e-06,
+      "loss": 0.8309,
+      "step": 29145
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.806843017144521,
+      "learning_rate": 5.044088851614391e-06,
+      "loss": 0.7765,
+      "step": 29150
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.324009815413327,
+      "learning_rate": 5.0424499009888165e-06,
+      "loss": 0.8049,
+      "step": 29155
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.711334284462785,
+      "learning_rate": 5.040810945801812e-06,
+      "loss": 0.7728,
+      "step": 29160
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.8084473931244585,
+      "learning_rate": 5.039171986229491e-06,
+      "loss": 0.8313,
+      "step": 29165
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.524139245799343,
+      "learning_rate": 5.037533022447967e-06,
+      "loss": 0.8093,
+      "step": 29170
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.73566929702185,
+      "learning_rate": 5.0358940546333555e-06,
+      "loss": 0.8271,
+      "step": 29175
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.6962732035184604,
+      "learning_rate": 5.034255082961768e-06,
+      "loss": 0.7423,
+      "step": 29180
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.533090871706895,
+      "learning_rate": 5.032616107609322e-06,
+      "loss": 0.8142,
+      "step": 29185
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.408774450153409,
+      "learning_rate": 5.0309771287521295e-06,
+      "loss": 0.8627,
+      "step": 29190
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.156362265169408,
+      "learning_rate": 5.029338146566307e-06,
+      "loss": 0.7984,
+      "step": 29195
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.754148694201769,
+      "learning_rate": 5.027699161227975e-06,
+      "loss": 0.8132,
+      "step": 29200
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.874061883796169,
+      "learning_rate": 5.026060172913243e-06,
+      "loss": 0.785,
+      "step": 29205
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.070895173347584,
+      "learning_rate": 5.024421181798231e-06,
+      "loss": 0.7783,
+      "step": 29210
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.08443732562972,
+      "learning_rate": 5.022782188059056e-06,
+      "loss": 0.8524,
+      "step": 29215
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.392687815404849,
+      "learning_rate": 5.021143191871834e-06,
+      "loss": 0.7729,
+      "step": 29220
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.464790980083453,
+      "learning_rate": 5.019504193412682e-06,
+      "loss": 0.8182,
+      "step": 29225
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.628883840032938,
+      "learning_rate": 5.01786519285772e-06,
+      "loss": 0.7888,
+      "step": 29230
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.04323033822842,
+      "learning_rate": 5.016226190383061e-06,
+      "loss": 0.8187,
+      "step": 29235
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.726518575213119,
+      "learning_rate": 5.014587186164828e-06,
+      "loss": 0.7887,
+      "step": 29240
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.160618859165872,
+      "learning_rate": 5.012948180379138e-06,
+      "loss": 0.7665,
+      "step": 29245
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.291806868973104,
+      "learning_rate": 5.011309173202108e-06,
+      "loss": 0.8027,
+      "step": 29250
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.45783451634518,
+      "learning_rate": 5.009670164809858e-06,
+      "loss": 0.7657,
+      "step": 29255
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.4830726465245,
+      "learning_rate": 5.0080311553785075e-06,
+      "loss": 0.7731,
+      "step": 29260
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.546063778080442,
+      "learning_rate": 5.006392145084172e-06,
+      "loss": 0.7672,
+      "step": 29265
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.775772578098124,
+      "learning_rate": 5.004753134102973e-06,
+      "loss": 0.8346,
+      "step": 29270
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.5794976908939935,
+      "learning_rate": 5.003114122611028e-06,
+      "loss": 0.7771,
+      "step": 29275
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.762264800576519,
+      "learning_rate": 5.0014751107844576e-06,
+      "loss": 0.8228,
+      "step": 29280
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.9314046914117835,
+      "learning_rate": 4.999836098799379e-06,
+      "loss": 0.7801,
+      "step": 29285
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.2197424090684,
+      "learning_rate": 4.998197086831912e-06,
+      "loss": 0.8282,
+      "step": 29290
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.863892250220879,
+      "learning_rate": 4.996558075058179e-06,
+      "loss": 0.7849,
+      "step": 29295
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.88712624080577,
+      "learning_rate": 4.994919063654295e-06,
+      "loss": 0.8118,
+      "step": 29300
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.899540341700078,
+      "learning_rate": 4.993280052796378e-06,
+      "loss": 0.8211,
+      "step": 29305
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.366164063496896,
+      "learning_rate": 4.99164104266055e-06,
+      "loss": 0.7674,
+      "step": 29310
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.0300986008164665,
+      "learning_rate": 4.990002033422928e-06,
+      "loss": 0.8071,
+      "step": 29315
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.294204839684711,
+      "learning_rate": 4.988363025259635e-06,
+      "loss": 0.7988,
+      "step": 29320
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.682600958364213,
+      "learning_rate": 4.986724018346782e-06,
+      "loss": 0.7934,
+      "step": 29325
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.708390759590833,
+      "learning_rate": 4.985085012860493e-06,
+      "loss": 0.829,
+      "step": 29330
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.758827242074737,
+      "learning_rate": 4.983446008976887e-06,
+      "loss": 0.7857,
+      "step": 29335
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 24.59076245051149,
+      "learning_rate": 4.9818070068720785e-06,
+      "loss": 0.8186,
+      "step": 29340
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.659853242933357,
+      "learning_rate": 4.980168006722189e-06,
+      "loss": 0.8073,
+      "step": 29345
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.466058576157828,
+      "learning_rate": 4.978529008703333e-06,
+      "loss": 0.8042,
+      "step": 29350
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.89881613598247,
+      "learning_rate": 4.976890012991631e-06,
+      "loss": 0.7984,
+      "step": 29355
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 18.077320619114,
+      "learning_rate": 4.9752510197632e-06,
+      "loss": 0.8476,
+      "step": 29360
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.424010477811986,
+      "learning_rate": 4.973612029194154e-06,
+      "loss": 0.8136,
+      "step": 29365
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.770990192838112,
+      "learning_rate": 4.971973041460613e-06,
+      "loss": 0.8047,
+      "step": 29370
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.662160218025914,
+      "learning_rate": 4.970334056738692e-06,
+      "loss": 0.8108,
+      "step": 29375
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.411098912694886,
+      "learning_rate": 4.968695075204508e-06,
+      "loss": 0.7702,
+      "step": 29380
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.7155223274576015,
+      "learning_rate": 4.967056097034177e-06,
+      "loss": 0.8121,
+      "step": 29385
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.371586893614241,
+      "learning_rate": 4.965417122403811e-06,
+      "loss": 0.7996,
+      "step": 29390
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.282520779370752,
+      "learning_rate": 4.9637781514895315e-06,
+      "loss": 0.8254,
+      "step": 29395
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.877233785679275,
+      "learning_rate": 4.962139184467446e-06,
+      "loss": 0.7821,
+      "step": 29400
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.062370566173023,
+      "learning_rate": 4.960500221513675e-06,
+      "loss": 0.8165,
+      "step": 29405
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.062883174542722,
+      "learning_rate": 4.958861262804327e-06,
+      "loss": 0.8096,
+      "step": 29410
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.344973583028538,
+      "learning_rate": 4.95722230851552e-06,
+      "loss": 0.7902,
+      "step": 29415
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.269743878820086,
+      "learning_rate": 4.955583358823364e-06,
+      "loss": 0.7993,
+      "step": 29420
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.024395665734921,
+      "learning_rate": 4.953944413903971e-06,
+      "loss": 0.8432,
+      "step": 29425
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.421035473683998,
+      "learning_rate": 4.952305473933456e-06,
+      "loss": 0.7857,
+      "step": 29430
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.588717028826829,
+      "learning_rate": 4.950666539087927e-06,
+      "loss": 0.78,
+      "step": 29435
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.632604368468013,
+      "learning_rate": 4.949027609543497e-06,
+      "loss": 0.7669,
+      "step": 29440
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.630758459389194,
+      "learning_rate": 4.947388685476276e-06,
+      "loss": 0.7864,
+      "step": 29445
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.018305897013323,
+      "learning_rate": 4.94574976706237e-06,
+      "loss": 0.7852,
+      "step": 29450
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.948471414354077,
+      "learning_rate": 4.944110854477894e-06,
+      "loss": 0.8035,
+      "step": 29455
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.109579858646023,
+      "learning_rate": 4.9424719478989515e-06,
+      "loss": 0.7884,
+      "step": 29460
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.266712112485102,
+      "learning_rate": 4.940833047501652e-06,
+      "loss": 0.8263,
+      "step": 29465
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.535824088404048,
+      "learning_rate": 4.939194153462103e-06,
+      "loss": 0.8198,
+      "step": 29470
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.107778575655036,
+      "learning_rate": 4.937555265956411e-06,
+      "loss": 0.8127,
+      "step": 29475
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.346506795232058,
+      "learning_rate": 4.935916385160683e-06,
+      "loss": 0.8092,
+      "step": 29480
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.557190475765134,
+      "learning_rate": 4.934277511251019e-06,
+      "loss": 0.8308,
+      "step": 29485
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.675250707152214,
+      "learning_rate": 4.932638644403529e-06,
+      "loss": 0.8183,
+      "step": 29490
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.956476470838433,
+      "learning_rate": 4.930999784794311e-06,
+      "loss": 0.8023,
+      "step": 29495
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.7709799701961035,
+      "learning_rate": 4.929360932599474e-06,
+      "loss": 0.8033,
+      "step": 29500
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.2637823075954335,
+      "learning_rate": 4.927722087995115e-06,
+      "loss": 0.776,
+      "step": 29505
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.495390031682441,
+      "learning_rate": 4.926083251157336e-06,
+      "loss": 0.8069,
+      "step": 29510
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.54021499002464,
+      "learning_rate": 4.924444422262239e-06,
+      "loss": 0.807,
+      "step": 29515
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.066093691697654,
+      "learning_rate": 4.922805601485921e-06,
+      "loss": 0.7629,
+      "step": 29520
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.9980763077862305,
+      "learning_rate": 4.9211667890044835e-06,
+      "loss": 0.7896,
+      "step": 29525
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.452395057883908,
+      "learning_rate": 4.919527984994023e-06,
+      "loss": 0.806,
+      "step": 29530
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 18.906589815923898,
+      "learning_rate": 4.9178891896306326e-06,
+      "loss": 0.8162,
+      "step": 29535
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.809767699924567,
+      "learning_rate": 4.916250403090414e-06,
+      "loss": 0.8011,
+      "step": 29540
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.363769083652064,
+      "learning_rate": 4.914611625549456e-06,
+      "loss": 0.7945,
+      "step": 29545
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.202218114203786,
+      "learning_rate": 4.912972857183859e-06,
+      "loss": 0.786,
+      "step": 29550
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.660272967854556,
+      "learning_rate": 4.91133409816971e-06,
+      "loss": 0.8108,
+      "step": 29555
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.476911932929125,
+      "learning_rate": 4.909695348683104e-06,
+      "loss": 0.7923,
+      "step": 29560
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.806783499374322,
+      "learning_rate": 4.908056608900133e-06,
+      "loss": 0.8095,
+      "step": 29565
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.138858436291346,
+      "learning_rate": 4.906417878996883e-06,
+      "loss": 0.7477,
+      "step": 29570
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.895158280181477,
+      "learning_rate": 4.9047791591494475e-06,
+      "loss": 0.8601,
+      "step": 29575
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.47706193638025,
+      "learning_rate": 4.903140449533909e-06,
+      "loss": 0.8103,
+      "step": 29580
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.993411544865312,
+      "learning_rate": 4.901501750326359e-06,
+      "loss": 0.8425,
+      "step": 29585
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.702668991067775,
+      "learning_rate": 4.899863061702881e-06,
+      "loss": 0.8218,
+      "step": 29590
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.2106236377777675,
+      "learning_rate": 4.898224383839556e-06,
+      "loss": 0.8052,
+      "step": 29595
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.815829897033973,
+      "learning_rate": 4.8965857169124714e-06,
+      "loss": 0.8285,
+      "step": 29600
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.7280018731393865,
+      "learning_rate": 4.894947061097708e-06,
+      "loss": 0.7939,
+      "step": 29605
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.512877569309113,
+      "learning_rate": 4.893308416571347e-06,
+      "loss": 0.7994,
+      "step": 29610
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.532257727972996,
+      "learning_rate": 4.891669783509468e-06,
+      "loss": 0.8087,
+      "step": 29615
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.9606685351204,
+      "learning_rate": 4.8900311620881464e-06,
+      "loss": 0.7874,
+      "step": 29620
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.126604625139472,
+      "learning_rate": 4.888392552483465e-06,
+      "loss": 0.7716,
+      "step": 29625
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.565607606006656,
+      "learning_rate": 4.8867539548714936e-06,
+      "loss": 0.7736,
+      "step": 29630
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.165534621343919,
+      "learning_rate": 4.885115369428311e-06,
+      "loss": 0.8528,
+      "step": 29635
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.433104211919282,
+      "learning_rate": 4.883476796329988e-06,
+      "loss": 0.8179,
+      "step": 29640
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.976847490224073,
+      "learning_rate": 4.881838235752599e-06,
+      "loss": 0.7883,
+      "step": 29645
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.353315632211304,
+      "learning_rate": 4.8801996878722115e-06,
+      "loss": 0.7951,
+      "step": 29650
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.689167601321623,
+      "learning_rate": 4.878561152864896e-06,
+      "loss": 0.7685,
+      "step": 29655
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.763086204971069,
+      "learning_rate": 4.8769226309067225e-06,
+      "loss": 0.808,
+      "step": 29660
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.235599067821967,
+      "learning_rate": 4.875284122173753e-06,
+      "loss": 0.8003,
+      "step": 29665
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.036504165449317,
+      "learning_rate": 4.873645626842058e-06,
+      "loss": 0.8137,
+      "step": 29670
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.135836833289432,
+      "learning_rate": 4.872007145087698e-06,
+      "loss": 0.7674,
+      "step": 29675
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.506328651735684,
+      "learning_rate": 4.870368677086732e-06,
+      "loss": 0.8207,
+      "step": 29680
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.629886198396468,
+      "learning_rate": 4.868730223015227e-06,
+      "loss": 0.8029,
+      "step": 29685
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.824316072670978,
+      "learning_rate": 4.867091783049236e-06,
+      "loss": 0.8197,
+      "step": 29690
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.418927302091987,
+      "learning_rate": 4.865453357364821e-06,
+      "loss": 0.7727,
+      "step": 29695
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.571949221645414,
+      "learning_rate": 4.8638149461380364e-06,
+      "loss": 0.7894,
+      "step": 29700
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.467381923539769,
+      "learning_rate": 4.8621765495449354e-06,
+      "loss": 0.8143,
+      "step": 29705
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.121398006803582,
+      "learning_rate": 4.860538167761575e-06,
+      "loss": 0.8145,
+      "step": 29710
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.141009686601915,
+      "learning_rate": 4.858899800964002e-06,
+      "loss": 0.8161,
+      "step": 29715
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.815248446997574,
+      "learning_rate": 4.85726144932827e-06,
+      "loss": 0.7666,
+      "step": 29720
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.300760267108112,
+      "learning_rate": 4.855623113030423e-06,
+      "loss": 0.7932,
+      "step": 29725
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.222183631700164,
+      "learning_rate": 4.8539847922465115e-06,
+      "loss": 0.8275,
+      "step": 29730
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.599627578657428,
+      "learning_rate": 4.8523464871525785e-06,
+      "loss": 0.7501,
+      "step": 29735
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.447940565078039,
+      "learning_rate": 4.850708197924666e-06,
+      "loss": 0.7678,
+      "step": 29740
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.126041805046706,
+      "learning_rate": 4.8490699247388165e-06,
+      "loss": 0.8297,
+      "step": 29745
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.746869691429128,
+      "learning_rate": 4.8474316677710695e-06,
+      "loss": 0.8213,
+      "step": 29750
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.25782994771493,
+      "learning_rate": 4.845793427197465e-06,
+      "loss": 0.7632,
+      "step": 29755
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 14.112215242568517,
+      "learning_rate": 4.844155203194036e-06,
+      "loss": 0.8075,
+      "step": 29760
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.247021941465757,
+      "learning_rate": 4.842516995936817e-06,
+      "loss": 0.7925,
+      "step": 29765
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.407277703325768,
+      "learning_rate": 4.840878805601845e-06,
+      "loss": 0.7922,
+      "step": 29770
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.07675960183057,
+      "learning_rate": 4.839240632365144e-06,
+      "loss": 0.7974,
+      "step": 29775
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.337796061320492,
+      "learning_rate": 4.83760247640275e-06,
+      "loss": 0.7534,
+      "step": 29780
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 28.93401640164091,
+      "learning_rate": 4.835964337890684e-06,
+      "loss": 0.7885,
+      "step": 29785
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.14149833183626,
+      "learning_rate": 4.834326217004974e-06,
+      "loss": 0.8125,
+      "step": 29790
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.61939976416113,
+      "learning_rate": 4.8326881139216446e-06,
+      "loss": 0.7885,
+      "step": 29795
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.308973243131987,
+      "learning_rate": 4.831050028816713e-06,
+      "loss": 0.7826,
+      "step": 29800
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.166485818628002,
+      "learning_rate": 4.829411961866205e-06,
+      "loss": 0.7902,
+      "step": 29805
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.235460655298076,
+      "learning_rate": 4.827773913246132e-06,
+      "loss": 0.7853,
+      "step": 29810
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.012300261771474,
+      "learning_rate": 4.826135883132514e-06,
+      "loss": 0.7916,
+      "step": 29815
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.4966381445275,
+      "learning_rate": 4.824497871701363e-06,
+      "loss": 0.7992,
+      "step": 29820
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.7356534738914435,
+      "learning_rate": 4.822859879128687e-06,
+      "loss": 0.7997,
+      "step": 29825
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.403234122506506,
+      "learning_rate": 4.821221905590501e-06,
+      "loss": 0.8447,
+      "step": 29830
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.960980550710589,
+      "learning_rate": 4.81958395126281e-06,
+      "loss": 0.7782,
+      "step": 29835
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.2252945598010765,
+      "learning_rate": 4.817946016321619e-06,
+      "loss": 0.8208,
+      "step": 29840
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.996738294284722,
+      "learning_rate": 4.816308100942935e-06,
+      "loss": 0.7859,
+      "step": 29845
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.681301926252344,
+      "learning_rate": 4.8146702053027525e-06,
+      "loss": 0.7857,
+      "step": 29850
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.932112347604734,
+      "learning_rate": 4.813032329577076e-06,
+      "loss": 0.7957,
+      "step": 29855
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.849151009617056,
+      "learning_rate": 4.8113944739419e-06,
+      "loss": 0.8329,
+      "step": 29860
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.625374211275668,
+      "learning_rate": 4.809756638573222e-06,
+      "loss": 0.8123,
+      "step": 29865
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.362709994828513,
+      "learning_rate": 4.8081188236470315e-06,
+      "loss": 0.7906,
+      "step": 29870
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.880402705932108,
+      "learning_rate": 4.806481029339321e-06,
+      "loss": 0.7437,
+      "step": 29875
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.69619434877722,
+      "learning_rate": 4.8048432558260765e-06,
+      "loss": 0.7687,
+      "step": 29880
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.418283962974726,
+      "learning_rate": 4.803205503283285e-06,
+      "loss": 0.8102,
+      "step": 29885
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.137728309150263,
+      "learning_rate": 4.801567771886933e-06,
+      "loss": 0.8295,
+      "step": 29890
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.98234463446497,
+      "learning_rate": 4.7999300618129976e-06,
+      "loss": 0.7419,
+      "step": 29895
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.746731612064017,
+      "learning_rate": 4.798292373237462e-06,
+      "loss": 0.8162,
+      "step": 29900
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.864035556704577,
+      "learning_rate": 4.796654706336301e-06,
+      "loss": 0.8153,
+      "step": 29905
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.749274723640831,
+      "learning_rate": 4.795017061285487e-06,
+      "loss": 0.8055,
+      "step": 29910
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.622351675017694,
+      "learning_rate": 4.793379438260997e-06,
+      "loss": 0.8148,
+      "step": 29915
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.051848746409057,
+      "learning_rate": 4.7917418374387955e-06,
+      "loss": 0.7543,
+      "step": 29920
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.258996248407842,
+      "learning_rate": 4.790104258994855e-06,
+      "loss": 0.7768,
+      "step": 29925
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.097692331287701,
+      "learning_rate": 4.788466703105138e-06,
+      "loss": 0.7802,
+      "step": 29930
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.725187362488219,
+      "learning_rate": 4.786829169945606e-06,
+      "loss": 0.7569,
+      "step": 29935
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.264251665243966,
+      "learning_rate": 4.785191659692222e-06,
+      "loss": 0.824,
+      "step": 29940
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.892749424970571,
+      "learning_rate": 4.78355417252094e-06,
+      "loss": 0.7977,
+      "step": 29945
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.152772257028792,
+      "learning_rate": 4.78191670860772e-06,
+      "loss": 0.7677,
+      "step": 29950
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.7418885709930025,
+      "learning_rate": 4.78027926812851e-06,
+      "loss": 0.8035,
+      "step": 29955
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.703054725185859,
+      "learning_rate": 4.778641851259264e-06,
+      "loss": 0.8008,
+      "step": 29960
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 16.151759196592305,
+      "learning_rate": 4.777004458175928e-06,
+      "loss": 0.8152,
+      "step": 29965
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.910392749323405,
+      "learning_rate": 4.775367089054446e-06,
+      "loss": 0.7963,
+      "step": 29970
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.67542640599338,
+      "learning_rate": 4.773729744070762e-06,
+      "loss": 0.7965,
+      "step": 29975
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.08530116647858,
+      "learning_rate": 4.772092423400815e-06,
+      "loss": 0.8069,
+      "step": 29980
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.09800617392457,
+      "learning_rate": 4.770455127220544e-06,
+      "loss": 0.7727,
+      "step": 29985
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 33.23931494505973,
+      "learning_rate": 4.7688178557058834e-06,
+      "loss": 0.7698,
+      "step": 29990
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.856668275163106,
+      "learning_rate": 4.7671806090327625e-06,
+      "loss": 0.7984,
+      "step": 29995
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 29.485450180586472,
+      "learning_rate": 4.765543387377115e-06,
+      "loss": 0.8264,
+      "step": 30000
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.888031482974746,
+      "learning_rate": 4.763906190914862e-06,
+      "loss": 0.8033,
+      "step": 30005
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.47157323006894,
+      "learning_rate": 4.762269019821933e-06,
+      "loss": 0.8286,
+      "step": 30010
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.841732621791369,
+      "learning_rate": 4.760631874274247e-06,
+      "loss": 0.8062,
+      "step": 30015
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.181694156915245,
+      "learning_rate": 4.758994754447721e-06,
+      "loss": 0.8381,
+      "step": 30020
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.925935935029198,
+      "learning_rate": 4.757357660518275e-06,
+      "loss": 0.8215,
+      "step": 30025
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.24092344867244,
+      "learning_rate": 4.755720592661816e-06,
+      "loss": 0.8017,
+      "step": 30030
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.308951157601896,
+      "learning_rate": 4.75408355105426e-06,
+      "loss": 0.7926,
+      "step": 30035
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.639978704363475,
+      "learning_rate": 4.752446535871509e-06,
+      "loss": 0.7751,
+      "step": 30040
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.448822047152843,
+      "learning_rate": 4.750809547289473e-06,
+      "loss": 0.8199,
+      "step": 30045
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.910283424670611,
+      "learning_rate": 4.7491725854840516e-06,
+      "loss": 0.7898,
+      "step": 30050
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.488005216345067,
+      "learning_rate": 4.74753565063114e-06,
+      "loss": 0.766,
+      "step": 30055
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.585112345932322,
+      "learning_rate": 4.7458987429066374e-06,
+      "loss": 0.8018,
+      "step": 30060
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.352788415478089,
+      "learning_rate": 4.744261862486438e-06,
+      "loss": 0.8029,
+      "step": 30065
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.564595572553094,
+      "learning_rate": 4.742625009546429e-06,
+      "loss": 0.7866,
+      "step": 30070
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.3516501002682055,
+      "learning_rate": 4.740988184262499e-06,
+      "loss": 0.7552,
+      "step": 30075
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.031661111625874,
+      "learning_rate": 4.73935138681053e-06,
+      "loss": 0.7799,
+      "step": 30080
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.047777136097501,
+      "learning_rate": 4.737714617366408e-06,
+      "loss": 0.797,
+      "step": 30085
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.841185629960563,
+      "learning_rate": 4.736077876106005e-06,
+      "loss": 0.7988,
+      "step": 30090
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.902412267033933,
+      "learning_rate": 4.734441163205202e-06,
+      "loss": 0.7972,
+      "step": 30095
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.899823930188095,
+      "learning_rate": 4.7328044788398655e-06,
+      "loss": 0.822,
+      "step": 30100
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.244104167681226,
+      "learning_rate": 4.731167823185867e-06,
+      "loss": 0.7611,
+      "step": 30105
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.031387987540888,
+      "learning_rate": 4.729531196419074e-06,
+      "loss": 0.7859,
+      "step": 30110
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.551106765441283,
+      "learning_rate": 4.727894598715345e-06,
+      "loss": 0.8148,
+      "step": 30115
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.9615410825668045,
+      "learning_rate": 4.726258030250545e-06,
+      "loss": 0.774,
+      "step": 30120
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.299898513292915,
+      "learning_rate": 4.7246214912005255e-06,
+      "loss": 0.7852,
+      "step": 30125
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.080392471708402,
+      "learning_rate": 4.7229849817411445e-06,
+      "loss": 0.7968,
+      "step": 30130
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.385539634015399,
+      "learning_rate": 4.7213485020482495e-06,
+      "loss": 0.8311,
+      "step": 30135
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.118844227317314,
+      "learning_rate": 4.7197120522976855e-06,
+      "loss": 0.7762,
+      "step": 30140
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.017188656833131,
+      "learning_rate": 4.718075632665301e-06,
+      "loss": 0.7788,
+      "step": 30145
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.61556780088029,
+      "learning_rate": 4.7164392433269314e-06,
+      "loss": 0.8016,
+      "step": 30150
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.635523592967406,
+      "learning_rate": 4.7148028844584185e-06,
+      "loss": 0.7737,
+      "step": 30155
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.846318759564663,
+      "learning_rate": 4.713166556235594e-06,
+      "loss": 0.8064,
+      "step": 30160
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.30424642437604,
+      "learning_rate": 4.7115302588342885e-06,
+      "loss": 0.7453,
+      "step": 30165
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.831306235957011,
+      "learning_rate": 4.709893992430331e-06,
+      "loss": 0.7741,
+      "step": 30170
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.330872623207505,
+      "learning_rate": 4.708257757199542e-06,
+      "loss": 0.7999,
+      "step": 30175
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 25.378696340898358,
+      "learning_rate": 4.706621553317746e-06,
+      "loss": 0.7982,
+      "step": 30180
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.688725928598445,
+      "learning_rate": 4.7049853809607604e-06,
+      "loss": 0.7798,
+      "step": 30185
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.230638946395093,
+      "learning_rate": 4.703349240304395e-06,
+      "loss": 0.7861,
+      "step": 30190
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.038961700679986,
+      "learning_rate": 4.701713131524464e-06,
+      "loss": 0.7871,
+      "step": 30195
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 25.196247234254738,
+      "learning_rate": 4.7000770547967735e-06,
+      "loss": 0.8176,
+      "step": 30200
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.882688341047091,
+      "learning_rate": 4.6984410102971265e-06,
+      "loss": 0.7958,
+      "step": 30205
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.982819953624391,
+      "learning_rate": 4.696804998201324e-06,
+      "loss": 0.7492,
+      "step": 30210
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.202877011319478,
+      "learning_rate": 4.695169018685164e-06,
+      "loss": 0.7599,
+      "step": 30215
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 18.016970163694605,
+      "learning_rate": 4.693533071924438e-06,
+      "loss": 0.8026,
+      "step": 30220
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.749427803192396,
+      "learning_rate": 4.691897158094934e-06,
+      "loss": 0.8056,
+      "step": 30225
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.395352516790499,
+      "learning_rate": 4.690261277372442e-06,
+      "loss": 0.754,
+      "step": 30230
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.312698473040356,
+      "learning_rate": 4.68862542993274e-06,
+      "loss": 0.8053,
+      "step": 30235
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.961063223845596,
+      "learning_rate": 4.686989615951613e-06,
+      "loss": 0.8001,
+      "step": 30240
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.015842572499906,
+      "learning_rate": 4.685353835604832e-06,
+      "loss": 0.8056,
+      "step": 30245
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.8779439199749115,
+      "learning_rate": 4.683718089068169e-06,
+      "loss": 0.8126,
+      "step": 30250
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.033629123670667,
+      "learning_rate": 4.682082376517395e-06,
+      "loss": 0.796,
+      "step": 30255
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.04106153601148,
+      "learning_rate": 4.680446698128271e-06,
+      "loss": 0.7696,
+      "step": 30260
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.949650591888291,
+      "learning_rate": 4.678811054076561e-06,
+      "loss": 0.8016,
+      "step": 30265
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.103165173015239,
+      "learning_rate": 4.677175444538021e-06,
+      "loss": 0.8282,
+      "step": 30270
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.357339839085992,
+      "learning_rate": 4.675539869688402e-06,
+      "loss": 0.8257,
+      "step": 30275
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.004583256288209,
+      "learning_rate": 4.673904329703459e-06,
+      "loss": 0.734,
+      "step": 30280
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.032854359968172,
+      "learning_rate": 4.672268824758932e-06,
+      "loss": 0.8103,
+      "step": 30285
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.704184503885026,
+      "learning_rate": 4.6706333550305675e-06,
+      "loss": 0.7754,
+      "step": 30290
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.83805772876524,
+      "learning_rate": 4.668997920694103e-06,
+      "loss": 0.7956,
+      "step": 30295
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.60882884163727,
+      "learning_rate": 4.667362521925272e-06,
+      "loss": 0.7972,
+      "step": 30300
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.4931881686723205,
+      "learning_rate": 4.665727158899807e-06,
+      "loss": 0.7891,
+      "step": 30305
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.717858080434424,
+      "learning_rate": 4.664091831793432e-06,
+      "loss": 0.8242,
+      "step": 30310
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.591795154861806,
+      "learning_rate": 4.662456540781875e-06,
+      "loss": 0.8148,
+      "step": 30315
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.324604322295022,
+      "learning_rate": 4.66082128604085e-06,
+      "loss": 0.7701,
+      "step": 30320
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.807349808563159,
+      "learning_rate": 4.6591860677460765e-06,
+      "loss": 0.7871,
+      "step": 30325
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.379971415663782,
+      "learning_rate": 4.657550886073263e-06,
+      "loss": 0.7308,
+      "step": 30330
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.692979415457627,
+      "learning_rate": 4.6559157411981185e-06,
+      "loss": 0.8314,
+      "step": 30335
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.322567919421736,
+      "learning_rate": 4.6542806332963455e-06,
+      "loss": 0.783,
+      "step": 30340
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.504274834555929,
+      "learning_rate": 4.6526455625436436e-06,
+      "loss": 0.798,
+      "step": 30345
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.137756434797579,
+      "learning_rate": 4.651010529115712e-06,
+      "loss": 0.7444,
+      "step": 30350
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.048530234730052,
+      "learning_rate": 4.649375533188239e-06,
+      "loss": 0.8339,
+      "step": 30355
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.266714893016973,
+      "learning_rate": 4.6477405749369095e-06,
+      "loss": 0.7527,
+      "step": 30360
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.202683717523678,
+      "learning_rate": 4.646105654537412e-06,
+      "loss": 0.766,
+      "step": 30365
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.908417280112789,
+      "learning_rate": 4.644470772165422e-06,
+      "loss": 0.8348,
+      "step": 30370
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.92922707124643,
+      "learning_rate": 4.642835927996618e-06,
+      "loss": 0.7651,
+      "step": 30375
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.221855566146902,
+      "learning_rate": 4.641201122206669e-06,
+      "loss": 0.8098,
+      "step": 30380
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.945490575664971,
+      "learning_rate": 4.639566354971243e-06,
+      "loss": 0.7882,
+      "step": 30385
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.631437553930325,
+      "learning_rate": 4.637931626466004e-06,
+      "loss": 0.8199,
+      "step": 30390
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.748414024473478,
+      "learning_rate": 4.636296936866608e-06,
+      "loss": 0.8084,
+      "step": 30395
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.093336736761858,
+      "learning_rate": 4.634662286348713e-06,
+      "loss": 0.7911,
+      "step": 30400
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.89393963512712,
+      "learning_rate": 4.633027675087965e-06,
+      "loss": 0.7509,
+      "step": 30405
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.572528867939662,
+      "learning_rate": 4.6313931032600166e-06,
+      "loss": 0.7688,
+      "step": 30410
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.864991872589657,
+      "learning_rate": 4.629758571040505e-06,
+      "loss": 0.8021,
+      "step": 30415
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.761432934035888,
+      "learning_rate": 4.628124078605067e-06,
+      "loss": 0.7702,
+      "step": 30420
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.137330379569389,
+      "learning_rate": 4.626489626129339e-06,
+      "loss": 0.7672,
+      "step": 30425
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.002866723263471,
+      "learning_rate": 4.624855213788949e-06,
+      "loss": 0.794,
+      "step": 30430
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.807448760548121,
+      "learning_rate": 4.623220841759522e-06,
+      "loss": 0.7887,
+      "step": 30435
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.982114532743113,
+      "learning_rate": 4.621586510216677e-06,
+      "loss": 0.8064,
+      "step": 30440
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.19786776108365,
+      "learning_rate": 4.619952219336034e-06,
+      "loss": 0.7866,
+      "step": 30445
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.185672257390578,
+      "learning_rate": 4.618317969293203e-06,
+      "loss": 0.7995,
+      "step": 30450
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.677456015521814,
+      "learning_rate": 4.616683760263788e-06,
+      "loss": 0.7774,
+      "step": 30455
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.902543365821851,
+      "learning_rate": 4.615049592423396e-06,
+      "loss": 0.7847,
+      "step": 30460
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.054153778972065,
+      "learning_rate": 4.6134154659476234e-06,
+      "loss": 0.825,
+      "step": 30465
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.093642378000999,
+      "learning_rate": 4.6117813810120666e-06,
+      "loss": 0.7743,
+      "step": 30470
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 23.83979502928781,
+      "learning_rate": 4.610147337792313e-06,
+      "loss": 0.7427,
+      "step": 30475
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.785042569240462,
+      "learning_rate": 4.608513336463946e-06,
+      "loss": 0.848,
+      "step": 30480
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.86235805737726,
+      "learning_rate": 4.6068793772025535e-06,
+      "loss": 0.7708,
+      "step": 30485
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.12690200621559,
+      "learning_rate": 4.6052454601837025e-06,
+      "loss": 0.77,
+      "step": 30490
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.213617836410733,
+      "learning_rate": 4.603611585582972e-06,
+      "loss": 0.7688,
+      "step": 30495
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.9635968095081715,
+      "learning_rate": 4.601977753575927e-06,
+      "loss": 0.7621,
+      "step": 30500
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.520427801040054,
+      "learning_rate": 4.600343964338125e-06,
+      "loss": 0.7929,
+      "step": 30505
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.457308644107953,
+      "learning_rate": 4.598710218045131e-06,
+      "loss": 0.7794,
+      "step": 30510
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.932348994994766,
+      "learning_rate": 4.597076514872493e-06,
+      "loss": 0.8306,
+      "step": 30515
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.583724636856987,
+      "learning_rate": 4.595442854995762e-06,
+      "loss": 0.7993,
+      "step": 30520
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.618874416324449,
+      "learning_rate": 4.5938092385904826e-06,
+      "loss": 0.7803,
+      "step": 30525
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.633748954413597,
+      "learning_rate": 4.592175665832193e-06,
+      "loss": 0.7983,
+      "step": 30530
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.884533804620512,
+      "learning_rate": 4.590542136896428e-06,
+      "loss": 0.7917,
+      "step": 30535
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.802421179470391,
+      "learning_rate": 4.588908651958716e-06,
+      "loss": 0.758,
+      "step": 30540
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.870770880892959,
+      "learning_rate": 4.587275211194586e-06,
+      "loss": 0.7458,
+      "step": 30545
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.876659920577412,
+      "learning_rate": 4.585641814779554e-06,
+      "loss": 0.8006,
+      "step": 30550
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.575556345491801,
+      "learning_rate": 4.58400846288914e-06,
+      "loss": 0.8246,
+      "step": 30555
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.054746527256716,
+      "learning_rate": 4.5823751556988515e-06,
+      "loss": 0.7718,
+      "step": 30560
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.787702906973506,
+      "learning_rate": 4.580741893384195e-06,
+      "loss": 0.7374,
+      "step": 30565
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.496357982882074,
+      "learning_rate": 4.579108676120673e-06,
+      "loss": 0.7297,
+      "step": 30570
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.188715676796305,
+      "learning_rate": 4.57747550408378e-06,
+      "loss": 0.7705,
+      "step": 30575
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.96915600723857,
+      "learning_rate": 4.575842377449011e-06,
+      "loss": 0.7701,
+      "step": 30580
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.27471628944124,
+      "learning_rate": 4.57420929639185e-06,
+      "loss": 0.7919,
+      "step": 30585
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.954474080427729,
+      "learning_rate": 4.572576261087777e-06,
+      "loss": 0.8263,
+      "step": 30590
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.967339068825614,
+      "learning_rate": 4.570943271712274e-06,
+      "loss": 0.8065,
+      "step": 30595
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.95307628650065,
+      "learning_rate": 4.569310328440808e-06,
+      "loss": 0.7714,
+      "step": 30600
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.1858974515440295,
+      "learning_rate": 4.56767743144885e-06,
+      "loss": 0.7888,
+      "step": 30605
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.603289368621448,
+      "learning_rate": 4.566044580911857e-06,
+      "loss": 0.7874,
+      "step": 30610
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.212870890259156,
+      "learning_rate": 4.564411777005291e-06,
+      "loss": 0.7233,
+      "step": 30615
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.021863111667034,
+      "learning_rate": 4.562779019904603e-06,
+      "loss": 0.7685,
+      "step": 30620
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 13.57482857299841,
+      "learning_rate": 4.561146309785236e-06,
+      "loss": 0.743,
+      "step": 30625
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.933359595418446,
+      "learning_rate": 4.5595136468226374e-06,
+      "loss": 0.7689,
+      "step": 30630
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.644422904804832,
+      "learning_rate": 4.5578810311922396e-06,
+      "loss": 0.7854,
+      "step": 30635
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.985852582945892,
+      "learning_rate": 4.556248463069478e-06,
+      "loss": 0.7833,
+      "step": 30640
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.616850299489971,
+      "learning_rate": 4.5546159426297775e-06,
+      "loss": 0.7798,
+      "step": 30645
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.40064317403941,
+      "learning_rate": 4.552983470048558e-06,
+      "loss": 0.7806,
+      "step": 30650
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.304983524882694,
+      "learning_rate": 4.551351045501238e-06,
+      "loss": 0.7595,
+      "step": 30655
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.481968054889494,
+      "learning_rate": 4.549718669163228e-06,
+      "loss": 0.7718,
+      "step": 30660
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.860778228699741,
+      "learning_rate": 4.548086341209934e-06,
+      "loss": 0.7637,
+      "step": 30665
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.648316604195691,
+      "learning_rate": 4.546454061816759e-06,
+      "loss": 0.8252,
+      "step": 30670
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.1501342867187,
+      "learning_rate": 4.544821831159094e-06,
+      "loss": 0.7855,
+      "step": 30675
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.679821449145516,
+      "learning_rate": 4.543189649412335e-06,
+      "loss": 0.7714,
+      "step": 30680
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.658851702787414,
+      "learning_rate": 4.541557516751861e-06,
+      "loss": 0.8068,
+      "step": 30685
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.5092821750011645,
+      "learning_rate": 4.5399254333530575e-06,
+      "loss": 0.7554,
+      "step": 30690
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.18908964575066,
+      "learning_rate": 4.538293399391295e-06,
+      "loss": 0.77,
+      "step": 30695
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 19.764968967316946,
+      "learning_rate": 4.536661415041947e-06,
+      "loss": 0.8045,
+      "step": 30700
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.07756035863198,
+      "learning_rate": 4.535029480480373e-06,
+      "loss": 0.7665,
+      "step": 30705
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.869028974400354,
+      "learning_rate": 4.533397595881932e-06,
+      "loss": 0.7533,
+      "step": 30710
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 18.928500894149803,
+      "learning_rate": 4.531765761421981e-06,
+      "loss": 0.7466,
+      "step": 30715
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.129437668117854,
+      "learning_rate": 4.530133977275863e-06,
+      "loss": 0.7936,
+      "step": 30720
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.992181502887697,
+      "learning_rate": 4.528502243618927e-06,
+      "loss": 0.7721,
+      "step": 30725
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.03655983735585,
+      "learning_rate": 4.526870560626503e-06,
+      "loss": 0.7886,
+      "step": 30730
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.84818588439893,
+      "learning_rate": 4.525238928473924e-06,
+      "loss": 0.7785,
+      "step": 30735
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.044420929900029,
+      "learning_rate": 4.52360734733652e-06,
+      "loss": 0.768,
+      "step": 30740
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.70383824810081,
+      "learning_rate": 4.521975817389607e-06,
+      "loss": 0.8228,
+      "step": 30745
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.220618034970984,
+      "learning_rate": 4.520344338808503e-06,
+      "loss": 0.7752,
+      "step": 30750
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.439677389351129,
+      "learning_rate": 4.518712911768518e-06,
+      "loss": 0.7824,
+      "step": 30755
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.365446726376254,
+      "learning_rate": 4.517081536444953e-06,
+      "loss": 0.8092,
+      "step": 30760
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.978277276892899,
+      "learning_rate": 4.515450213013109e-06,
+      "loss": 0.8105,
+      "step": 30765
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.593006875897456,
+      "learning_rate": 4.5138189416482765e-06,
+      "loss": 0.759,
+      "step": 30770
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.87811568062077,
+      "learning_rate": 4.5121877225257465e-06,
+      "loss": 0.7763,
+      "step": 30775
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.072742428425823,
+      "learning_rate": 4.510556555820797e-06,
+      "loss": 0.7657,
+      "step": 30780
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.017659350867577,
+      "learning_rate": 4.508925441708707e-06,
+      "loss": 0.7766,
+      "step": 30785
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.291281139526044,
+      "learning_rate": 4.5072943803647455e-06,
+      "loss": 0.8064,
+      "step": 30790
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.205925312444233,
+      "learning_rate": 4.505663371964176e-06,
+      "loss": 0.7985,
+      "step": 30795
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.836958271735716,
+      "learning_rate": 4.50403241668226e-06,
+      "loss": 0.7831,
+      "step": 30800
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.205153316243075,
+      "learning_rate": 4.502401514694248e-06,
+      "loss": 0.7513,
+      "step": 30805
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 23.470695761410514,
+      "learning_rate": 4.500770666175391e-06,
+      "loss": 0.8096,
+      "step": 30810
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.630883665656526,
+      "learning_rate": 4.499139871300931e-06,
+      "loss": 0.7521,
+      "step": 30815
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.350718036561842,
+      "learning_rate": 4.497509130246098e-06,
+      "loss": 0.8045,
+      "step": 30820
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.879727503033959,
+      "learning_rate": 4.49587844318613e-06,
+      "loss": 0.7993,
+      "step": 30825
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.544375792731871,
+      "learning_rate": 4.494247810296245e-06,
+      "loss": 0.8047,
+      "step": 30830
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.72640514434046,
+      "learning_rate": 4.492617231751668e-06,
+      "loss": 0.8267,
+      "step": 30835
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.91031449738027,
+      "learning_rate": 4.490986707727607e-06,
+      "loss": 0.7463,
+      "step": 30840
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.68250037970916,
+      "learning_rate": 4.489356238399271e-06,
+      "loss": 0.8151,
+      "step": 30845
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.799596402429676,
+      "learning_rate": 4.487725823941861e-06,
+      "loss": 0.7721,
+      "step": 30850
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.655085216244254,
+      "learning_rate": 4.486095464530572e-06,
+      "loss": 0.7693,
+      "step": 30855
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.772520859631326,
+      "learning_rate": 4.4844651603405935e-06,
+      "loss": 0.7378,
+      "step": 30860
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.892676890718672,
+      "learning_rate": 4.482834911547108e-06,
+      "loss": 0.7642,
+      "step": 30865
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 4.921523137293614,
+      "learning_rate": 4.481204718325295e-06,
+      "loss": 0.7459,
+      "step": 30870
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.110797146339437,
+      "learning_rate": 4.479574580850325e-06,
+      "loss": 0.7745,
+      "step": 30875
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.715277224027325,
+      "learning_rate": 4.477944499297361e-06,
+      "loss": 0.7393,
+      "step": 30880
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.57049664161394,
+      "learning_rate": 4.4763144738415655e-06,
+      "loss": 0.7739,
+      "step": 30885
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.051762173851264,
+      "learning_rate": 4.474684504658091e-06,
+      "loss": 0.7689,
+      "step": 30890
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.070437940829969,
+      "learning_rate": 4.473054591922085e-06,
+      "loss": 0.782,
+      "step": 30895
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.6936418351549305,
+      "learning_rate": 4.471424735808689e-06,
+      "loss": 0.7977,
+      "step": 30900
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.944237077514344,
+      "learning_rate": 4.469794936493037e-06,
+      "loss": 0.7659,
+      "step": 30905
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.0988930121814,
+      "learning_rate": 4.46816519415026e-06,
+      "loss": 0.8011,
+      "step": 30910
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.813932156741568,
+      "learning_rate": 4.4665355089554786e-06,
+      "loss": 0.8182,
+      "step": 30915
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.032759838191125,
+      "learning_rate": 4.464905881083813e-06,
+      "loss": 0.788,
+      "step": 30920
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.10155654552586,
+      "learning_rate": 4.46327631071037e-06,
+      "loss": 0.781,
+      "step": 30925
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.770885476095118,
+      "learning_rate": 4.461646798010259e-06,
+      "loss": 0.7785,
+      "step": 30930
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.366305357005817,
+      "learning_rate": 4.4600173431585734e-06,
+      "loss": 0.7825,
+      "step": 30935
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.0720709688726275,
+      "learning_rate": 4.458387946330406e-06,
+      "loss": 0.7828,
+      "step": 30940
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.104130471865807,
+      "learning_rate": 4.4567586077008475e-06,
+      "loss": 0.8041,
+      "step": 30945
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.382667558508729,
+      "learning_rate": 4.455129327444971e-06,
+      "loss": 0.7559,
+      "step": 30950
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.282227207214767,
+      "learning_rate": 4.453500105737856e-06,
+      "loss": 0.7512,
+      "step": 30955
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 4.630416905290494,
+      "learning_rate": 4.451870942754566e-06,
+      "loss": 0.7707,
+      "step": 30960
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.823400173313469,
+      "learning_rate": 4.4502418386701595e-06,
+      "loss": 0.7944,
+      "step": 30965
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.140146188727773,
+      "learning_rate": 4.4486127936596966e-06,
+      "loss": 0.8237,
+      "step": 30970
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.267334861579995,
+      "learning_rate": 4.44698380789822e-06,
+      "loss": 0.7765,
+      "step": 30975
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.091711751004377,
+      "learning_rate": 4.4453548815607756e-06,
+      "loss": 0.7869,
+      "step": 30980
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.673789695748567,
+      "learning_rate": 4.443726014822398e-06,
+      "loss": 0.7586,
+      "step": 30985
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.978048564926374,
+      "learning_rate": 4.442097207858113e-06,
+      "loss": 0.7226,
+      "step": 30990
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.043454928860996,
+      "learning_rate": 4.4404684608429465e-06,
+      "loss": 0.7785,
+      "step": 30995
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.780883133711187,
+      "learning_rate": 4.438839773951912e-06,
+      "loss": 0.8075,
+      "step": 31000
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.869682890959952,
+      "learning_rate": 4.4372111473600216e-06,
+      "loss": 0.8423,
+      "step": 31005
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 14.382298328227915,
+      "learning_rate": 4.435582581242276e-06,
+      "loss": 0.7843,
+      "step": 31010
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.298800869774029,
+      "learning_rate": 4.433954075773675e-06,
+      "loss": 0.7579,
+      "step": 31015
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.500762773898244,
+      "learning_rate": 4.432325631129205e-06,
+      "loss": 0.7621,
+      "step": 31020
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.278527125856053,
+      "learning_rate": 4.430697247483852e-06,
+      "loss": 0.7504,
+      "step": 31025
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.789904700356915,
+      "learning_rate": 4.4290689250125915e-06,
+      "loss": 0.7909,
+      "step": 31030
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.029927064891336,
+      "learning_rate": 4.427440663890394e-06,
+      "loss": 0.7907,
+      "step": 31035
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.080109042755826,
+      "learning_rate": 4.425812464292226e-06,
+      "loss": 0.8349,
+      "step": 31040
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 23.86416052534052,
+      "learning_rate": 4.424184326393044e-06,
+      "loss": 0.7762,
+      "step": 31045
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.087622115151375,
+      "learning_rate": 4.422556250367794e-06,
+      "loss": 0.7698,
+      "step": 31050
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.38325541238791,
+      "learning_rate": 4.420928236391426e-06,
+      "loss": 0.7936,
+      "step": 31055
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.691124138180824,
+      "learning_rate": 4.419300284638872e-06,
+      "loss": 0.7878,
+      "step": 31060
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.738360449067118,
+      "learning_rate": 4.417672395285068e-06,
+      "loss": 0.7893,
+      "step": 31065
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.551122192832034,
+      "learning_rate": 4.416044568504934e-06,
+      "loss": 0.7862,
+      "step": 31070
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.612648392437366,
+      "learning_rate": 4.414416804473387e-06,
+      "loss": 0.7469,
+      "step": 31075
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.452026622214422,
+      "learning_rate": 4.412789103365341e-06,
+      "loss": 0.7527,
+      "step": 31080
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.494616714621361,
+      "learning_rate": 4.411161465355695e-06,
+      "loss": 0.7527,
+      "step": 31085
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.440267287183166,
+      "learning_rate": 4.409533890619351e-06,
+      "loss": 0.7898,
+      "step": 31090
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.065609967411705,
+      "learning_rate": 4.407906379331193e-06,
+      "loss": 0.7662,
+      "step": 31095
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.339445680539315,
+      "learning_rate": 4.4062789316661105e-06,
+      "loss": 0.772,
+      "step": 31100
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.746744948171015,
+      "learning_rate": 4.4046515477989774e-06,
+      "loss": 0.7662,
+      "step": 31105
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.6492187131747675,
+      "learning_rate": 4.403024227904661e-06,
+      "loss": 0.7758,
+      "step": 31110
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.544339588254074,
+      "learning_rate": 4.401396972158027e-06,
+      "loss": 0.8084,
+      "step": 31115
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.785846514801537,
+      "learning_rate": 4.399769780733929e-06,
+      "loss": 0.7735,
+      "step": 31120
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.752272418122944,
+      "learning_rate": 4.398142653807219e-06,
+      "loss": 0.7645,
+      "step": 31125
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.285973621470024,
+      "learning_rate": 4.396515591552737e-06,
+      "loss": 0.7613,
+      "step": 31130
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.805614312727597,
+      "learning_rate": 4.394888594145315e-06,
+      "loss": 0.8157,
+      "step": 31135
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.420765425722168,
+      "learning_rate": 4.393261661759788e-06,
+      "loss": 0.7925,
+      "step": 31140
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 18.84362929653387,
+      "learning_rate": 4.39163479457097e-06,
+      "loss": 0.7682,
+      "step": 31145
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.936863052636425,
+      "learning_rate": 4.390007992753681e-06,
+      "loss": 0.7413,
+      "step": 31150
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.861988183192306,
+      "learning_rate": 4.388381256482724e-06,
+      "loss": 0.7679,
+      "step": 31155
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.59897921077883,
+      "learning_rate": 4.3867545859329e-06,
+      "loss": 0.7404,
+      "step": 31160
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.008417029745514,
+      "learning_rate": 4.385127981279002e-06,
+      "loss": 0.7653,
+      "step": 31165
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.941394647427479,
+      "learning_rate": 4.383501442695816e-06,
+      "loss": 0.742,
+      "step": 31170
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 17.08816107208935,
+      "learning_rate": 4.381874970358122e-06,
+      "loss": 0.7713,
+      "step": 31175
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.618820013739727,
+      "learning_rate": 4.380248564440688e-06,
+      "loss": 0.7792,
+      "step": 31180
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.683833882800295,
+      "learning_rate": 4.378622225118284e-06,
+      "loss": 0.7603,
+      "step": 31185
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.099153329246707,
+      "learning_rate": 4.376995952565664e-06,
+      "loss": 0.7828,
+      "step": 31190
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.300304520294611,
+      "learning_rate": 4.3753697469575765e-06,
+      "loss": 0.7546,
+      "step": 31195
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.326420135609376,
+      "learning_rate": 4.373743608468768e-06,
+      "loss": 0.7401,
+      "step": 31200
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.875584867433873,
+      "learning_rate": 4.372117537273971e-06,
+      "loss": 0.7677,
+      "step": 31205
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.972293807215018,
+      "learning_rate": 4.3704915335479156e-06,
+      "loss": 0.7481,
+      "step": 31210
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.580515434200049,
+      "learning_rate": 4.3688655974653254e-06,
+      "loss": 0.7869,
+      "step": 31215
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.886669050135439,
+      "learning_rate": 4.367239729200909e-06,
+      "loss": 0.7538,
+      "step": 31220
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.5301429470913055,
+      "learning_rate": 4.36561392892938e-06,
+      "loss": 0.7447,
+      "step": 31225
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.048714691653655,
+      "learning_rate": 4.363988196825431e-06,
+      "loss": 0.7584,
+      "step": 31230
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.978489615631539,
+      "learning_rate": 4.362362533063759e-06,
+      "loss": 0.7716,
+      "step": 31235
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.246958008123747,
+      "learning_rate": 4.360736937819048e-06,
+      "loss": 0.7752,
+      "step": 31240
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.179731498250461,
+      "learning_rate": 4.359111411265971e-06,
+      "loss": 0.783,
+      "step": 31245
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.243791979682189,
+      "learning_rate": 4.357485953579203e-06,
+      "loss": 0.7304,
+      "step": 31250
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.291653960712617,
+      "learning_rate": 4.3558605649334055e-06,
+      "loss": 0.7588,
+      "step": 31255
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.915867816626169,
+      "learning_rate": 4.354235245503232e-06,
+      "loss": 0.7684,
+      "step": 31260
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.745634037995895,
+      "learning_rate": 4.352609995463331e-06,
+      "loss": 0.7692,
+      "step": 31265
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.857619438194422,
+      "learning_rate": 4.350984814988344e-06,
+      "loss": 0.7721,
+      "step": 31270
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 20.924372231052487,
+      "learning_rate": 4.349359704252904e-06,
+      "loss": 0.7709,
+      "step": 31275
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 17.05477162542986,
+      "learning_rate": 4.347734663431633e-06,
+      "loss": 0.7975,
+      "step": 31280
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.919630400985666,
+      "learning_rate": 4.3461096926991535e-06,
+      "loss": 0.7838,
+      "step": 31285
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.775409566108691,
+      "learning_rate": 4.3444847922300715e-06,
+      "loss": 0.7536,
+      "step": 31290
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 15.137755669208488,
+      "learning_rate": 4.342859962198993e-06,
+      "loss": 0.75,
+      "step": 31295
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.083390967951964,
+      "learning_rate": 4.341235202780511e-06,
+      "loss": 0.7942,
+      "step": 31300
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.314667301433014,
+      "learning_rate": 4.339610514149212e-06,
+      "loss": 0.7906,
+      "step": 31305
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.78234183383961,
+      "learning_rate": 4.337985896479681e-06,
+      "loss": 0.777,
+      "step": 31310
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.542191665357444,
+      "learning_rate": 4.336361349946485e-06,
+      "loss": 0.8141,
+      "step": 31315
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 16.02455848399195,
+      "learning_rate": 4.334736874724192e-06,
+      "loss": 0.7968,
+      "step": 31320
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.411196553036113,
+      "learning_rate": 4.3331124709873564e-06,
+      "loss": 0.7965,
+      "step": 31325
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.814419711407543,
+      "learning_rate": 4.331488138910531e-06,
+      "loss": 0.7867,
+      "step": 31330
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.251291883369472,
+      "learning_rate": 4.329863878668255e-06,
+      "loss": 0.7708,
+      "step": 31335
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 25.343195757766704,
+      "learning_rate": 4.328239690435061e-06,
+      "loss": 0.7544,
+      "step": 31340
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.396865937727929,
+      "learning_rate": 4.326615574385477e-06,
+      "loss": 0.7425,
+      "step": 31345
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.548083644899822,
+      "learning_rate": 4.324991530694023e-06,
+      "loss": 0.7863,
+      "step": 31350
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.892588628574988,
+      "learning_rate": 4.323367559535208e-06,
+      "loss": 0.8153,
+      "step": 31355
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.445878857866565,
+      "learning_rate": 4.321743661083535e-06,
+      "loss": 0.7565,
+      "step": 31360
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.8100511915903486,
+      "learning_rate": 4.320119835513498e-06,
+      "loss": 0.7827,
+      "step": 31365
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.39383204652131,
+      "learning_rate": 4.318496082999586e-06,
+      "loss": 0.7498,
+      "step": 31370
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.997854485220208,
+      "learning_rate": 4.3168724037162765e-06,
+      "loss": 0.793,
+      "step": 31375
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.213261918647128,
+      "learning_rate": 4.315248797838044e-06,
+      "loss": 0.795,
+      "step": 31380
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.120675960037323,
+      "learning_rate": 4.313625265539349e-06,
+      "loss": 0.7762,
+      "step": 31385
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.35815778137239,
+      "learning_rate": 4.312001806994648e-06,
+      "loss": 0.7231,
+      "step": 31390
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.5415263799436385,
+      "learning_rate": 4.31037842237839e-06,
+      "loss": 0.7535,
+      "step": 31395
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.395032082967594,
+      "learning_rate": 4.3087551118650125e-06,
+      "loss": 0.7911,
+      "step": 31400
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.655500905307801,
+      "learning_rate": 4.307131875628951e-06,
+      "loss": 0.7518,
+      "step": 31405
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.283945441297352,
+      "learning_rate": 4.305508713844626e-06,
+      "loss": 0.7708,
+      "step": 31410
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.225466470945415,
+      "learning_rate": 4.3038856266864556e-06,
+      "loss": 0.7566,
+      "step": 31415
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.5559474300909475,
+      "learning_rate": 4.302262614328847e-06,
+      "loss": 0.7392,
+      "step": 31420
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.88652884886417,
+      "learning_rate": 4.300639676946199e-06,
+      "loss": 0.7742,
+      "step": 31425
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.239389019150582,
+      "learning_rate": 4.299016814712905e-06,
+      "loss": 0.7435,
+      "step": 31430
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.925998101904268,
+      "learning_rate": 4.2973940278033466e-06,
+      "loss": 0.7206,
+      "step": 31435
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.790237032822464,
+      "learning_rate": 4.295771316391901e-06,
+      "loss": 0.7315,
+      "step": 31440
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.876794687580935,
+      "learning_rate": 4.2941486806529375e-06,
+      "loss": 0.716,
+      "step": 31445
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.942558109633507,
+      "learning_rate": 4.292526120760811e-06,
+      "loss": 0.7528,
+      "step": 31450
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.385608527512987,
+      "learning_rate": 4.2909036368898775e-06,
+      "loss": 0.7606,
+      "step": 31455
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.041549066237203,
+      "learning_rate": 4.289281229214475e-06,
+      "loss": 0.8232,
+      "step": 31460
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.097921714615842,
+      "learning_rate": 4.2876588979089425e-06,
+      "loss": 0.7271,
+      "step": 31465
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.035814420161072,
+      "learning_rate": 4.286036643147607e-06,
+      "loss": 0.784,
+      "step": 31470
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.912060853703544,
+      "learning_rate": 4.284414465104782e-06,
+      "loss": 0.7881,
+      "step": 31475
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 17.53554338247057,
+      "learning_rate": 4.282792363954782e-06,
+      "loss": 0.7949,
+      "step": 31480
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.127032572018969,
+      "learning_rate": 4.281170339871908e-06,
+      "loss": 0.7658,
+      "step": 31485
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.656023041910718,
+      "learning_rate": 4.2795483930304525e-06,
+      "loss": 0.7279,
+      "step": 31490
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.3379778108073594,
+      "learning_rate": 4.277926523604702e-06,
+      "loss": 0.7454,
+      "step": 31495
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.628433221272852,
+      "learning_rate": 4.276304731768934e-06,
+      "loss": 0.789,
+      "step": 31500
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.739220700964836,
+      "learning_rate": 4.274683017697418e-06,
+      "loss": 0.7568,
+      "step": 31505
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.022702406494936,
+      "learning_rate": 4.27306138156441e-06,
+      "loss": 0.7864,
+      "step": 31510
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.801256168210536,
+      "learning_rate": 4.271439823544167e-06,
+      "loss": 0.7255,
+      "step": 31515
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.218142837687559,
+      "learning_rate": 4.269818343810928e-06,
+      "loss": 0.7971,
+      "step": 31520
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.722214871650178,
+      "learning_rate": 4.2681969425389325e-06,
+      "loss": 0.7171,
+      "step": 31525
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.350522850546013,
+      "learning_rate": 4.266575619902404e-06,
+      "loss": 0.7584,
+      "step": 31530
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.80285127418216,
+      "learning_rate": 4.264954376075562e-06,
+      "loss": 0.8044,
+      "step": 31535
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.624145849686642,
+      "learning_rate": 4.263333211232618e-06,
+      "loss": 0.7942,
+      "step": 31540
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.101323984067481,
+      "learning_rate": 4.261712125547769e-06,
+      "loss": 0.7647,
+      "step": 31545
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.033069267818239,
+      "learning_rate": 4.260091119195212e-06,
+      "loss": 0.7906,
+      "step": 31550
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.1912348970513715,
+      "learning_rate": 4.25847019234913e-06,
+      "loss": 0.7714,
+      "step": 31555
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.752930101317553,
+      "learning_rate": 4.256849345183698e-06,
+      "loss": 0.7395,
+      "step": 31560
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 21.63767299952961,
+      "learning_rate": 4.255228577873084e-06,
+      "loss": 0.7129,
+      "step": 31565
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.792642421828407,
+      "learning_rate": 4.253607890591445e-06,
+      "loss": 0.7597,
+      "step": 31570
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.253362064768006,
+      "learning_rate": 4.2519872835129325e-06,
+      "loss": 0.7456,
+      "step": 31575
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.066988823586071,
+      "learning_rate": 4.250366756811688e-06,
+      "loss": 0.771,
+      "step": 31580
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.51477458227472,
+      "learning_rate": 4.248746310661845e-06,
+      "loss": 0.7672,
+      "step": 31585
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.071531395847083,
+      "learning_rate": 4.247125945237526e-06,
+      "loss": 0.7474,
+      "step": 31590
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.23186061281606,
+      "learning_rate": 4.245505660712844e-06,
+      "loss": 0.7482,
+      "step": 31595
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.865601469920776,
+      "learning_rate": 4.243885457261913e-06,
+      "loss": 0.7795,
+      "step": 31600
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.505457591036943,
+      "learning_rate": 4.242265335058823e-06,
+      "loss": 0.7553,
+      "step": 31605
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.736200050014133,
+      "learning_rate": 4.24064529427767e-06,
+      "loss": 0.8207,
+      "step": 31610
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.936635635092902,
+      "learning_rate": 4.2390253350925295e-06,
+      "loss": 0.7855,
+      "step": 31615
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.733425246430409,
+      "learning_rate": 4.237405457677476e-06,
+      "loss": 0.7755,
+      "step": 31620
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.201225649268791,
+      "learning_rate": 4.235785662206571e-06,
+      "loss": 0.7493,
+      "step": 31625
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 14.019078487520051,
+      "learning_rate": 4.2341659488538674e-06,
+      "loss": 0.7154,
+      "step": 31630
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.258473765302222,
+      "learning_rate": 4.2325463177934165e-06,
+      "loss": 0.7571,
+      "step": 31635
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.124520434246618,
+      "learning_rate": 4.23092676919925e-06,
+      "loss": 0.7879,
+      "step": 31640
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.297941820023116,
+      "learning_rate": 4.229307303245393e-06,
+      "loss": 0.7946,
+      "step": 31645
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.095887150025404,
+      "learning_rate": 4.227687920105871e-06,
+      "loss": 0.7504,
+      "step": 31650
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.817195496141892,
+      "learning_rate": 4.226068619954688e-06,
+      "loss": 0.7546,
+      "step": 31655
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.720075996295372,
+      "learning_rate": 4.22444940296585e-06,
+      "loss": 0.7872,
+      "step": 31660
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.413683138621276,
+      "learning_rate": 4.222830269313344e-06,
+      "loss": 0.7135,
+      "step": 31665
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.4658339719402305,
+      "learning_rate": 4.221211219171156e-06,
+      "loss": 0.7615,
+      "step": 31670
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.810946550099423,
+      "learning_rate": 4.2195922527132615e-06,
+      "loss": 0.8076,
+      "step": 31675
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.951481803182469,
+      "learning_rate": 4.217973370113622e-06,
+      "loss": 0.7608,
+      "step": 31680
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.847401075071992,
+      "learning_rate": 4.216354571546195e-06,
+      "loss": 0.7598,
+      "step": 31685
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.7501901210612845,
+      "learning_rate": 4.214735857184928e-06,
+      "loss": 0.745,
+      "step": 31690
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.958178236177581,
+      "learning_rate": 4.21311722720376e-06,
+      "loss": 0.7506,
+      "step": 31695
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.857799563866652,
+      "learning_rate": 4.211498681776619e-06,
+      "loss": 0.7283,
+      "step": 31700
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.835098289625643,
+      "learning_rate": 4.209880221077422e-06,
+      "loss": 0.7535,
+      "step": 31705
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.266675177355465,
+      "learning_rate": 4.208261845280084e-06,
+      "loss": 0.7948,
+      "step": 31710
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.742704098505882,
+      "learning_rate": 4.206643554558504e-06,
+      "loss": 0.777,
+      "step": 31715
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.663389553754389,
+      "learning_rate": 4.205025349086576e-06,
+      "loss": 0.7606,
+      "step": 31720
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.655969345425756,
+      "learning_rate": 4.2034072290381835e-06,
+      "loss": 0.7612,
+      "step": 31725
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.77392702473303,
+      "learning_rate": 4.201789194587198e-06,
+      "loss": 0.7682,
+      "step": 31730
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.160671480717663,
+      "learning_rate": 4.200171245907488e-06,
+      "loss": 0.7027,
+      "step": 31735
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 22.467000756344046,
+      "learning_rate": 4.1985533831729064e-06,
+      "loss": 0.7651,
+      "step": 31740
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.860220719417435,
+      "learning_rate": 4.196935606557302e-06,
+      "loss": 0.751,
+      "step": 31745
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.693644957629103,
+      "learning_rate": 4.195317916234509e-06,
+      "loss": 0.764,
+      "step": 31750
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.04868919779745,
+      "learning_rate": 4.19370031237836e-06,
+      "loss": 0.7589,
+      "step": 31755
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.029436333630056,
+      "learning_rate": 4.192082795162669e-06,
+      "loss": 0.7573,
+      "step": 31760
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.80530077768993,
+      "learning_rate": 4.190465364761247e-06,
+      "loss": 0.7856,
+      "step": 31765
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.832406430152098,
+      "learning_rate": 4.188848021347895e-06,
+      "loss": 0.758,
+      "step": 31770
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.94655597758038,
+      "learning_rate": 4.1872307650964025e-06,
+      "loss": 0.7592,
+      "step": 31775
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.648696778511196,
+      "learning_rate": 4.185613596180553e-06,
+      "loss": 0.7683,
+      "step": 31780
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.204299199503271,
+      "learning_rate": 4.183996514774116e-06,
+      "loss": 0.7188,
+      "step": 31785
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.507753091151947,
+      "learning_rate": 4.182379521050853e-06,
+      "loss": 0.7693,
+      "step": 31790
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.454095096314965,
+      "learning_rate": 4.180762615184521e-06,
+      "loss": 0.7189,
+      "step": 31795
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.288290058462045,
+      "learning_rate": 4.17914579734886e-06,
+      "loss": 0.722,
+      "step": 31800
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.642839848928702,
+      "learning_rate": 4.177529067717607e-06,
+      "loss": 0.7602,
+      "step": 31805
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.890058280333357,
+      "learning_rate": 4.175912426464485e-06,
+      "loss": 0.7351,
+      "step": 31810
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.571984915838518,
+      "learning_rate": 4.1742958737632095e-06,
+      "loss": 0.8144,
+      "step": 31815
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.200743083525845,
+      "learning_rate": 4.172679409787488e-06,
+      "loss": 0.7529,
+      "step": 31820
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.3007219715715586,
+      "learning_rate": 4.171063034711012e-06,
+      "loss": 0.7663,
+      "step": 31825
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.05394089101257,
+      "learning_rate": 4.169446748707475e-06,
+      "loss": 0.7661,
+      "step": 31830
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.6760887125767425,
+      "learning_rate": 4.167830551950548e-06,
+      "loss": 0.7724,
+      "step": 31835
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.962089281564885,
+      "learning_rate": 4.166214444613902e-06,
+      "loss": 0.7636,
+      "step": 31840
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.12419493624849,
+      "learning_rate": 4.164598426871193e-06,
+      "loss": 0.7596,
+      "step": 31845
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.593275584109431,
+      "learning_rate": 4.16298249889607e-06,
+      "loss": 0.7726,
+      "step": 31850
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.8467002907292915,
+      "learning_rate": 4.161366660862171e-06,
+      "loss": 0.7463,
+      "step": 31855
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.195673443419892,
+      "learning_rate": 4.159750912943125e-06,
+      "loss": 0.7892,
+      "step": 31860
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.188407050668656,
+      "learning_rate": 4.158135255312554e-06,
+      "loss": 0.7294,
+      "step": 31865
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.178572546014093,
+      "learning_rate": 4.156519688144065e-06,
+      "loss": 0.7363,
+      "step": 31870
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.89703040686469,
+      "learning_rate": 4.154904211611256e-06,
+      "loss": 0.7305,
+      "step": 31875
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.000373430155154,
+      "learning_rate": 4.153288825887721e-06,
+      "loss": 0.7757,
+      "step": 31880
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.363742379551147,
+      "learning_rate": 4.151673531147037e-06,
+      "loss": 0.7347,
+      "step": 31885
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.149094319391997,
+      "learning_rate": 4.1500583275627784e-06,
+      "loss": 0.7516,
+      "step": 31890
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.262698444911576,
+      "learning_rate": 4.1484432153085e-06,
+      "loss": 0.7484,
+      "step": 31895
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.415766897917479,
+      "learning_rate": 4.146828194557758e-06,
+      "loss": 0.7656,
+      "step": 31900
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.81800745427543,
+      "learning_rate": 4.145213265484094e-06,
+      "loss": 0.7295,
+      "step": 31905
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.304317114743605,
+      "learning_rate": 4.143598428261034e-06,
+      "loss": 0.7087,
+      "step": 31910
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.1246612257695965,
+      "learning_rate": 4.141983683062104e-06,
+      "loss": 0.7485,
+      "step": 31915
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.959453601082129,
+      "learning_rate": 4.140369030060813e-06,
+      "loss": 0.7555,
+      "step": 31920
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.836317126015636,
+      "learning_rate": 4.1387544694306646e-06,
+      "loss": 0.7428,
+      "step": 31925
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.812818822265653,
+      "learning_rate": 4.13714000134515e-06,
+      "loss": 0.8003,
+      "step": 31930
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.2243360596537975,
+      "learning_rate": 4.1355256259777485e-06,
+      "loss": 0.756,
+      "step": 31935
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.630192128442654,
+      "learning_rate": 4.1339113435019345e-06,
+      "loss": 0.7846,
+      "step": 31940
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.166171758026652,
+      "learning_rate": 4.132297154091169e-06,
+      "loss": 0.7431,
+      "step": 31945
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.57799394956431,
+      "learning_rate": 4.130683057918904e-06,
+      "loss": 0.7716,
+      "step": 31950
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.935452628773687,
+      "learning_rate": 4.129069055158582e-06,
+      "loss": 0.8137,
+      "step": 31955
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.891814026575228,
+      "learning_rate": 4.127455145983631e-06,
+      "loss": 0.7307,
+      "step": 31960
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.038499061185585,
+      "learning_rate": 4.125841330567478e-06,
+      "loss": 0.7278,
+      "step": 31965
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.284721168296819,
+      "learning_rate": 4.12422760908353e-06,
+      "loss": 0.7421,
+      "step": 31970
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.111371571582586,
+      "learning_rate": 4.122613981705192e-06,
+      "loss": 0.7344,
+      "step": 31975
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.60813408911817,
+      "learning_rate": 4.121000448605852e-06,
+      "loss": 0.7979,
+      "step": 31980
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 23.050318204325972,
+      "learning_rate": 4.119387009958896e-06,
+      "loss": 0.7527,
+      "step": 31985
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.453800733158293,
+      "learning_rate": 4.11777366593769e-06,
+      "loss": 0.7193,
+      "step": 31990
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.929249730939494,
+      "learning_rate": 4.116160416715597e-06,
+      "loss": 0.766,
+      "step": 31995
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.514173746917727,
+      "learning_rate": 4.11454726246597e-06,
+      "loss": 0.7525,
+      "step": 32000
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.732700795478207,
+      "learning_rate": 4.112934203362145e-06,
+      "loss": 0.7508,
+      "step": 32005
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.244566091873192,
+      "learning_rate": 4.1113212395774575e-06,
+      "loss": 0.7686,
+      "step": 32010
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 15.900792083792036,
+      "learning_rate": 4.109708371285224e-06,
+      "loss": 0.7596,
+      "step": 32015
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.53746529168734,
+      "learning_rate": 4.108095598658753e-06,
+      "loss": 0.7583,
+      "step": 32020
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.774204067156112,
+      "learning_rate": 4.1064829218713485e-06,
+      "loss": 0.7658,
+      "step": 32025
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.529234728684587,
+      "learning_rate": 4.104870341096295e-06,
+      "loss": 0.7511,
+      "step": 32030
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.594346783118083,
+      "learning_rate": 4.103257856506876e-06,
+      "loss": 0.7331,
+      "step": 32035
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.646534008548479,
+      "learning_rate": 4.101645468276358e-06,
+      "loss": 0.7753,
+      "step": 32040
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.008223688407055,
+      "learning_rate": 4.100033176577998e-06,
+      "loss": 0.7485,
+      "step": 32045
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.050050521919625,
+      "learning_rate": 4.098420981585046e-06,
+      "loss": 0.7368,
+      "step": 32050
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.798072441024582,
+      "learning_rate": 4.096808883470737e-06,
+      "loss": 0.7529,
+      "step": 32055
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.027194373653681,
+      "learning_rate": 4.095196882408302e-06,
+      "loss": 0.7647,
+      "step": 32060
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.869935624755382,
+      "learning_rate": 4.093584978570952e-06,
+      "loss": 0.7633,
+      "step": 32065
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.831572464823065,
+      "learning_rate": 4.0919731721319e-06,
+      "loss": 0.7689,
+      "step": 32070
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.366645756409165,
+      "learning_rate": 4.090361463264336e-06,
+      "loss": 0.7306,
+      "step": 32075
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.910519208879556,
+      "learning_rate": 4.088749852141447e-06,
+      "loss": 0.7194,
+      "step": 32080
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.941402669165022,
+      "learning_rate": 4.087138338936409e-06,
+      "loss": 0.7607,
+      "step": 32085
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.276830228065567,
+      "learning_rate": 4.085526923822383e-06,
+      "loss": 0.7706,
+      "step": 32090
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.313920519669926,
+      "learning_rate": 4.0839156069725286e-06,
+      "loss": 0.7602,
+      "step": 32095
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.744053032382729,
+      "learning_rate": 4.082304388559984e-06,
+      "loss": 0.77,
+      "step": 32100
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.480837100740973,
+      "learning_rate": 4.080693268757881e-06,
+      "loss": 0.7392,
+      "step": 32105
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.434959462726063,
+      "learning_rate": 4.079082247739346e-06,
+      "loss": 0.7386,
+      "step": 32110
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.855344306413238,
+      "learning_rate": 4.077471325677485e-06,
+      "loss": 0.791,
+      "step": 32115
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.09501038994656,
+      "learning_rate": 4.0758605027454045e-06,
+      "loss": 0.7569,
+      "step": 32120
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.245371001691915,
+      "learning_rate": 4.07424977911619e-06,
+      "loss": 0.7775,
+      "step": 32125
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.365929182518692,
+      "learning_rate": 4.072639154962922e-06,
+      "loss": 0.7653,
+      "step": 32130
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.027473029988867,
+      "learning_rate": 4.071028630458671e-06,
+      "loss": 0.7751,
+      "step": 32135
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.575579205576042,
+      "learning_rate": 4.069418205776491e-06,
+      "loss": 0.7451,
+      "step": 32140
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.401146694741413,
+      "learning_rate": 4.067807881089434e-06,
+      "loss": 0.7245,
+      "step": 32145
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.96988948956007,
+      "learning_rate": 4.066197656570533e-06,
+      "loss": 0.7655,
+      "step": 32150
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.77269771981435,
+      "learning_rate": 4.064587532392817e-06,
+      "loss": 0.7382,
+      "step": 32155
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.637743519235023,
+      "learning_rate": 4.062977508729298e-06,
+      "loss": 0.7631,
+      "step": 32160
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.616435366436222,
+      "learning_rate": 4.0613675857529795e-06,
+      "loss": 0.7532,
+      "step": 32165
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.412442146282715,
+      "learning_rate": 4.059757763636858e-06,
+      "loss": 0.7628,
+      "step": 32170
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.549725817186147,
+      "learning_rate": 4.0581480425539135e-06,
+      "loss": 0.7391,
+      "step": 32175
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.089406068427861,
+      "learning_rate": 4.056538422677119e-06,
+      "loss": 0.7484,
+      "step": 32180
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.697875326602824,
+      "learning_rate": 4.054928904179435e-06,
+      "loss": 0.7399,
+      "step": 32185
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.837444253034605,
+      "learning_rate": 4.05331948723381e-06,
+      "loss": 0.7692,
+      "step": 32190
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.681669853385758,
+      "learning_rate": 4.051710172013186e-06,
+      "loss": 0.7832,
+      "step": 32195
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.878806670156813,
+      "learning_rate": 4.050100958690487e-06,
+      "loss": 0.7259,
+      "step": 32200
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.740676691401125,
+      "learning_rate": 4.048491847438634e-06,
+      "loss": 0.7078,
+      "step": 32205
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.002376127598739,
+      "learning_rate": 4.0468828384305316e-06,
+      "loss": 0.7415,
+      "step": 32210
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.982992071388394,
+      "learning_rate": 4.045273931839073e-06,
+      "loss": 0.7274,
+      "step": 32215
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.126164475039666,
+      "learning_rate": 4.043665127837145e-06,
+      "loss": 0.741,
+      "step": 32220
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.459062382502498,
+      "learning_rate": 4.042056426597618e-06,
+      "loss": 0.7541,
+      "step": 32225
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.096543727090931,
+      "learning_rate": 4.040447828293359e-06,
+      "loss": 0.737,
+      "step": 32230
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.059764483812855,
+      "learning_rate": 4.038839333097212e-06,
+      "loss": 0.7343,
+      "step": 32235
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.2378777671462196,
+      "learning_rate": 4.0372309411820235e-06,
+      "loss": 0.7547,
+      "step": 32240
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 108.92030761273968,
+      "learning_rate": 4.035622652720619e-06,
+      "loss": 0.7671,
+      "step": 32245
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 43.745775882897625,
+      "learning_rate": 4.034014467885814e-06,
+      "loss": 0.7815,
+      "step": 32250
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 49.83487826438429,
+      "learning_rate": 4.032406386850421e-06,
+      "loss": 0.7428,
+      "step": 32255
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 24.095843495208634,
+      "learning_rate": 4.030798409787229e-06,
+      "loss": 0.7578,
+      "step": 32260
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.34996733573595,
+      "learning_rate": 4.029190536869027e-06,
+      "loss": 0.7143,
+      "step": 32265
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.93765065178245,
+      "learning_rate": 4.027582768268587e-06,
+      "loss": 0.7293,
+      "step": 32270
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.382284251542114,
+      "learning_rate": 4.025975104158668e-06,
+      "loss": 0.7425,
+      "step": 32275
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.361834811889404,
+      "learning_rate": 4.024367544712025e-06,
+      "loss": 0.7611,
+      "step": 32280
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.7089254654863355,
+      "learning_rate": 4.022760090101393e-06,
+      "loss": 0.7659,
+      "step": 32285
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.064004428185941,
+      "learning_rate": 4.0211527404995045e-06,
+      "loss": 0.7625,
+      "step": 32290
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.697429299709736,
+      "learning_rate": 4.0195454960790715e-06,
+      "loss": 0.7129,
+      "step": 32295
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.4286762907772586,
+      "learning_rate": 4.017938357012804e-06,
+      "loss": 0.7145,
+      "step": 32300
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.945370514148755,
+      "learning_rate": 4.016331323473393e-06,
+      "loss": 0.7162,
+      "step": 32305
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.789212200932463,
+      "learning_rate": 4.014724395633523e-06,
+      "loss": 0.7424,
+      "step": 32310
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.104215820493675,
+      "learning_rate": 4.013117573665866e-06,
+      "loss": 0.7726,
+      "step": 32315
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.668650501945677,
+      "learning_rate": 4.011510857743079e-06,
+      "loss": 0.7428,
+      "step": 32320
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 19.135462965548953,
+      "learning_rate": 4.0099042480378155e-06,
+      "loss": 0.7649,
+      "step": 32325
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.276693225153347,
+      "learning_rate": 4.008297744722709e-06,
+      "loss": 0.7273,
+      "step": 32330
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.683834211031284,
+      "learning_rate": 4.006691347970386e-06,
+      "loss": 0.759,
+      "step": 32335
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.1302909383361275,
+      "learning_rate": 4.005085057953463e-06,
+      "loss": 0.7441,
+      "step": 32340
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.763589626580232,
+      "learning_rate": 4.0034788748445385e-06,
+      "loss": 0.6944,
+      "step": 32345
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.823536433778182,
+      "learning_rate": 4.00187279881621e-06,
+      "loss": 0.7211,
+      "step": 32350
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.890491462853178,
+      "learning_rate": 4.000266830041053e-06,
+      "loss": 0.7386,
+      "step": 32355
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.279807004199837,
+      "learning_rate": 3.998660968691636e-06,
+      "loss": 0.7236,
+      "step": 32360
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.480624980131441,
+      "learning_rate": 3.99705521494052e-06,
+      "loss": 0.7493,
+      "step": 32365
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.14373201760105,
+      "learning_rate": 3.995449568960244e-06,
+      "loss": 0.6911,
+      "step": 32370
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.308649146086617,
+      "learning_rate": 3.9938440309233485e-06,
+      "loss": 0.7192,
+      "step": 32375
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.723555952214085,
+      "learning_rate": 3.99223860100235e-06,
+      "loss": 0.8161,
+      "step": 32380
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.9143905317605245,
+      "learning_rate": 3.990633279369763e-06,
+      "loss": 0.7385,
+      "step": 32385
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.874189250447786,
+      "learning_rate": 3.989028066198085e-06,
+      "loss": 0.7598,
+      "step": 32390
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.392860507148786,
+      "learning_rate": 3.987422961659801e-06,
+      "loss": 0.7392,
+      "step": 32395
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.752071227232578,
+      "learning_rate": 3.98581796592739e-06,
+      "loss": 0.7702,
+      "step": 32400
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.491806669605895,
+      "learning_rate": 3.984213079173314e-06,
+      "loss": 0.7225,
+      "step": 32405
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.027790376668704,
+      "learning_rate": 3.982608301570027e-06,
+      "loss": 0.7446,
+      "step": 32410
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.539158941213466,
+      "learning_rate": 3.981003633289968e-06,
+      "loss": 0.7488,
+      "step": 32415
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.506726147339293,
+      "learning_rate": 3.979399074505564e-06,
+      "loss": 0.7352,
+      "step": 32420
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.977666467236214,
+      "learning_rate": 3.977794625389235e-06,
+      "loss": 0.7606,
+      "step": 32425
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.131229336936734,
+      "learning_rate": 3.976190286113384e-06,
+      "loss": 0.7475,
+      "step": 32430
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.437401289549152,
+      "learning_rate": 3.974586056850406e-06,
+      "loss": 0.7811,
+      "step": 32435
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.246324271116518,
+      "learning_rate": 3.972981937772681e-06,
+      "loss": 0.7579,
+      "step": 32440
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.721942524027835,
+      "learning_rate": 3.97137792905258e-06,
+      "loss": 0.8232,
+      "step": 32445
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.269173428548643,
+      "learning_rate": 3.96977403086246e-06,
+      "loss": 0.7119,
+      "step": 32450
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.602933566525741,
+      "learning_rate": 3.968170243374666e-06,
+      "loss": 0.7188,
+      "step": 32455
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.749783608193544,
+      "learning_rate": 3.966566566761535e-06,
+      "loss": 0.7708,
+      "step": 32460
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.059023493743448,
+      "learning_rate": 3.964963001195385e-06,
+      "loss": 0.714,
+      "step": 32465
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.045492016603623,
+      "learning_rate": 3.963359546848531e-06,
+      "loss": 0.7396,
+      "step": 32470
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.890409114883049,
+      "learning_rate": 3.961756203893269e-06,
+      "loss": 0.7636,
+      "step": 32475
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.739718233715035,
+      "learning_rate": 3.960152972501881e-06,
+      "loss": 0.755,
+      "step": 32480
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.876632833186473,
+      "learning_rate": 3.9585498528466495e-06,
+      "loss": 0.7837,
+      "step": 32485
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.417276614614368,
+      "learning_rate": 3.95694684509983e-06,
+      "loss": 0.7206,
+      "step": 32490
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 29.818927515288983,
+      "learning_rate": 3.955343949433677e-06,
+      "loss": 0.7615,
+      "step": 32495
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.838857168759922,
+      "learning_rate": 3.9537411660204275e-06,
+      "loss": 0.7185,
+      "step": 32500
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.803653681464658,
+      "learning_rate": 3.952138495032304e-06,
+      "loss": 0.7308,
+      "step": 32505
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.529705026388461,
+      "learning_rate": 3.950535936641527e-06,
+      "loss": 0.741,
+      "step": 32510
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.914597086797618,
+      "learning_rate": 3.948933491020293e-06,
+      "loss": 0.7171,
+      "step": 32515
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.0065243443796,
+      "learning_rate": 3.947331158340796e-06,
+      "loss": 0.751,
+      "step": 32520
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.829841352091856,
+      "learning_rate": 3.945728938775213e-06,
+      "loss": 0.7525,
+      "step": 32525
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.874341182645252,
+      "learning_rate": 3.944126832495705e-06,
+      "loss": 0.7685,
+      "step": 32530
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.961402296516516,
+      "learning_rate": 3.94252483967443e-06,
+      "loss": 0.7186,
+      "step": 32535
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.212277421026527,
+      "learning_rate": 3.940922960483528e-06,
+      "loss": 0.7439,
+      "step": 32540
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.139824947731008,
+      "learning_rate": 3.939321195095128e-06,
+      "loss": 0.7,
+      "step": 32545
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.384602426818018,
+      "learning_rate": 3.9377195436813455e-06,
+      "loss": 0.7088,
+      "step": 32550
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.924989819051035,
+      "learning_rate": 3.936118006414289e-06,
+      "loss": 0.7283,
+      "step": 32555
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.195762346201736,
+      "learning_rate": 3.934516583466048e-06,
+      "loss": 0.7697,
+      "step": 32560
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.924731308787463,
+      "learning_rate": 3.9329152750086996e-06,
+      "loss": 0.7106,
+      "step": 32565
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.19683831568055,
+      "learning_rate": 3.931314081214317e-06,
+      "loss": 0.7297,
+      "step": 32570
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.780190002870143,
+      "learning_rate": 3.92971300225495e-06,
+      "loss": 0.7038,
+      "step": 32575
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.429819226463751,
+      "learning_rate": 3.928112038302647e-06,
+      "loss": 0.7694,
+      "step": 32580
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.061228380966536,
+      "learning_rate": 3.926511189529435e-06,
+      "loss": 0.7275,
+      "step": 32585
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.880655402822012,
+      "learning_rate": 3.924910456107333e-06,
+      "loss": 0.7909,
+      "step": 32590
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.148758703138844,
+      "learning_rate": 3.923309838208349e-06,
+      "loss": 0.7439,
+      "step": 32595
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.109903427231474,
+      "learning_rate": 3.921709336004471e-06,
+      "loss": 0.7278,
+      "step": 32600
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.192494836222281,
+      "learning_rate": 3.920108949667687e-06,
+      "loss": 0.7814,
+      "step": 32605
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.978464096329849,
+      "learning_rate": 3.9185086793699625e-06,
+      "loss": 0.751,
+      "step": 32610
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.304591997611985,
+      "learning_rate": 3.916908525283252e-06,
+      "loss": 0.7367,
+      "step": 32615
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.2673916109453796,
+      "learning_rate": 3.915308487579501e-06,
+      "loss": 0.7647,
+      "step": 32620
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.867734512133385,
+      "learning_rate": 3.913708566430639e-06,
+      "loss": 0.734,
+      "step": 32625
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.756938881781644,
+      "learning_rate": 3.912108762008586e-06,
+      "loss": 0.7547,
+      "step": 32630
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.629908597637033,
+      "learning_rate": 3.910509074485249e-06,
+      "loss": 0.721,
+      "step": 32635
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.891973908041338,
+      "learning_rate": 3.908909504032519e-06,
+      "loss": 0.7008,
+      "step": 32640
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.809622620180986,
+      "learning_rate": 3.90731005082228e-06,
+      "loss": 0.7223,
+      "step": 32645
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.444578271276455,
+      "learning_rate": 3.905710715026396e-06,
+      "loss": 0.7038,
+      "step": 32650
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.431709121181191,
+      "learning_rate": 3.904111496816727e-06,
+      "loss": 0.7937,
+      "step": 32655
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.224569572129378,
+      "learning_rate": 3.902512396365113e-06,
+      "loss": 0.7325,
+      "step": 32660
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.706558355927059,
+      "learning_rate": 3.900913413843387e-06,
+      "loss": 0.7295,
+      "step": 32665
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.198248517035201,
+      "learning_rate": 3.899314549423364e-06,
+      "loss": 0.7261,
+      "step": 32670
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.716783980003042,
+      "learning_rate": 3.89771580327685e-06,
+      "loss": 0.7579,
+      "step": 32675
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.870755285483422,
+      "learning_rate": 3.8961171755756385e-06,
+      "loss": 0.7657,
+      "step": 32680
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.1597131749413405,
+      "learning_rate": 3.894518666491507e-06,
+      "loss": 0.7335,
+      "step": 32685
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.69179036006,
+      "learning_rate": 3.892920276196226e-06,
+      "loss": 0.7355,
+      "step": 32690
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.888652128757496,
+      "learning_rate": 3.891322004861545e-06,
+      "loss": 0.7774,
+      "step": 32695
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.745532279358965,
+      "learning_rate": 3.88972385265921e-06,
+      "loss": 0.7423,
+      "step": 32700
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.46785471070572,
+      "learning_rate": 3.888125819760947e-06,
+      "loss": 0.7153,
+      "step": 32705
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.664777069083331,
+      "learning_rate": 3.8865279063384695e-06,
+      "loss": 0.7199,
+      "step": 32710
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.250637047504215,
+      "learning_rate": 3.884930112563484e-06,
+      "loss": 0.7388,
+      "step": 32715
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.100398930029195,
+      "learning_rate": 3.883332438607677e-06,
+      "loss": 0.7264,
+      "step": 32720
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.878890099982266,
+      "learning_rate": 3.88173488464273e-06,
+      "loss": 0.7085,
+      "step": 32725
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.998895867735593,
+      "learning_rate": 3.880137450840305e-06,
+      "loss": 0.7531,
+      "step": 32730
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.192499568647865,
+      "learning_rate": 3.878540137372051e-06,
+      "loss": 0.7167,
+      "step": 32735
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.935148044262539,
+      "learning_rate": 3.876942944409612e-06,
+      "loss": 0.7391,
+      "step": 32740
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.713431584224038,
+      "learning_rate": 3.8753458721246065e-06,
+      "loss": 0.7335,
+      "step": 32745
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.154152487697281,
+      "learning_rate": 3.873748920688653e-06,
+      "loss": 0.7053,
+      "step": 32750
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.250284148721991,
+      "learning_rate": 3.872152090273349e-06,
+      "loss": 0.7735,
+      "step": 32755
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.628606587302714,
+      "learning_rate": 3.870555381050279e-06,
+      "loss": 0.7185,
+      "step": 32760
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.948409626958064,
+      "learning_rate": 3.868958793191017e-06,
+      "loss": 0.7253,
+      "step": 32765
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.46461713091953,
+      "learning_rate": 3.867362326867126e-06,
+      "loss": 0.6974,
+      "step": 32770
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.726005685383899,
+      "learning_rate": 3.86576598225015e-06,
+      "loss": 0.6606,
+      "step": 32775
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.835234711108583,
+      "learning_rate": 3.864169759511625e-06,
+      "loss": 0.7467,
+      "step": 32780
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.697925745596195,
+      "learning_rate": 3.862573658823075e-06,
+      "loss": 0.7096,
+      "step": 32785
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.802825289105715,
+      "learning_rate": 3.860977680356004e-06,
+      "loss": 0.7519,
+      "step": 32790
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.749256680968329,
+      "learning_rate": 3.859381824281906e-06,
+      "loss": 0.7712,
+      "step": 32795
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.726770427574885,
+      "learning_rate": 3.857786090772268e-06,
+      "loss": 0.7582,
+      "step": 32800
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.721761866857864,
+      "learning_rate": 3.856190479998553e-06,
+      "loss": 0.7177,
+      "step": 32805
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.158233470231522,
+      "learning_rate": 3.854594992132221e-06,
+      "loss": 0.7203,
+      "step": 32810
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.827405380450159,
+      "learning_rate": 3.852999627344711e-06,
+      "loss": 0.6741,
+      "step": 32815
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.828920311623465,
+      "learning_rate": 3.851404385807452e-06,
+      "loss": 0.7503,
+      "step": 32820
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.042129343470339,
+      "learning_rate": 3.849809267691863e-06,
+      "loss": 0.7249,
+      "step": 32825
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.8341905136822865,
+      "learning_rate": 3.8482142731693425e-06,
+      "loss": 0.7749,
+      "step": 32830
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.158277977080501,
+      "learning_rate": 3.846619402411283e-06,
+      "loss": 0.6899,
+      "step": 32835
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.736043582193756,
+      "learning_rate": 3.8450246555890595e-06,
+      "loss": 0.6957,
+      "step": 32840
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.758376958474109,
+      "learning_rate": 3.843430032874032e-06,
+      "loss": 0.7464,
+      "step": 32845
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.30633803296278,
+      "learning_rate": 3.841835534437554e-06,
+      "loss": 0.7352,
+      "step": 32850
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.835532010013905,
+      "learning_rate": 3.840241160450957e-06,
+      "loss": 0.7794,
+      "step": 32855
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.039623406688664,
+      "learning_rate": 3.838646911085566e-06,
+      "loss": 0.7566,
+      "step": 32860
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.281947195613584,
+      "learning_rate": 3.837052786512691e-06,
+      "loss": 0.7357,
+      "step": 32865
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.168850515003406,
+      "learning_rate": 3.835458786903627e-06,
+      "loss": 0.7608,
+      "step": 32870
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.867820813372498,
+      "learning_rate": 3.833864912429655e-06,
+      "loss": 0.734,
+      "step": 32875
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.53356939155142,
+      "learning_rate": 3.8322711632620434e-06,
+      "loss": 0.7068,
+      "step": 32880
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.84947732054856,
+      "learning_rate": 3.830677539572052e-06,
+      "loss": 0.7188,
+      "step": 32885
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.341087049202473,
+      "learning_rate": 3.829084041530916e-06,
+      "loss": 0.742,
+      "step": 32890
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.51850578911999,
+      "learning_rate": 3.827490669309871e-06,
+      "loss": 0.7338,
+      "step": 32895
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.777830711988462,
+      "learning_rate": 3.825897423080126e-06,
+      "loss": 0.6848,
+      "step": 32900
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.090353498447795,
+      "learning_rate": 3.824304303012884e-06,
+      "loss": 0.7496,
+      "step": 32905
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.6346802442528725,
+      "learning_rate": 3.8227113092793355e-06,
+      "loss": 0.7971,
+      "step": 32910
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.615440735398652,
+      "learning_rate": 3.82111844205065e-06,
+      "loss": 0.7531,
+      "step": 32915
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.5549202926848125,
+      "learning_rate": 3.819525701497994e-06,
+      "loss": 0.7452,
+      "step": 32920
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.884484306822319,
+      "learning_rate": 3.81793308779251e-06,
+      "loss": 0.7312,
+      "step": 32925
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.954313405873962,
+      "learning_rate": 3.816340601105331e-06,
+      "loss": 0.7356,
+      "step": 32930
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.771480389507657,
+      "learning_rate": 3.81474824160758e-06,
+      "loss": 0.7476,
+      "step": 32935
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.646446813417095,
+      "learning_rate": 3.8131560094703597e-06,
+      "loss": 0.7214,
+      "step": 32940
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.139651060425786,
+      "learning_rate": 3.8115639048647656e-06,
+      "loss": 0.735,
+      "step": 32945
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.570965526611558,
+      "learning_rate": 3.8099719279618735e-06,
+      "loss": 0.7725,
+      "step": 32950
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.016560523626401,
+      "learning_rate": 3.8083800789327508e-06,
+      "loss": 0.6994,
+      "step": 32955
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.0086040446261455,
+      "learning_rate": 3.8067883579484483e-06,
+      "loss": 0.7449,
+      "step": 32960
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.251575214555384,
+      "learning_rate": 3.805196765180001e-06,
+      "loss": 0.7212,
+      "step": 32965
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.609832168174631,
+      "learning_rate": 3.803605300798436e-06,
+      "loss": 0.7261,
+      "step": 32970
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.085790608966236,
+      "learning_rate": 3.80201396497476e-06,
+      "loss": 0.7001,
+      "step": 32975
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.789936282743735,
+      "learning_rate": 3.8004227578799723e-06,
+      "loss": 0.727,
+      "step": 32980
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.770146878627893,
+      "learning_rate": 3.7988316796850533e-06,
+      "loss": 0.762,
+      "step": 32985
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.06776056933221,
+      "learning_rate": 3.797240730560969e-06,
+      "loss": 0.7501,
+      "step": 32990
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.152363973893559,
+      "learning_rate": 3.795649910678679e-06,
+      "loss": 0.7479,
+      "step": 32995
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.858114408066479,
+      "learning_rate": 3.79405922020912e-06,
+      "loss": 0.7296,
+      "step": 33000
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.219572523046081,
+      "learning_rate": 3.79246865932322e-06,
+      "loss": 0.7491,
+      "step": 33005
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.22601817351407,
+      "learning_rate": 3.7908782281918934e-06,
+      "loss": 0.7763,
+      "step": 33010
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.923123165622001,
+      "learning_rate": 3.7892879269860348e-06,
+      "loss": 0.7335,
+      "step": 33015
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.878097118926254,
+      "learning_rate": 3.7876977558765344e-06,
+      "loss": 0.7478,
+      "step": 33020
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.740463168941638,
+      "learning_rate": 3.786107715034258e-06,
+      "loss": 0.6969,
+      "step": 33025
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.83498617013336,
+      "learning_rate": 3.784517804630067e-06,
+      "loss": 0.7134,
+      "step": 33030
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.607155632496121,
+      "learning_rate": 3.7829280248348e-06,
+      "loss": 0.7279,
+      "step": 33035
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.254495005295007,
+      "learning_rate": 3.7813383758192906e-06,
+      "loss": 0.7649,
+      "step": 33040
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.339475826830785,
+      "learning_rate": 3.7797488577543494e-06,
+      "loss": 0.7334,
+      "step": 33045
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.010207978408473,
+      "learning_rate": 3.778159470810778e-06,
+      "loss": 0.7294,
+      "step": 33050
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.52219248375397,
+      "learning_rate": 3.7765702151593663e-06,
+      "loss": 0.7453,
+      "step": 33055
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.864643104171137,
+      "learning_rate": 3.7749810909708824e-06,
+      "loss": 0.7276,
+      "step": 33060
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.555456985361319,
+      "learning_rate": 3.773392098416089e-06,
+      "loss": 0.6888,
+      "step": 33065
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.639870938612942,
+      "learning_rate": 3.771803237665727e-06,
+      "loss": 0.7415,
+      "step": 33070
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.419973200174649,
+      "learning_rate": 3.770214508890526e-06,
+      "loss": 0.7374,
+      "step": 33075
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.743170008193136,
+      "learning_rate": 3.7686259122612057e-06,
+      "loss": 0.7429,
+      "step": 33080
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.7084056169955,
+      "learning_rate": 3.7670374479484635e-06,
+      "loss": 0.7328,
+      "step": 33085
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.5140714440406375,
+      "learning_rate": 3.76544911612299e-06,
+      "loss": 0.7574,
+      "step": 33090
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.230336189391053,
+      "learning_rate": 3.7638609169554584e-06,
+      "loss": 0.7611,
+      "step": 33095
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.12368552845721,
+      "learning_rate": 3.7622728506165247e-06,
+      "loss": 0.7384,
+      "step": 33100
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.381520027505094,
+      "learning_rate": 3.760684917276838e-06,
+      "loss": 0.7302,
+      "step": 33105
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.858433678077069,
+      "learning_rate": 3.759097117107024e-06,
+      "loss": 0.7331,
+      "step": 33110
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.432298389929707,
+      "learning_rate": 3.7575094502777038e-06,
+      "loss": 0.7029,
+      "step": 33115
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.598382904597247,
+      "learning_rate": 3.7559219169594742e-06,
+      "loss": 0.6795,
+      "step": 33120
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.787806713123113,
+      "learning_rate": 3.7543345173229274e-06,
+      "loss": 0.6868,
+      "step": 33125
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.749646146539174,
+      "learning_rate": 3.752747251538633e-06,
+      "loss": 0.7897,
+      "step": 33130
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.679588388653044,
+      "learning_rate": 3.751160119777152e-06,
+      "loss": 0.7627,
+      "step": 33135
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.333879132444865,
+      "learning_rate": 3.7495731222090263e-06,
+      "loss": 0.6887,
+      "step": 33140
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.838699235196046,
+      "learning_rate": 3.7479862590047868e-06,
+      "loss": 0.7324,
+      "step": 33145
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.495523715311052,
+      "learning_rate": 3.7463995303349516e-06,
+      "loss": 0.7341,
+      "step": 33150
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.962021011382005,
+      "learning_rate": 3.744812936370019e-06,
+      "loss": 0.7222,
+      "step": 33155
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.06668332820931,
+      "learning_rate": 3.743226477280474e-06,
+      "loss": 0.7076,
+      "step": 33160
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.789074340187547,
+      "learning_rate": 3.741640153236793e-06,
+      "loss": 0.7071,
+      "step": 33165
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.063639419988654,
+      "learning_rate": 3.740053964409428e-06,
+      "loss": 0.7113,
+      "step": 33170
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.783536229577145,
+      "learning_rate": 3.7384679109688284e-06,
+      "loss": 0.7867,
+      "step": 33175
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.940079076665862,
+      "learning_rate": 3.7368819930854165e-06,
+      "loss": 0.7687,
+      "step": 33180
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.1263887119496525,
+      "learning_rate": 3.7352962109296103e-06,
+      "loss": 0.734,
+      "step": 33185
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.428773217140986,
+      "learning_rate": 3.7337105646718086e-06,
+      "loss": 0.738,
+      "step": 33190
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.187692467521277,
+      "learning_rate": 3.732125054482394e-06,
+      "loss": 0.7588,
+      "step": 33195
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.5226357128826224,
+      "learning_rate": 3.7305396805317394e-06,
+      "loss": 0.7376,
+      "step": 33200
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.837206423449839,
+      "learning_rate": 3.7289544429901962e-06,
+      "loss": 0.6971,
+      "step": 33205
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.115755968885098,
+      "learning_rate": 3.7273693420281097e-06,
+      "loss": 0.7183,
+      "step": 33210
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.403755501235667,
+      "learning_rate": 3.725784377815805e-06,
+      "loss": 0.6442,
+      "step": 33215
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.774291127446088,
+      "learning_rate": 3.7241995505235893e-06,
+      "loss": 0.698,
+      "step": 33220
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.9821489305670195,
+      "learning_rate": 3.722614860321764e-06,
+      "loss": 0.7453,
+      "step": 33225
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.582326438191975,
+      "learning_rate": 3.72103030738061e-06,
+      "loss": 0.7474,
+      "step": 33230
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.116080293871856,
+      "learning_rate": 3.719445891870393e-06,
+      "loss": 0.7642,
+      "step": 33235
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.805096225135345,
+      "learning_rate": 3.7178616139613676e-06,
+      "loss": 0.753,
+      "step": 33240
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.1202490150168085,
+      "learning_rate": 3.716277473823768e-06,
+      "loss": 0.7356,
+      "step": 33245
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.983644991572485,
+      "learning_rate": 3.714693471627823e-06,
+      "loss": 0.7536,
+      "step": 33250
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.85707263913975,
+      "learning_rate": 3.713109607543733e-06,
+      "loss": 0.7287,
+      "step": 33255
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.3630504564674,
+      "learning_rate": 3.711525881741698e-06,
+      "loss": 0.7633,
+      "step": 33260
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.331488525980054,
+      "learning_rate": 3.709942294391892e-06,
+      "loss": 0.7158,
+      "step": 33265
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.552899225732661,
+      "learning_rate": 3.708358845664481e-06,
+      "loss": 0.7345,
+      "step": 33270
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.969317872162896,
+      "learning_rate": 3.706775535729612e-06,
+      "loss": 0.727,
+      "step": 33275
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.639046499207537,
+      "learning_rate": 3.7051923647574172e-06,
+      "loss": 0.7539,
+      "step": 33280
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.504805829946549,
+      "learning_rate": 3.7036093329180206e-06,
+      "loss": 0.7676,
+      "step": 33285
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 17.476568490832356,
+      "learning_rate": 3.70202644038152e-06,
+      "loss": 0.7336,
+      "step": 33290
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.832739640367027,
+      "learning_rate": 3.7004436873180083e-06,
+      "loss": 0.7267,
+      "step": 33295
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.767689201380548,
+      "learning_rate": 3.6988610738975585e-06,
+      "loss": 0.71,
+      "step": 33300
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.826464465345845,
+      "learning_rate": 3.697278600290226e-06,
+      "loss": 0.7865,
+      "step": 33305
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.330790330324903,
+      "learning_rate": 3.695696266666059e-06,
+      "loss": 0.7579,
+      "step": 33310
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.028007907363844,
+      "learning_rate": 3.694114073195084e-06,
+      "loss": 0.658,
+      "step": 33315
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.069186402198259,
+      "learning_rate": 3.692532020047316e-06,
+      "loss": 0.684,
+      "step": 33320
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.097981862953665,
+      "learning_rate": 3.690950107392752e-06,
+      "loss": 0.7477,
+      "step": 33325
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.876920740713685,
+      "learning_rate": 3.6893683354013752e-06,
+      "loss": 0.6985,
+      "step": 33330
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.8937221390226355,
+      "learning_rate": 3.6877867042431575e-06,
+      "loss": 0.7423,
+      "step": 33335
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.050457359324089,
+      "learning_rate": 3.6862052140880474e-06,
+      "loss": 0.7166,
+      "step": 33340
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.19579683854763,
+      "learning_rate": 3.6846238651059874e-06,
+      "loss": 0.7302,
+      "step": 33345
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.521345566681065,
+      "learning_rate": 3.6830426574668953e-06,
+      "loss": 0.7746,
+      "step": 33350
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.151783416296471,
+      "learning_rate": 3.6814615913406846e-06,
+      "loss": 0.7431,
+      "step": 33355
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.38410660950617,
+      "learning_rate": 3.6798806668972443e-06,
+      "loss": 0.7377,
+      "step": 33360
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.897383699146823,
+      "learning_rate": 3.678299884306451e-06,
+      "loss": 0.6312,
+      "step": 33365
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.248718368567392,
+      "learning_rate": 3.6767192437381705e-06,
+      "loss": 0.7071,
+      "step": 33370
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.236465170126863,
+      "learning_rate": 3.675138745362245e-06,
+      "loss": 0.7521,
+      "step": 33375
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.390162974629451,
+      "learning_rate": 3.6735583893485104e-06,
+      "loss": 0.7272,
+      "step": 33380
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.016337995684026,
+      "learning_rate": 3.671978175866782e-06,
+      "loss": 0.7513,
+      "step": 33385
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.835615259113764,
+      "learning_rate": 3.6703981050868563e-06,
+      "loss": 0.6902,
+      "step": 33390
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.795694439318842,
+      "learning_rate": 3.6688181771785246e-06,
+      "loss": 0.7399,
+      "step": 33395
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.1015435728597565,
+      "learning_rate": 3.667238392311553e-06,
+      "loss": 0.7502,
+      "step": 33400
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.619985644047551,
+      "learning_rate": 3.6656587506556984e-06,
+      "loss": 0.7173,
+      "step": 33405
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.303040301753041,
+      "learning_rate": 3.664079252380701e-06,
+      "loss": 0.7417,
+      "step": 33410
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.9985413869010635,
+      "learning_rate": 3.6624998976562824e-06,
+      "loss": 0.7207,
+      "step": 33415
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.825821084031015,
+      "learning_rate": 3.6609206866521546e-06,
+      "loss": 0.7142,
+      "step": 33420
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.997454816895571,
+      "learning_rate": 3.659341619538006e-06,
+      "loss": 0.6869,
+      "step": 33425
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.52407682161989,
+      "learning_rate": 3.65776269648352e-06,
+      "loss": 0.736,
+      "step": 33430
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.977158052908027,
+      "learning_rate": 3.6561839176583536e-06,
+      "loss": 0.7837,
+      "step": 33435
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.333562391025884,
+      "learning_rate": 3.654605283232159e-06,
+      "loss": 0.7257,
+      "step": 33440
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.123224172007378,
+      "learning_rate": 3.6530267933745623e-06,
+      "loss": 0.7105,
+      "step": 33445
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 21.807940846621342,
+      "learning_rate": 3.6514484482551817e-06,
+      "loss": 0.7555,
+      "step": 33450
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.509095728610213,
+      "learning_rate": 3.6498702480436167e-06,
+      "loss": 0.7124,
+      "step": 33455
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.424522271011536,
+      "learning_rate": 3.6482921929094507e-06,
+      "loss": 0.7087,
+      "step": 33460
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.596304477963848,
+      "learning_rate": 3.6467142830222566e-06,
+      "loss": 0.759,
+      "step": 33465
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.76604585438323,
+      "learning_rate": 3.645136518551585e-06,
+      "loss": 0.7096,
+      "step": 33470
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.844500528948266,
+      "learning_rate": 3.6435588996669726e-06,
+      "loss": 0.7187,
+      "step": 33475
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.83270439365789,
+      "learning_rate": 3.641981426537945e-06,
+      "loss": 0.6937,
+      "step": 33480
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.993599292802973,
+      "learning_rate": 3.6404040993340038e-06,
+      "loss": 0.7443,
+      "step": 33485
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.47810486396018,
+      "learning_rate": 3.6388269182246444e-06,
+      "loss": 0.7402,
+      "step": 33490
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.301075576795128,
+      "learning_rate": 3.637249883379339e-06,
+      "loss": 0.6943,
+      "step": 33495
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.864076173611569,
+      "learning_rate": 3.6356729949675474e-06,
+      "loss": 0.7046,
+      "step": 33500
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.756086870112118,
+      "learning_rate": 3.6340962531587153e-06,
+      "loss": 0.7301,
+      "step": 33505
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.617545638067697,
+      "learning_rate": 3.632519658122268e-06,
+      "loss": 0.7547,
+      "step": 33510
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.277245488928964,
+      "learning_rate": 3.6309432100276193e-06,
+      "loss": 0.7378,
+      "step": 33515
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.168528630142161,
+      "learning_rate": 3.629366909044163e-06,
+      "loss": 0.7512,
+      "step": 33520
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.352948888128745,
+      "learning_rate": 3.627790755341284e-06,
+      "loss": 0.7119,
+      "step": 33525
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.735412290404789,
+      "learning_rate": 3.6262147490883436e-06,
+      "loss": 0.7822,
+      "step": 33530
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.92245499413314,
+      "learning_rate": 3.6246388904546905e-06,
+      "loss": 0.6771,
+      "step": 33535
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.61480618008788,
+      "learning_rate": 3.6230631796096584e-06,
+      "loss": 0.7735,
+      "step": 33540
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.943676178400943,
+      "learning_rate": 3.6214876167225648e-06,
+      "loss": 0.6994,
+      "step": 33545
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.898413698136308,
+      "learning_rate": 3.619912201962711e-06,
+      "loss": 0.7006,
+      "step": 33550
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.480076412106544,
+      "learning_rate": 3.6183369354993827e-06,
+      "loss": 0.7271,
+      "step": 33555
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.202830880100374,
+      "learning_rate": 3.616761817501846e-06,
+      "loss": 0.7322,
+      "step": 33560
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.678698441288536,
+      "learning_rate": 3.6151868481393592e-06,
+      "loss": 0.7178,
+      "step": 33565
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.551095927024385,
+      "learning_rate": 3.6136120275811554e-06,
+      "loss": 0.6984,
+      "step": 33570
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.91493372463631,
+      "learning_rate": 3.6120373559964593e-06,
+      "loss": 0.6605,
+      "step": 33575
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.399261163381923,
+      "learning_rate": 3.6104628335544743e-06,
+      "loss": 0.7247,
+      "step": 33580
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.399695530011383,
+      "learning_rate": 3.60888846042439e-06,
+      "loss": 0.7189,
+      "step": 33585
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.4772390707745995,
+      "learning_rate": 3.6073142367753805e-06,
+      "loss": 0.701,
+      "step": 33590
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.842214399711519,
+      "learning_rate": 3.6057401627766007e-06,
+      "loss": 0.7019,
+      "step": 33595
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.611602914394154,
+      "learning_rate": 3.604166238597197e-06,
+      "loss": 0.6859,
+      "step": 33600
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.802945363645624,
+      "learning_rate": 3.602592464406288e-06,
+      "loss": 0.7219,
+      "step": 33605
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.753396496273493,
+      "learning_rate": 3.6010188403729893e-06,
+      "loss": 0.6851,
+      "step": 33610
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.238619830630508,
+      "learning_rate": 3.59944536666639e-06,
+      "loss": 0.7184,
+      "step": 33615
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.131524168937709,
+      "learning_rate": 3.5978720434555646e-06,
+      "loss": 0.7288,
+      "step": 33620
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.224120621845438,
+      "learning_rate": 3.596298870909578e-06,
+      "loss": 0.7214,
+      "step": 33625
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.805090207311748,
+      "learning_rate": 3.594725849197471e-06,
+      "loss": 0.7196,
+      "step": 33630
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.394838594112325,
+      "learning_rate": 3.5931529784882748e-06,
+      "loss": 0.7105,
+      "step": 33635
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.980071892581502,
+      "learning_rate": 3.5915802589510005e-06,
+      "loss": 0.7879,
+      "step": 33640
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.32604819181313,
+      "learning_rate": 3.590007690754642e-06,
+      "loss": 0.7558,
+      "step": 33645
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.763935024362741,
+      "learning_rate": 3.5884352740681816e-06,
+      "loss": 0.7389,
+      "step": 33650
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.46121061875503,
+      "learning_rate": 3.5868630090605784e-06,
+      "loss": 0.7454,
+      "step": 33655
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.305726612329147,
+      "learning_rate": 3.585290895900784e-06,
+      "loss": 0.6992,
+      "step": 33660
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.174971837692179,
+      "learning_rate": 3.5837189347577234e-06,
+      "loss": 0.7343,
+      "step": 33665
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.650068931154648,
+      "learning_rate": 3.582147125800316e-06,
+      "loss": 0.715,
+      "step": 33670
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.887708001835577,
+      "learning_rate": 3.580575469197457e-06,
+      "loss": 0.7363,
+      "step": 33675
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.662671837795487,
+      "learning_rate": 3.579003965118028e-06,
+      "loss": 0.7382,
+      "step": 33680
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.762266468819768,
+      "learning_rate": 3.5774326137308944e-06,
+      "loss": 0.7154,
+      "step": 33685
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 15.601980321054873,
+      "learning_rate": 3.575861415204904e-06,
+      "loss": 0.7213,
+      "step": 33690
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.108175915047928,
+      "learning_rate": 3.574290369708891e-06,
+      "loss": 0.6971,
+      "step": 33695
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.761560473533494,
+      "learning_rate": 3.5727194774116702e-06,
+      "loss": 0.7546,
+      "step": 33700
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 30.499099568455062,
+      "learning_rate": 3.571148738482039e-06,
+      "loss": 0.7239,
+      "step": 33705
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.189839583208315,
+      "learning_rate": 3.5695781530887834e-06,
+      "loss": 0.6914,
+      "step": 33710
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.638723198655214,
+      "learning_rate": 3.5680077214006664e-06,
+      "loss": 0.7422,
+      "step": 33715
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.994479886348189,
+      "learning_rate": 3.5664374435864414e-06,
+      "loss": 0.7155,
+      "step": 33720
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.134008146783192,
+      "learning_rate": 3.5648673198148394e-06,
+      "loss": 0.7359,
+      "step": 33725
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.486958398123869,
+      "learning_rate": 3.563297350254576e-06,
+      "loss": 0.7086,
+      "step": 33730
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.670895683960037,
+      "learning_rate": 3.5617275350743553e-06,
+      "loss": 0.7719,
+      "step": 33735
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.827079092448349,
+      "learning_rate": 3.5601578744428565e-06,
+      "loss": 0.6919,
+      "step": 33740
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.4745233786034015,
+      "learning_rate": 3.5585883685287504e-06,
+      "loss": 0.72,
+      "step": 33745
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.621025207694066,
+      "learning_rate": 3.5570190175006836e-06,
+      "loss": 0.6989,
+      "step": 33750
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.806551040034962,
+      "learning_rate": 3.5554498215272938e-06,
+      "loss": 0.7461,
+      "step": 33755
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.199203112108226,
+      "learning_rate": 3.5538807807771956e-06,
+      "loss": 0.663,
+      "step": 33760
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.8222320361041895,
+      "learning_rate": 3.552311895418988e-06,
+      "loss": 0.6926,
+      "step": 33765
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.32950681888661,
+      "learning_rate": 3.5507431656212575e-06,
+      "loss": 0.7439,
+      "step": 33770
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.892683554255738,
+      "learning_rate": 3.54917459155257e-06,
+      "loss": 0.7089,
+      "step": 33775
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.152434360621356,
+      "learning_rate": 3.547606173381476e-06,
+      "loss": 0.7351,
+      "step": 33780
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.34090279185434,
+      "learning_rate": 3.546037911276509e-06,
+      "loss": 0.7556,
+      "step": 33785
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.751353575444856,
+      "learning_rate": 3.5444698054061834e-06,
+      "loss": 0.7285,
+      "step": 33790
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.224222411745411,
+      "learning_rate": 3.5429018559390026e-06,
+      "loss": 0.6832,
+      "step": 33795
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.734189281023405,
+      "learning_rate": 3.5413340630434463e-06,
+      "loss": 0.6813,
+      "step": 33800
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.424637752154393,
+      "learning_rate": 3.5397664268879844e-06,
+      "loss": 0.6999,
+      "step": 33805
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.183836385427888,
+      "learning_rate": 3.5381989476410637e-06,
+      "loss": 0.7281,
+      "step": 33810
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.462419754167492,
+      "learning_rate": 3.5366316254711176e-06,
+      "loss": 0.7442,
+      "step": 33815
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.56039018391325,
+      "learning_rate": 3.5350644605465612e-06,
+      "loss": 0.6912,
+      "step": 33820
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.727771553841209,
+      "learning_rate": 3.5334974530357928e-06,
+      "loss": 0.6866,
+      "step": 33825
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.508507776083809,
+      "learning_rate": 3.5319306031071977e-06,
+      "loss": 0.7128,
+      "step": 33830
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.069321571392466,
+      "learning_rate": 3.530363910929136e-06,
+      "loss": 0.7264,
+      "step": 33835
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.145850859528512,
+      "learning_rate": 3.52879737666996e-06,
+      "loss": 0.7524,
+      "step": 33840
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.3433397719816345,
+      "learning_rate": 3.527231000497999e-06,
+      "loss": 0.7089,
+      "step": 33845
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.354804226800192,
+      "learning_rate": 3.5256647825815647e-06,
+      "loss": 0.7212,
+      "step": 33850
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.674002308860486,
+      "learning_rate": 3.5240987230889588e-06,
+      "loss": 0.6799,
+      "step": 33855
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.746560071597657,
+      "learning_rate": 3.522532822188456e-06,
+      "loss": 0.6992,
+      "step": 33860
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.797984805419982,
+      "learning_rate": 3.520967080048323e-06,
+      "loss": 0.6963,
+      "step": 33865
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.040668604292417,
+      "learning_rate": 3.5194014968368064e-06,
+      "loss": 0.7128,
+      "step": 33870
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.759423601666029,
+      "learning_rate": 3.517836072722131e-06,
+      "loss": 0.7009,
+      "step": 33875
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.9749233946751374,
+      "learning_rate": 3.5162708078725127e-06,
+      "loss": 0.7393,
+      "step": 33880
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.402478597008374,
+      "learning_rate": 3.5147057024561426e-06,
+      "loss": 0.7264,
+      "step": 33885
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.206926525704742,
+      "learning_rate": 3.5131407566412017e-06,
+      "loss": 0.7052,
+      "step": 33890
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.9274820775913835,
+      "learning_rate": 3.5115759705958486e-06,
+      "loss": 0.7663,
+      "step": 33895
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.039420048469241,
+      "learning_rate": 3.5100113444882246e-06,
+      "loss": 0.7327,
+      "step": 33900
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.3161812038229606,
+      "learning_rate": 3.508446878486458e-06,
+      "loss": 0.6992,
+      "step": 33905
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.615202691945004,
+      "learning_rate": 3.5068825727586575e-06,
+      "loss": 0.6852,
+      "step": 33910
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.301251088519508,
+      "learning_rate": 3.5053184274729135e-06,
+      "loss": 0.7051,
+      "step": 33915
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.2922900641527155,
+      "learning_rate": 3.5037544427973e-06,
+      "loss": 0.7436,
+      "step": 33920
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.819673966870845,
+      "learning_rate": 3.502190618899878e-06,
+      "loss": 0.7124,
+      "step": 33925
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.339869741268306,
+      "learning_rate": 3.5006269559486837e-06,
+      "loss": 0.7239,
+      "step": 33930
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.903721571102292,
+      "learning_rate": 3.4990634541117383e-06,
+      "loss": 0.719,
+      "step": 33935
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.3735729078079535,
+      "learning_rate": 3.4975001135570506e-06,
+      "loss": 0.7214,
+      "step": 33940
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.223521645337661,
+      "learning_rate": 3.495936934452605e-06,
+      "loss": 0.6887,
+      "step": 33945
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.04464718672468,
+      "learning_rate": 3.494373916966376e-06,
+      "loss": 0.6751,
+      "step": 33950
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.491194395049443,
+      "learning_rate": 3.4928110612663135e-06,
+      "loss": 0.7298,
+      "step": 33955
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.885557763092124,
+      "learning_rate": 3.491248367520352e-06,
+      "loss": 0.6642,
+      "step": 33960
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.193732538609462,
+      "learning_rate": 3.489685835896416e-06,
+      "loss": 0.6913,
+      "step": 33965
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.772620079382848,
+      "learning_rate": 3.4881234665623984e-06,
+      "loss": 0.7063,
+      "step": 33970
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.207586008164692,
+      "learning_rate": 3.4865612596861903e-06,
+      "loss": 0.7483,
+      "step": 33975
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.241460421403689,
+      "learning_rate": 3.484999215435654e-06,
+      "loss": 0.7176,
+      "step": 33980
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.130022812908069,
+      "learning_rate": 3.4834373339786354e-06,
+      "loss": 0.6604,
+      "step": 33985
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.902386369940835,
+      "learning_rate": 3.481875615482971e-06,
+      "loss": 0.7335,
+      "step": 33990
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.211343008779445,
+      "learning_rate": 3.4803140601164697e-06,
+      "loss": 0.7406,
+      "step": 33995
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.992851876311853,
+      "learning_rate": 3.4787526680469297e-06,
+      "loss": 0.7392,
+      "step": 34000
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.987220985759034,
+      "learning_rate": 3.4771914394421303e-06,
+      "loss": 0.716,
+      "step": 34005
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.046684606330692,
+      "learning_rate": 3.4756303744698306e-06,
+      "loss": 0.7175,
+      "step": 34010
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.7129756591946945,
+      "learning_rate": 3.4740694732977766e-06,
+      "loss": 0.6947,
+      "step": 34015
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.1962218242675675,
+      "learning_rate": 3.472508736093689e-06,
+      "loss": 0.6953,
+      "step": 34020
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.007780165649483,
+      "learning_rate": 3.4709481630252822e-06,
+      "loss": 0.6763,
+      "step": 34025
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.4200840203855805,
+      "learning_rate": 3.469387754260241e-06,
+      "loss": 0.7273,
+      "step": 34030
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.22927267066624,
+      "learning_rate": 3.467827509966243e-06,
+      "loss": 0.6916,
+      "step": 34035
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.363394374762744,
+      "learning_rate": 3.4662674303109394e-06,
+      "loss": 0.7153,
+      "step": 34040
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.122240000977692,
+      "learning_rate": 3.4647075154619695e-06,
+      "loss": 0.7313,
+      "step": 34045
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.417056892609173,
+      "learning_rate": 3.4631477655869535e-06,
+      "loss": 0.6756,
+      "step": 34050
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.9046395257729305,
+      "learning_rate": 3.461588180853491e-06,
+      "loss": 0.7144,
+      "step": 34055
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.948375831133929,
+      "learning_rate": 3.4600287614291704e-06,
+      "loss": 0.7268,
+      "step": 34060
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.81818192888927,
+      "learning_rate": 3.4584695074815556e-06,
+      "loss": 0.6954,
+      "step": 34065
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.056126713134114,
+      "learning_rate": 3.4569104191781934e-06,
+      "loss": 0.6938,
+      "step": 34070
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.97982971155647,
+      "learning_rate": 3.4553514966866187e-06,
+      "loss": 0.7434,
+      "step": 34075
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.0804405657430225,
+      "learning_rate": 3.4537927401743408e-06,
+      "loss": 0.6989,
+      "step": 34080
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.232370335315725,
+      "learning_rate": 3.4522341498088596e-06,
+      "loss": 0.7343,
+      "step": 34085
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.6929324607892315,
+      "learning_rate": 3.4506757257576474e-06,
+      "loss": 0.6838,
+      "step": 34090
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.847684941651293,
+      "learning_rate": 3.4491174681881667e-06,
+      "loss": 0.7998,
+      "step": 34095
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.044830561057324,
+      "learning_rate": 3.4475593772678608e-06,
+      "loss": 0.6488,
+      "step": 34100
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.711354549085556,
+      "learning_rate": 3.446001453164148e-06,
+      "loss": 0.706,
+      "step": 34105
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.048989575698363,
+      "learning_rate": 3.4444436960444393e-06,
+      "loss": 0.6854,
+      "step": 34110
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.257284662689208,
+      "learning_rate": 3.4428861060761198e-06,
+      "loss": 0.6976,
+      "step": 34115
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 21.04206986966952,
+      "learning_rate": 3.4413286834265615e-06,
+      "loss": 0.7153,
+      "step": 34120
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.270449177250299,
+      "learning_rate": 3.4397714282631157e-06,
+      "loss": 0.7192,
+      "step": 34125
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.781081714048268,
+      "learning_rate": 3.4382143407531126e-06,
+      "loss": 0.7171,
+      "step": 34130
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.091730267944795,
+      "learning_rate": 3.436657421063873e-06,
+      "loss": 0.7316,
+      "step": 34135
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.933812731737008,
+      "learning_rate": 3.4351006693626933e-06,
+      "loss": 0.702,
+      "step": 34140
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.372362762285348,
+      "learning_rate": 3.4335440858168523e-06,
+      "loss": 0.7398,
+      "step": 34145
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.8455483710133995,
+      "learning_rate": 3.431987670593613e-06,
+      "loss": 0.7201,
+      "step": 34150
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.572167621918964,
+      "learning_rate": 3.4304314238602195e-06,
+      "loss": 0.7274,
+      "step": 34155
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.318767412363042,
+      "learning_rate": 3.4288753457838965e-06,
+      "loss": 0.7392,
+      "step": 34160
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.545359597527433,
+      "learning_rate": 3.4273194365318494e-06,
+      "loss": 0.7101,
+      "step": 34165
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.267444881515774,
+      "learning_rate": 3.425763696271273e-06,
+      "loss": 0.7128,
+      "step": 34170
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.28827760130778,
+      "learning_rate": 3.4242081251693315e-06,
+      "loss": 0.6922,
+      "step": 34175
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.857576972315481,
+      "learning_rate": 3.422652723393184e-06,
+      "loss": 0.7162,
+      "step": 34180
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.239104105271066,
+      "learning_rate": 3.421097491109962e-06,
+      "loss": 0.7343,
+      "step": 34185
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.458692609742278,
+      "learning_rate": 3.4195424284867816e-06,
+      "loss": 0.7026,
+      "step": 34190
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.642379913819716,
+      "learning_rate": 3.4179875356907443e-06,
+      "loss": 0.7026,
+      "step": 34195
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.044826048221263,
+      "learning_rate": 3.4164328128889264e-06,
+      "loss": 0.6843,
+      "step": 34200
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.799607512515753,
+      "learning_rate": 3.4148782602483933e-06,
+      "loss": 0.6681,
+      "step": 34205
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.921927178432561,
+      "learning_rate": 3.4133238779361867e-06,
+      "loss": 0.6966,
+      "step": 34210
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.056524593418422,
+      "learning_rate": 3.41176966611933e-06,
+      "loss": 0.7182,
+      "step": 34215
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.086715413497012,
+      "learning_rate": 3.4102156249648344e-06,
+      "loss": 0.6893,
+      "step": 34220
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.818002941235232,
+      "learning_rate": 3.4086617546396844e-06,
+      "loss": 0.7129,
+      "step": 34225
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.5628820242695225,
+      "learning_rate": 3.4071080553108527e-06,
+      "loss": 0.7204,
+      "step": 34230
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.692911681535861,
+      "learning_rate": 3.4055545271452907e-06,
+      "loss": 0.7289,
+      "step": 34235
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.05193984747011,
+      "learning_rate": 3.4040011703099308e-06,
+      "loss": 0.6832,
+      "step": 34240
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.04610368711286,
+      "learning_rate": 3.40244798497169e-06,
+      "loss": 0.7035,
+      "step": 34245
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.981004110026776,
+      "learning_rate": 3.400894971297462e-06,
+      "loss": 0.7044,
+      "step": 34250
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.02871742649193,
+      "learning_rate": 3.399342129454129e-06,
+      "loss": 0.673,
+      "step": 34255
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.0891710187106565,
+      "learning_rate": 3.397789459608546e-06,
+      "loss": 0.6471,
+      "step": 34260
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.509849143862967,
+      "learning_rate": 3.3962369619275594e-06,
+      "loss": 0.7028,
+      "step": 34265
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.21109221125414,
+      "learning_rate": 3.3946846365779874e-06,
+      "loss": 0.7047,
+      "step": 34270
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.069199609629398,
+      "learning_rate": 3.3931324837266356e-06,
+      "loss": 0.6802,
+      "step": 34275
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.995862586321246,
+      "learning_rate": 3.391580503540291e-06,
+      "loss": 0.6839,
+      "step": 34280
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.225962015890741,
+      "learning_rate": 3.390028696185718e-06,
+      "loss": 0.7228,
+      "step": 34285
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.7826968736433955,
+      "learning_rate": 3.3884770618296696e-06,
+      "loss": 0.6992,
+      "step": 34290
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.215579942389127,
+      "learning_rate": 3.3869256006388727e-06,
+      "loss": 0.6615,
+      "step": 34295
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.8171014225910485,
+      "learning_rate": 3.3853743127800375e-06,
+      "loss": 0.6712,
+      "step": 34300
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.154691343420838,
+      "learning_rate": 3.38382319841986e-06,
+      "loss": 0.712,
+      "step": 34305
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.728149003069718,
+      "learning_rate": 3.382272257725011e-06,
+      "loss": 0.6618,
+      "step": 34310
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.180089532624329,
+      "learning_rate": 3.380721490862149e-06,
+      "loss": 0.7202,
+      "step": 34315
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.852505187684544,
+      "learning_rate": 3.3791708979979073e-06,
+      "loss": 0.7027,
+      "step": 34320
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.424161113556859,
+      "learning_rate": 3.3776204792989064e-06,
+      "loss": 0.7438,
+      "step": 34325
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.969015351711339,
+      "learning_rate": 3.3760702349317464e-06,
+      "loss": 0.6872,
+      "step": 34330
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.092021792098919,
+      "learning_rate": 3.3745201650630055e-06,
+      "loss": 0.7189,
+      "step": 34335
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.83639774476023,
+      "learning_rate": 3.3729702698592478e-06,
+      "loss": 0.7595,
+      "step": 34340
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.044030685192272,
+      "learning_rate": 3.3714205494870134e-06,
+      "loss": 0.7027,
+      "step": 34345
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.060222704112878,
+      "learning_rate": 3.3698710041128303e-06,
+      "loss": 0.6926,
+      "step": 34350
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.331748524456981,
+      "learning_rate": 3.3683216339032023e-06,
+      "loss": 0.6764,
+      "step": 34355
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.327864521187992,
+      "learning_rate": 3.3667724390246137e-06,
+      "loss": 0.7165,
+      "step": 34360
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.99815157543103,
+      "learning_rate": 3.3652234196435357e-06,
+      "loss": 0.7129,
+      "step": 34365
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.466694965390953,
+      "learning_rate": 3.3636745759264155e-06,
+      "loss": 0.7277,
+      "step": 34370
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.379631250627289,
+      "learning_rate": 3.3621259080396847e-06,
+      "loss": 0.7197,
+      "step": 34375
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.017399773871821,
+      "learning_rate": 3.360577416149754e-06,
+      "loss": 0.7095,
+      "step": 34380
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.37714353547322,
+      "learning_rate": 3.3590291004230134e-06,
+      "loss": 0.7209,
+      "step": 34385
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.21106889210857,
+      "learning_rate": 3.3574809610258403e-06,
+      "loss": 0.7217,
+      "step": 34390
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.020769066940849,
+      "learning_rate": 3.3559329981245845e-06,
+      "loss": 0.6942,
+      "step": 34395
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.535013418423105,
+      "learning_rate": 3.3543852118855863e-06,
+      "loss": 0.7216,
+      "step": 34400
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.833457039317395,
+      "learning_rate": 3.3528376024751572e-06,
+      "loss": 0.6834,
+      "step": 34405
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.466722213019253,
+      "learning_rate": 3.3512901700596e-06,
+      "loss": 0.687,
+      "step": 34410
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.699603331606502,
+      "learning_rate": 3.3497429148051886e-06,
+      "loss": 0.7176,
+      "step": 34415
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 14.078155850945397,
+      "learning_rate": 3.348195836878183e-06,
+      "loss": 0.7339,
+      "step": 34420
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.371680160155005,
+      "learning_rate": 3.3466489364448273e-06,
+      "loss": 0.7319,
+      "step": 34425
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.967862303147182,
+      "learning_rate": 3.3451022136713387e-06,
+      "loss": 0.7241,
+      "step": 34430
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.160505088786074,
+      "learning_rate": 3.343555668723922e-06,
+      "loss": 0.7006,
+      "step": 34435
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.370597144765361,
+      "learning_rate": 3.342009301768759e-06,
+      "loss": 0.6614,
+      "step": 34440
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.908732261224134,
+      "learning_rate": 3.3404631129720123e-06,
+      "loss": 0.728,
+      "step": 34445
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.464322679582324,
+      "learning_rate": 3.3389171024998302e-06,
+      "loss": 0.7296,
+      "step": 34450
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.446735218347529,
+      "learning_rate": 3.3373712705183335e-06,
+      "loss": 0.719,
+      "step": 34455
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.430665155686357,
+      "learning_rate": 3.3358256171936334e-06,
+      "loss": 0.7079,
+      "step": 34460
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.07094148537387,
+      "learning_rate": 3.334280142691816e-06,
+      "loss": 0.7402,
+      "step": 34465
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.958396301774057,
+      "learning_rate": 3.3327348471789465e-06,
+      "loss": 0.669,
+      "step": 34470
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.3002737488699205,
+      "learning_rate": 3.3311897308210784e-06,
+      "loss": 0.6967,
+      "step": 34475
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.901253329894225,
+      "learning_rate": 3.3296447937842365e-06,
+      "loss": 0.7611,
+      "step": 34480
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.409048107446193,
+      "learning_rate": 3.328100036234435e-06,
+      "loss": 0.6883,
+      "step": 34485
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.825450357177326,
+      "learning_rate": 3.3265554583376613e-06,
+      "loss": 0.7539,
+      "step": 34490
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.540915432854963,
+      "learning_rate": 3.3250110602598915e-06,
+      "loss": 0.7062,
+      "step": 34495
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.981442890497445,
+      "learning_rate": 3.3234668421670745e-06,
+      "loss": 0.6951,
+      "step": 34500
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.436484307213503,
+      "learning_rate": 3.3219228042251444e-06,
+      "loss": 0.6662,
+      "step": 34505
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.757233251489815,
+      "learning_rate": 3.3203789466000145e-06,
+      "loss": 0.6971,
+      "step": 34510
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.516152299769269,
+      "learning_rate": 3.318835269457579e-06,
+      "loss": 0.701,
+      "step": 34515
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.311327387586023,
+      "learning_rate": 3.3172917729637153e-06,
+      "loss": 0.6577,
+      "step": 34520
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.142789565685974,
+      "learning_rate": 3.3157484572842767e-06,
+      "loss": 0.6826,
+      "step": 34525
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.767767972624377,
+      "learning_rate": 3.3142053225850977e-06,
+      "loss": 0.6983,
+      "step": 34530
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.93963829978935,
+      "learning_rate": 3.312662369031999e-06,
+      "loss": 0.7298,
+      "step": 34535
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.955742890723362,
+      "learning_rate": 3.311119596790773e-06,
+      "loss": 0.7179,
+      "step": 34540
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.4240534447418,
+      "learning_rate": 3.3095770060272027e-06,
+      "loss": 0.6544,
+      "step": 34545
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.177569405537723,
+      "learning_rate": 3.3080345969070406e-06,
+      "loss": 0.7166,
+      "step": 34550
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.40433047937323,
+      "learning_rate": 3.3064923695960293e-06,
+      "loss": 0.6646,
+      "step": 34555
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.958549895696814,
+      "learning_rate": 3.304950324259889e-06,
+      "loss": 0.6739,
+      "step": 34560
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.718905234860555,
+      "learning_rate": 3.303408461064314e-06,
+      "loss": 0.6344,
+      "step": 34565
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.886058161844482,
+      "learning_rate": 3.3018667801749905e-06,
+      "loss": 0.7017,
+      "step": 34570
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.861357601286246,
+      "learning_rate": 3.300325281757573e-06,
+      "loss": 0.7016,
+      "step": 34575
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.902415222533037,
+      "learning_rate": 3.2987839659777075e-06,
+      "loss": 0.7254,
+      "step": 34580
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.397106547037049,
+      "learning_rate": 3.2972428330010133e-06,
+      "loss": 0.7209,
+      "step": 34585
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.233412595583847,
+      "learning_rate": 3.295701882993089e-06,
+      "loss": 0.7159,
+      "step": 34590
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.898698516838979,
+      "learning_rate": 3.29416111611952e-06,
+      "loss": 0.7264,
+      "step": 34595
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.151982890610785,
+      "learning_rate": 3.292620532545868e-06,
+      "loss": 0.678,
+      "step": 34600
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.847001237003898,
+      "learning_rate": 3.291080132437674e-06,
+      "loss": 0.7123,
+      "step": 34605
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.519189860016127,
+      "learning_rate": 3.2895399159604643e-06,
+      "loss": 0.7116,
+      "step": 34610
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.0700122522342514,
+      "learning_rate": 3.2879998832797367e-06,
+      "loss": 0.6714,
+      "step": 34615
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.656954353614141,
+      "learning_rate": 3.2864600345609788e-06,
+      "loss": 0.7002,
+      "step": 34620
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.060961806551473,
+      "learning_rate": 3.284920369969651e-06,
+      "loss": 0.6637,
+      "step": 34625
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 27.81425414714497,
+      "learning_rate": 3.283380889671201e-06,
+      "loss": 0.7247,
+      "step": 34630
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.35180489889317,
+      "learning_rate": 3.2818415938310482e-06,
+      "loss": 0.6577,
+      "step": 34635
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.20720735762461,
+      "learning_rate": 3.2803024826146012e-06,
+      "loss": 0.7122,
+      "step": 34640
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.968011892621076,
+      "learning_rate": 3.2787635561872407e-06,
+      "loss": 0.6988,
+      "step": 34645
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.627838913286878,
+      "learning_rate": 3.2772248147143314e-06,
+      "loss": 0.6984,
+      "step": 34650
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 20.29426359885549,
+      "learning_rate": 3.2756862583612216e-06,
+      "loss": 0.6838,
+      "step": 34655
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.206233227106853,
+      "learning_rate": 3.2741478872932304e-06,
+      "loss": 0.6747,
+      "step": 34660
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.49570254513729,
+      "learning_rate": 3.2726097016756673e-06,
+      "loss": 0.6961,
+      "step": 34665
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.765673366068342,
+      "learning_rate": 3.271071701673817e-06,
+      "loss": 0.6956,
+      "step": 34670
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 31.278675608315666,
+      "learning_rate": 3.269533887452939e-06,
+      "loss": 0.6775,
+      "step": 34675
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.296567576866515,
+      "learning_rate": 3.2679962591782843e-06,
+      "loss": 0.7264,
+      "step": 34680
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.480026200208731,
+      "learning_rate": 3.266458817015074e-06,
+      "loss": 0.7041,
+      "step": 34685
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.8713397591813,
+      "learning_rate": 3.2649215611285157e-06,
+      "loss": 0.7449,
+      "step": 34690
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.996493401901696,
+      "learning_rate": 3.2633844916837933e-06,
+      "loss": 0.7304,
+      "step": 34695
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.3405845221110715,
+      "learning_rate": 3.26184760884607e-06,
+      "loss": 0.7047,
+      "step": 34700
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.21416176409067,
+      "learning_rate": 3.260310912780494e-06,
+      "loss": 0.7197,
+      "step": 34705
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.478343052250641,
+      "learning_rate": 3.258774403652186e-06,
+      "loss": 0.6853,
+      "step": 34710
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.736976102011249,
+      "learning_rate": 3.2572380816262554e-06,
+      "loss": 0.6808,
+      "step": 34715
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.847908944633629,
+      "learning_rate": 3.255701946867782e-06,
+      "loss": 0.6935,
+      "step": 34720
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.995112079997536,
+      "learning_rate": 3.2541659995418352e-06,
+      "loss": 0.654,
+      "step": 34725
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.432544965454138,
+      "learning_rate": 3.252630239813455e-06,
+      "loss": 0.6921,
+      "step": 34730
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.41514470068817,
+      "learning_rate": 3.2510946678476677e-06,
+      "loss": 0.6501,
+      "step": 34735
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.04100676135012,
+      "learning_rate": 3.2495592838094768e-06,
+      "loss": 0.6964,
+      "step": 34740
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.132702607976753,
+      "learning_rate": 3.2480240878638646e-06,
+      "loss": 0.6656,
+      "step": 34745
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.984308471395036,
+      "learning_rate": 3.2464890801757987e-06,
+      "loss": 0.7329,
+      "step": 34750
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.37083283436118,
+      "learning_rate": 3.2449542609102203e-06,
+      "loss": 0.6692,
+      "step": 34755
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.065891822018525,
+      "learning_rate": 3.2434196302320498e-06,
+      "loss": 0.6915,
+      "step": 34760
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.506466007238628,
+      "learning_rate": 3.2418851883061942e-06,
+      "loss": 0.6818,
+      "step": 34765
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.540527651253667,
+      "learning_rate": 3.2403509352975325e-06,
+      "loss": 0.7146,
+      "step": 34770
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.253856929743469,
+      "learning_rate": 3.238816871370931e-06,
+      "loss": 0.6754,
+      "step": 34775
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.973745054767074,
+      "learning_rate": 3.2372829966912276e-06,
+      "loss": 0.7016,
+      "step": 34780
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.4143701951065,
+      "learning_rate": 3.235749311423245e-06,
+      "loss": 0.7051,
+      "step": 34785
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.256718611345136,
+      "learning_rate": 3.2342158157317875e-06,
+      "loss": 0.7005,
+      "step": 34790
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.430052768862708,
+      "learning_rate": 3.232682509781632e-06,
+      "loss": 0.6717,
+      "step": 34795
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.122382147228777,
+      "learning_rate": 3.2311493937375414e-06,
+      "loss": 0.7403,
+      "step": 34800
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.850098024675329,
+      "learning_rate": 3.2296164677642538e-06,
+      "loss": 0.7054,
+      "step": 34805
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.928735150390296,
+      "learning_rate": 3.2280837320264913e-06,
+      "loss": 0.7012,
+      "step": 34810
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.971053368580964,
+      "learning_rate": 3.2265511866889528e-06,
+      "loss": 0.7092,
+      "step": 34815
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.383201044098856,
+      "learning_rate": 3.2250188319163133e-06,
+      "loss": 0.6976,
+      "step": 34820
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.725235772265117,
+      "learning_rate": 3.2234866678732356e-06,
+      "loss": 0.6857,
+      "step": 34825
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.891194956713214,
+      "learning_rate": 3.221954694724355e-06,
+      "loss": 0.6958,
+      "step": 34830
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.258838383391693,
+      "learning_rate": 3.2204229126342903e-06,
+      "loss": 0.6743,
+      "step": 34835
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 16.058831689741183,
+      "learning_rate": 3.2188913217676384e-06,
+      "loss": 0.7022,
+      "step": 34840
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.434261438384548,
+      "learning_rate": 3.2173599222889716e-06,
+      "loss": 0.7183,
+      "step": 34845
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.054370650282337,
+      "learning_rate": 3.215828714362852e-06,
+      "loss": 0.7057,
+      "step": 34850
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.774936223876148,
+      "learning_rate": 3.2142976981538087e-06,
+      "loss": 0.6832,
+      "step": 34855
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.153082902012476,
+      "learning_rate": 3.2127668738263607e-06,
+      "loss": 0.7129,
+      "step": 34860
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.438502406162897,
+      "learning_rate": 3.2112362415449983e-06,
+      "loss": 0.7127,
+      "step": 34865
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.687628359593135,
+      "learning_rate": 3.2097058014741963e-06,
+      "loss": 0.6941,
+      "step": 34870
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.208079044546652,
+      "learning_rate": 3.208175553778407e-06,
+      "loss": 0.6832,
+      "step": 34875
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.083614473178835,
+      "learning_rate": 3.206645498622062e-06,
+      "loss": 0.735,
+      "step": 34880
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.662208474418582,
+      "learning_rate": 3.205115636169574e-06,
+      "loss": 0.6765,
+      "step": 34885
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.730622620037803,
+      "learning_rate": 3.203585966585331e-06,
+      "loss": 0.7013,
+      "step": 34890
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.724229039966694,
+      "learning_rate": 3.2020564900337057e-06,
+      "loss": 0.7409,
+      "step": 34895
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.749165581230447,
+      "learning_rate": 3.200527206679046e-06,
+      "loss": 0.6501,
+      "step": 34900
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.376249727339917,
+      "learning_rate": 3.1989981166856767e-06,
+      "loss": 0.6926,
+      "step": 34905
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.746816413591672,
+      "learning_rate": 3.1974692202179107e-06,
+      "loss": 0.7186,
+      "step": 34910
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 6.3990200178856345,
+      "learning_rate": 3.1959405174400303e-06,
+      "loss": 0.7007,
+      "step": 34915
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.916186341656027,
+      "learning_rate": 3.1944120085163045e-06,
+      "loss": 0.7204,
+      "step": 34920
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.14631137566495,
+      "learning_rate": 3.1928836936109776e-06,
+      "loss": 0.7184,
+      "step": 34925
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 6.220870589885455,
+      "learning_rate": 3.1913555728882706e-06,
+      "loss": 0.6801,
+      "step": 34930
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.683925784817751,
+      "learning_rate": 3.1898276465123923e-06,
+      "loss": 0.6963,
+      "step": 34935
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.498276001612888,
+      "learning_rate": 3.18829991464752e-06,
+      "loss": 0.7236,
+      "step": 34940
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 6.310380146467167,
+      "learning_rate": 3.18677237745782e-06,
+      "loss": 0.643,
+      "step": 34945
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.183976333866489,
+      "learning_rate": 3.18524503510743e-06,
+      "loss": 0.6871,
+      "step": 34950
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.843396474345437,
+      "learning_rate": 3.183717887760468e-06,
+      "loss": 0.7395,
+      "step": 34955
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.8160977151335,
+      "learning_rate": 3.182190935581036e-06,
+      "loss": 0.7199,
+      "step": 34960
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.728666747082901,
+      "learning_rate": 3.1806641787332105e-06,
+      "loss": 0.6712,
+      "step": 34965
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.3161951589156775,
+      "learning_rate": 3.1791376173810485e-06,
+      "loss": 0.7148,
+      "step": 34970
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.649087116983282,
+      "learning_rate": 3.177611251688585e-06,
+      "loss": 0.6711,
+      "step": 34975
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.358917128358095,
+      "learning_rate": 3.1760850818198364e-06,
+      "loss": 0.7179,
+      "step": 34980
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.315739908454177,
+      "learning_rate": 3.174559107938796e-06,
+      "loss": 0.6969,
+      "step": 34985
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.721395234434675,
+      "learning_rate": 3.173033330209433e-06,
+      "loss": 0.6818,
+      "step": 34990
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.9793097019291315,
+      "learning_rate": 3.1715077487957046e-06,
+      "loss": 0.6921,
+      "step": 34995
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.932958413519551,
+      "learning_rate": 3.1699823638615356e-06,
+      "loss": 0.7227,
+      "step": 35000
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 6.266642265568164,
+      "learning_rate": 3.1684571755708405e-06,
+      "loss": 0.6626,
+      "step": 35005
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.923864184283424,
+      "learning_rate": 3.1669321840875036e-06,
+      "loss": 0.6955,
+      "step": 35010
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.773322611824129,
+      "learning_rate": 3.1654073895753927e-06,
+      "loss": 0.6571,
+      "step": 35015
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.894107093078364,
+      "learning_rate": 3.163882792198357e-06,
+      "loss": 0.7053,
+      "step": 35020
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.90151505620469,
+      "learning_rate": 3.1623583921202163e-06,
+      "loss": 0.6834,
+      "step": 35025
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.962497029735831,
+      "learning_rate": 3.1608341895047784e-06,
+      "loss": 0.6902,
+      "step": 35030
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.26088204113902,
+      "learning_rate": 3.1593101845158215e-06,
+      "loss": 0.718,
+      "step": 35035
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.144310418494661,
+      "learning_rate": 3.1577863773171114e-06,
+      "loss": 0.6717,
+      "step": 35040
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.875648665970789,
+      "learning_rate": 3.156262768072385e-06,
+      "loss": 0.6907,
+      "step": 35045
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.664963596240274,
+      "learning_rate": 3.1547393569453596e-06,
+      "loss": 0.6864,
+      "step": 35050
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.27004589472345,
+      "learning_rate": 3.1532161440997346e-06,
+      "loss": 0.7028,
+      "step": 35055
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.323651522139277,
+      "learning_rate": 3.1516931296991852e-06,
+      "loss": 0.682,
+      "step": 35060
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.790824661300352,
+      "learning_rate": 3.150170313907367e-06,
+      "loss": 0.677,
+      "step": 35065
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.8124891660377775,
+      "learning_rate": 3.1486476968879133e-06,
+      "loss": 0.648,
+      "step": 35070
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.709249276514523,
+      "learning_rate": 3.1471252788044326e-06,
+      "loss": 0.7221,
+      "step": 35075
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.5922070158533845,
+      "learning_rate": 3.14560305982052e-06,
+      "loss": 0.6912,
+      "step": 35080
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.052275705629103,
+      "learning_rate": 3.1440810400997415e-06,
+      "loss": 0.7063,
+      "step": 35085
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.614530670321845,
+      "learning_rate": 3.1425592198056475e-06,
+      "loss": 0.6681,
+      "step": 35090
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.213499876329178,
+      "learning_rate": 3.1410375991017623e-06,
+      "loss": 0.6812,
+      "step": 35095
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.39707774006819,
+      "learning_rate": 3.1395161781515907e-06,
+      "loss": 0.6478,
+      "step": 35100
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.183880018678881,
+      "learning_rate": 3.1379949571186173e-06,
+      "loss": 0.7054,
+      "step": 35105
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.790529024945295,
+      "learning_rate": 3.136473936166302e-06,
+      "loss": 0.7118,
+      "step": 35110
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.790714271544594,
+      "learning_rate": 3.1349531154580885e-06,
+      "loss": 0.6785,
+      "step": 35115
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.363396283304108,
+      "learning_rate": 3.1334324951573925e-06,
+      "loss": 0.6911,
+      "step": 35120
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.67172662481413,
+      "learning_rate": 3.131912075427614e-06,
+      "loss": 0.7088,
+      "step": 35125
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.1825079244181635,
+      "learning_rate": 3.1303918564321276e-06,
+      "loss": 0.6561,
+      "step": 35130
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.127097794226152,
+      "learning_rate": 3.1288718383342853e-06,
+      "loss": 0.6708,
+      "step": 35135
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.517761463954615,
+      "learning_rate": 3.1273520212974244e-06,
+      "loss": 0.6742,
+      "step": 35140
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.768996000200358,
+      "learning_rate": 3.125832405484851e-06,
+      "loss": 0.7208,
+      "step": 35145
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.54907973874482,
+      "learning_rate": 3.1243129910598587e-06,
+      "loss": 0.6833,
+      "step": 35150
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.040848945756802,
+      "learning_rate": 3.1227937781857137e-06,
+      "loss": 0.6692,
+      "step": 35155
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.359439672587436,
+      "learning_rate": 3.1212747670256606e-06,
+      "loss": 0.6849,
+      "step": 35160
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.601260479168754,
+      "learning_rate": 3.119755957742927e-06,
+      "loss": 0.666,
+      "step": 35165
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.818423116929544,
+      "learning_rate": 3.1182373505007115e-06,
+      "loss": 0.7284,
+      "step": 35170
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.28760642783991,
+      "learning_rate": 3.1167189454622e-06,
+      "loss": 0.6949,
+      "step": 35175
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.919285582455202,
+      "learning_rate": 3.1152007427905496e-06,
+      "loss": 0.6965,
+      "step": 35180
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.602421231140983,
+      "learning_rate": 3.113682742648895e-06,
+      "loss": 0.6468,
+      "step": 35185
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.18047545646941,
+      "learning_rate": 3.1121649452003553e-06,
+      "loss": 0.6905,
+      "step": 35190
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.692723737245233,
+      "learning_rate": 3.1106473506080248e-06,
+      "loss": 0.6641,
+      "step": 35195
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.089993644237553,
+      "learning_rate": 3.1091299590349744e-06,
+      "loss": 0.7022,
+      "step": 35200
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 18.464348102001022,
+      "learning_rate": 3.107612770644254e-06,
+      "loss": 0.7242,
+      "step": 35205
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.417910039588145,
+      "learning_rate": 3.1060957855988956e-06,
+      "loss": 0.6852,
+      "step": 35210
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.911349146203593,
+      "learning_rate": 3.104579004061903e-06,
+      "loss": 0.6824,
+      "step": 35215
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.9585363784038154,
+      "learning_rate": 3.1030624261962595e-06,
+      "loss": 0.7463,
+      "step": 35220
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.927299731693227,
+      "learning_rate": 3.101546052164932e-06,
+      "loss": 0.7165,
+      "step": 35225
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 18.82840268503806,
+      "learning_rate": 3.1000298821308583e-06,
+      "loss": 0.6777,
+      "step": 35230
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 26.543236627565317,
+      "learning_rate": 3.098513916256961e-06,
+      "loss": 0.7074,
+      "step": 35235
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 18.053590209899767,
+      "learning_rate": 3.0969981547061333e-06,
+      "loss": 0.7189,
+      "step": 35240
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.09854583103734,
+      "learning_rate": 3.0954825976412528e-06,
+      "loss": 0.7315,
+      "step": 35245
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.441055497384916,
+      "learning_rate": 3.0939672452251736e-06,
+      "loss": 0.7075,
+      "step": 35250
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.874551333299127,
+      "learning_rate": 3.0924520976207238e-06,
+      "loss": 0.7354,
+      "step": 35255
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.548203979119037,
+      "learning_rate": 3.0909371549907175e-06,
+      "loss": 0.7256,
+      "step": 35260
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 16.1006428562444,
+      "learning_rate": 3.089422417497938e-06,
+      "loss": 0.7272,
+      "step": 35265
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.063491743642588,
+      "learning_rate": 3.0879078853051506e-06,
+      "loss": 0.6953,
+      "step": 35270
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.957301907855168,
+      "learning_rate": 3.0863935585751014e-06,
+      "loss": 0.6526,
+      "step": 35275
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.94978234221093,
+      "learning_rate": 3.0848794374705083e-06,
+      "loss": 0.7051,
+      "step": 35280
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.441672951653837,
+      "learning_rate": 3.083365522154072e-06,
+      "loss": 0.6943,
+      "step": 35285
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.792613427444279,
+      "learning_rate": 3.0818518127884693e-06,
+      "loss": 0.6771,
+      "step": 35290
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.209539089258062,
+      "learning_rate": 3.0803383095363547e-06,
+      "loss": 0.6916,
+      "step": 35295
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.5721290185461605,
+      "learning_rate": 3.0788250125603624e-06,
+      "loss": 0.6846,
+      "step": 35300
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.630528358446945,
+      "learning_rate": 3.0773119220230986e-06,
+      "loss": 0.6525,
+      "step": 35305
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.930745118257468,
+      "learning_rate": 3.0757990380871567e-06,
+      "loss": 0.6725,
+      "step": 35310
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.407960926449672,
+      "learning_rate": 3.074286360915098e-06,
+      "loss": 0.6855,
+      "step": 35315
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.699747509500149,
+      "learning_rate": 3.0727738906694703e-06,
+      "loss": 0.6812,
+      "step": 35320
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.908081523308288,
+      "learning_rate": 3.0712616275127937e-06,
+      "loss": 0.6618,
+      "step": 35325
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.064774130169687,
+      "learning_rate": 3.0697495716075665e-06,
+      "loss": 0.6746,
+      "step": 35330
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.04715897242109,
+      "learning_rate": 3.068237723116266e-06,
+      "loss": 0.722,
+      "step": 35335
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.107756854479823,
+      "learning_rate": 3.066726082201348e-06,
+      "loss": 0.6685,
+      "step": 35340
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.940410971828386,
+      "learning_rate": 3.0652146490252466e-06,
+      "loss": 0.6744,
+      "step": 35345
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.438299431843944,
+      "learning_rate": 3.0637034237503694e-06,
+      "loss": 0.6887,
+      "step": 35350
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.286761988035629,
+      "learning_rate": 3.062192406539103e-06,
+      "loss": 0.6548,
+      "step": 35355
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.1315447819054265,
+      "learning_rate": 3.0606815975538166e-06,
+      "loss": 0.6672,
+      "step": 35360
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.546797293429852,
+      "learning_rate": 3.0591709969568496e-06,
+      "loss": 0.6797,
+      "step": 35365
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.893488072686185,
+      "learning_rate": 3.0576606049105267e-06,
+      "loss": 0.711,
+      "step": 35370
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.333598157410917,
+      "learning_rate": 3.056150421577142e-06,
+      "loss": 0.7055,
+      "step": 35375
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.630504334203577,
+      "learning_rate": 3.054640447118974e-06,
+      "loss": 0.6762,
+      "step": 35380
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.818021257760977,
+      "learning_rate": 3.0531306816982764e-06,
+      "loss": 0.6943,
+      "step": 35385
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.459984953599493,
+      "learning_rate": 3.0516211254772777e-06,
+      "loss": 0.6862,
+      "step": 35390
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.05717280742564,
+      "learning_rate": 3.05011177861819e-06,
+      "loss": 0.6932,
+      "step": 35395
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.159867206313164,
+      "learning_rate": 3.048602641283195e-06,
+      "loss": 0.6862,
+      "step": 35400
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.700555157614799,
+      "learning_rate": 3.04709371363446e-06,
+      "loss": 0.6797,
+      "step": 35405
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.504531105194354,
+      "learning_rate": 3.045584995834124e-06,
+      "loss": 0.7213,
+      "step": 35410
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.583940979099118,
+      "learning_rate": 3.0440764880443045e-06,
+      "loss": 0.6655,
+      "step": 35415
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.2446812458027345,
+      "learning_rate": 3.0425681904270993e-06,
+      "loss": 0.6685,
+      "step": 35420
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.98702038239393,
+      "learning_rate": 3.0410601031445807e-06,
+      "loss": 0.6659,
+      "step": 35425
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.30765691435782,
+      "learning_rate": 3.0395522263587996e-06,
+      "loss": 0.6463,
+      "step": 35430
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.507169948067107,
+      "learning_rate": 3.038044560231785e-06,
+      "loss": 0.697,
+      "step": 35435
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.80795821968365,
+      "learning_rate": 3.036537104925539e-06,
+      "loss": 0.6565,
+      "step": 35440
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.580353059297831,
+      "learning_rate": 3.035029860602049e-06,
+      "loss": 0.6895,
+      "step": 35445
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.90561869904271,
+      "learning_rate": 3.0335228274232697e-06,
+      "loss": 0.7017,
+      "step": 35450
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.256834525834763,
+      "learning_rate": 3.032016005551144e-06,
+      "loss": 0.6792,
+      "step": 35455
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.568160748735181,
+      "learning_rate": 3.030509395147581e-06,
+      "loss": 0.6821,
+      "step": 35460
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.40151804531815,
+      "learning_rate": 3.0290029963744783e-06,
+      "loss": 0.6535,
+      "step": 35465
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.454250046103105,
+      "learning_rate": 3.0274968093937013e-06,
+      "loss": 0.6958,
+      "step": 35470
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.2595836898954405,
+      "learning_rate": 3.0259908343670964e-06,
+      "loss": 0.6589,
+      "step": 35475
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.171782526094596,
+      "learning_rate": 3.02448507145649e-06,
+      "loss": 0.6818,
+      "step": 35480
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.538920031530736,
+      "learning_rate": 3.0229795208236793e-06,
+      "loss": 0.6603,
+      "step": 35485
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.316265528218628,
+      "learning_rate": 3.0214741826304466e-06,
+      "loss": 0.6717,
+      "step": 35490
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.807201792994373,
+      "learning_rate": 3.019969057038545e-06,
+      "loss": 0.6614,
+      "step": 35495
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.1088519096374512,
+      "eval_runtime": 25.3868,
+      "eval_samples_per_second": 31.749,
+      "eval_steps_per_second": 3.978,
+      "step": 35496
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.624162199485662,
+      "learning_rate": 3.018464144209704e-06,
+      "loss": 0.4827,
+      "step": 35500
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.251439545333477,
+      "learning_rate": 3.0169594443056386e-06,
+      "loss": 0.4616,
+      "step": 35505
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.592487760809804,
+      "learning_rate": 3.015454957488031e-06,
+      "loss": 0.422,
+      "step": 35510
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.445055475848834,
+      "learning_rate": 3.0139506839185473e-06,
+      "loss": 0.4291,
+      "step": 35515
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.73350203023127,
+      "learning_rate": 3.012446623758827e-06,
+      "loss": 0.4073,
+      "step": 35520
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.610318191731628,
+      "learning_rate": 3.0109427771704903e-06,
+      "loss": 0.4327,
+      "step": 35525
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.798895152834253,
+      "learning_rate": 3.0094391443151306e-06,
+      "loss": 0.414,
+      "step": 35530
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.8561901802041305,
+      "learning_rate": 3.007935725354317e-06,
+      "loss": 0.4212,
+      "step": 35535
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.573398710816627,
+      "learning_rate": 3.0064325204496046e-06,
+      "loss": 0.418,
+      "step": 35540
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.95415125209879,
+      "learning_rate": 3.004929529762514e-06,
+      "loss": 0.4089,
+      "step": 35545
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.18703664328013,
+      "learning_rate": 3.0034267534545525e-06,
+      "loss": 0.4446,
+      "step": 35550
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.877676239718891,
+      "learning_rate": 3.0019241916871968e-06,
+      "loss": 0.438,
+      "step": 35555
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.423217455654268,
+      "learning_rate": 3.000421844621905e-06,
+      "loss": 0.425,
+      "step": 35560
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.086485665867204,
+      "learning_rate": 2.99891971242011e-06,
+      "loss": 0.4715,
+      "step": 35565
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.471798415265517,
+      "learning_rate": 2.9974177952432227e-06,
+      "loss": 0.4301,
+      "step": 35570
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.310256810777645,
+      "learning_rate": 2.9959160932526333e-06,
+      "loss": 0.437,
+      "step": 35575
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.945012475712052,
+      "learning_rate": 2.994414606609705e-06,
+      "loss": 0.4287,
+      "step": 35580
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.799289907826112,
+      "learning_rate": 2.9929133354757754e-06,
+      "loss": 0.4317,
+      "step": 35585
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.670616354730726,
+      "learning_rate": 2.9914122800121693e-06,
+      "loss": 0.436,
+      "step": 35590
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.938748044740877,
+      "learning_rate": 2.9899114403801754e-06,
+      "loss": 0.4419,
+      "step": 35595
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.820127001934945,
+      "learning_rate": 2.9884108167410703e-06,
+      "loss": 0.4261,
+      "step": 35600
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.644965024100297,
+      "learning_rate": 2.9869104092560993e-06,
+      "loss": 0.4271,
+      "step": 35605
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.087396027260219,
+      "learning_rate": 2.9854102180864895e-06,
+      "loss": 0.399,
+      "step": 35610
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.92111872490377,
+      "learning_rate": 2.9839102433934446e-06,
+      "loss": 0.4188,
+      "step": 35615
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.95032930360502,
+      "learning_rate": 2.9824104853381396e-06,
+      "loss": 0.4215,
+      "step": 35620
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.565444835470105,
+      "learning_rate": 2.980910944081734e-06,
+      "loss": 0.4269,
+      "step": 35625
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.299176971333008,
+      "learning_rate": 2.9794116197853564e-06,
+      "loss": 0.4459,
+      "step": 35630
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.964758881245675,
+      "learning_rate": 2.97791251261012e-06,
+      "loss": 0.4502,
+      "step": 35635
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.883842627489473,
+      "learning_rate": 2.976413622717108e-06,
+      "loss": 0.4043,
+      "step": 35640
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.6760575472965185,
+      "learning_rate": 2.974914950267381e-06,
+      "loss": 0.4433,
+      "step": 35645
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.589695042834639,
+      "learning_rate": 2.9734164954219812e-06,
+      "loss": 0.4467,
+      "step": 35650
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.199176387428726,
+      "learning_rate": 2.9719182583419233e-06,
+      "loss": 0.4263,
+      "step": 35655
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.4164360665117925,
+      "learning_rate": 2.9704202391881987e-06,
+      "loss": 0.4137,
+      "step": 35660
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.9496041243339395,
+      "learning_rate": 2.9689224381217774e-06,
+      "loss": 0.427,
+      "step": 35665
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.413073547077309,
+      "learning_rate": 2.9674248553036022e-06,
+      "loss": 0.4254,
+      "step": 35670
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.632442675004174,
+      "learning_rate": 2.965927490894599e-06,
+      "loss": 0.4133,
+      "step": 35675
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 7.153190155452668,
+      "learning_rate": 2.9644303450556617e-06,
+      "loss": 0.4454,
+      "step": 35680
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.114342055733055,
+      "learning_rate": 2.96293341794767e-06,
+      "loss": 0.4234,
+      "step": 35685
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.495183768494921,
+      "learning_rate": 2.96143670973147e-06,
+      "loss": 0.4171,
+      "step": 35690
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.216950966436851,
+      "learning_rate": 2.9599402205678945e-06,
+      "loss": 0.3834,
+      "step": 35695
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.836099168905307,
+      "learning_rate": 2.9584439506177453e-06,
+      "loss": 0.4115,
+      "step": 35700
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.9788850849651105,
+      "learning_rate": 2.956947900041802e-06,
+      "loss": 0.4315,
+      "step": 35705
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.335694047766549,
+      "learning_rate": 2.955452069000826e-06,
+      "loss": 0.4046,
+      "step": 35710
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.659384034868762,
+      "learning_rate": 2.9539564576555462e-06,
+      "loss": 0.4144,
+      "step": 35715
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.337211105652029,
+      "learning_rate": 2.952461066166677e-06,
+      "loss": 0.4186,
+      "step": 35720
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.139699583710478,
+      "learning_rate": 2.950965894694903e-06,
+      "loss": 0.451,
+      "step": 35725
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.061656896769589,
+      "learning_rate": 2.949470943400884e-06,
+      "loss": 0.41,
+      "step": 35730
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.797408797001112,
+      "learning_rate": 2.947976212445265e-06,
+      "loss": 0.4397,
+      "step": 35735
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 7.272477965943136,
+      "learning_rate": 2.946481701988655e-06,
+      "loss": 0.4468,
+      "step": 35740
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.313697508432313,
+      "learning_rate": 2.9449874121916504e-06,
+      "loss": 0.4348,
+      "step": 35745
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.078313122337126,
+      "learning_rate": 2.9434933432148194e-06,
+      "loss": 0.4323,
+      "step": 35750
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 15.740935713715533,
+      "learning_rate": 2.9419994952187013e-06,
+      "loss": 0.4118,
+      "step": 35755
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.6839709542325405,
+      "learning_rate": 2.9405058683638233e-06,
+      "loss": 0.4438,
+      "step": 35760
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 6.198213094334123,
+      "learning_rate": 2.939012462810677e-06,
+      "loss": 0.4368,
+      "step": 35765
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.0077983133807695,
+      "learning_rate": 2.9375192787197392e-06,
+      "loss": 0.427,
+      "step": 35770
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.8076979963613695,
+      "learning_rate": 2.9360263162514554e-06,
+      "loss": 0.3971,
+      "step": 35775
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.9645393002102365,
+      "learning_rate": 2.934533575566254e-06,
+      "loss": 0.4248,
+      "step": 35780
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.598701108429405,
+      "learning_rate": 2.933041056824536e-06,
+      "loss": 0.44,
+      "step": 35785
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.3368283017412175,
+      "learning_rate": 2.9315487601866777e-06,
+      "loss": 0.4002,
+      "step": 35790
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 6.185065483388467,
+      "learning_rate": 2.9300566858130343e-06,
+      "loss": 0.4707,
+      "step": 35795
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.604580349140249,
+      "learning_rate": 2.928564833863934e-06,
+      "loss": 0.4243,
+      "step": 35800
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.5612185838977135,
+      "learning_rate": 2.9270732044996864e-06,
+      "loss": 0.4131,
+      "step": 35805
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.813337345210239,
+      "learning_rate": 2.925581797880571e-06,
+      "loss": 0.4174,
+      "step": 35810
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.592818903498818,
+      "learning_rate": 2.9240906141668457e-06,
+      "loss": 0.4526,
+      "step": 35815
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.463786035143501,
+      "learning_rate": 2.9225996535187464e-06,
+      "loss": 0.3919,
+      "step": 35820
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.827176495670314,
+      "learning_rate": 2.9211089160964813e-06,
+      "loss": 0.4226,
+      "step": 35825
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.077025005993725,
+      "learning_rate": 2.9196184020602397e-06,
+      "loss": 0.4139,
+      "step": 35830
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.551503871376388,
+      "learning_rate": 2.9181281115701816e-06,
+      "loss": 0.4059,
+      "step": 35835
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 10.107936959906947,
+      "learning_rate": 2.9166380447864444e-06,
+      "loss": 0.4717,
+      "step": 35840
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.368268058379852,
+      "learning_rate": 2.915148201869146e-06,
+      "loss": 0.3893,
+      "step": 35845
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.909165442524433,
+      "learning_rate": 2.9136585829783727e-06,
+      "loss": 0.44,
+      "step": 35850
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.7662995337880565,
+      "learning_rate": 2.912169188274194e-06,
+      "loss": 0.4275,
+      "step": 35855
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.6140968907568825,
+      "learning_rate": 2.910680017916649e-06,
+      "loss": 0.419,
+      "step": 35860
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.5432678858739575,
+      "learning_rate": 2.9091910720657584e-06,
+      "loss": 0.4251,
+      "step": 35865
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.89283026920533,
+      "learning_rate": 2.9077023508815148e-06,
+      "loss": 0.4366,
+      "step": 35870
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.060170416042779,
+      "learning_rate": 2.9062138545238855e-06,
+      "loss": 0.4583,
+      "step": 35875
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.644859363623836,
+      "learning_rate": 2.9047255831528196e-06,
+      "loss": 0.4277,
+      "step": 35880
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.839867757789757,
+      "learning_rate": 2.903237536928236e-06,
+      "loss": 0.4145,
+      "step": 35885
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.422980378197918,
+      "learning_rate": 2.901749716010034e-06,
+      "loss": 0.4215,
+      "step": 35890
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.076494068953284,
+      "learning_rate": 2.9002621205580857e-06,
+      "loss": 0.4491,
+      "step": 35895
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.869737017386848,
+      "learning_rate": 2.8987747507322373e-06,
+      "loss": 0.4449,
+      "step": 35900
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.5067336712799815,
+      "learning_rate": 2.897287606692315e-06,
+      "loss": 0.461,
+      "step": 35905
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.730046760882599,
+      "learning_rate": 2.8958006885981216e-06,
+      "loss": 0.456,
+      "step": 35910
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.114492849485682,
+      "learning_rate": 2.894313996609429e-06,
+      "loss": 0.4181,
+      "step": 35915
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.616343883285859,
+      "learning_rate": 2.89282753088599e-06,
+      "loss": 0.4385,
+      "step": 35920
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.164659013183382,
+      "learning_rate": 2.891341291587535e-06,
+      "loss": 0.4289,
+      "step": 35925
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.5568870158173365,
+      "learning_rate": 2.8898552788737643e-06,
+      "loss": 0.4288,
+      "step": 35930
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.892143230641789,
+      "learning_rate": 2.8883694929043547e-06,
+      "loss": 0.4195,
+      "step": 35935
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.995254126052123,
+      "learning_rate": 2.8868839338389644e-06,
+      "loss": 0.4462,
+      "step": 35940
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.66834365046065,
+      "learning_rate": 2.8853986018372195e-06,
+      "loss": 0.427,
+      "step": 35945
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.984404586363791,
+      "learning_rate": 2.8839134970587284e-06,
+      "loss": 0.4164,
+      "step": 35950
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.299787020257302,
+      "learning_rate": 2.8824286196630712e-06,
+      "loss": 0.4321,
+      "step": 35955
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.275047001853112,
+      "learning_rate": 2.880943969809803e-06,
+      "loss": 0.4337,
+      "step": 35960
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.162259568469418,
+      "learning_rate": 2.8794595476584597e-06,
+      "loss": 0.4329,
+      "step": 35965
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.81831957712956,
+      "learning_rate": 2.877975353368544e-06,
+      "loss": 0.4145,
+      "step": 35970
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.528143333174243,
+      "learning_rate": 2.876491387099544e-06,
+      "loss": 0.3957,
+      "step": 35975
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.231655154828122,
+      "learning_rate": 2.875007649010916e-06,
+      "loss": 0.4281,
+      "step": 35980
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 7.461544004566585,
+      "learning_rate": 2.873524139262093e-06,
+      "loss": 0.4101,
+      "step": 35985
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.135560679089954,
+      "learning_rate": 2.872040858012487e-06,
+      "loss": 0.4174,
+      "step": 35990
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 10.469069434737028,
+      "learning_rate": 2.8705578054214813e-06,
+      "loss": 0.41,
+      "step": 35995
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.668290524868912,
+      "learning_rate": 2.8690749816484363e-06,
+      "loss": 0.4635,
+      "step": 36000
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.427870015516103,
+      "learning_rate": 2.8675923868526924e-06,
+      "loss": 0.4084,
+      "step": 36005
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.1801009084144525,
+      "learning_rate": 2.8661100211935544e-06,
+      "loss": 0.4329,
+      "step": 36010
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.8453693226143,
+      "learning_rate": 2.864627884830314e-06,
+      "loss": 0.4363,
+      "step": 36015
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.703891728875428,
+      "learning_rate": 2.8631459779222304e-06,
+      "loss": 0.3968,
+      "step": 36020
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.425389133222261,
+      "learning_rate": 2.861664300628544e-06,
+      "loss": 0.4379,
+      "step": 36025
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.521892542208789,
+      "learning_rate": 2.8601828531084635e-06,
+      "loss": 0.3961,
+      "step": 36030
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.841144524461962,
+      "learning_rate": 2.858701635521181e-06,
+      "loss": 0.4389,
+      "step": 36035
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.51812169464078,
+      "learning_rate": 2.857220648025858e-06,
+      "loss": 0.4123,
+      "step": 36040
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.435356632303837,
+      "learning_rate": 2.8557398907816315e-06,
+      "loss": 0.3939,
+      "step": 36045
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.02000080834046,
+      "learning_rate": 2.854259363947619e-06,
+      "loss": 0.4284,
+      "step": 36050
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.810024106384719,
+      "learning_rate": 2.852779067682906e-06,
+      "loss": 0.4115,
+      "step": 36055
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.726470021066952,
+      "learning_rate": 2.85129900214656e-06,
+      "loss": 0.4044,
+      "step": 36060
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.019711301925271,
+      "learning_rate": 2.8498191674976196e-06,
+      "loss": 0.4631,
+      "step": 36065
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.993576964009328,
+      "learning_rate": 2.8483395638950965e-06,
+      "loss": 0.4265,
+      "step": 36070
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.984751673257927,
+      "learning_rate": 2.846860191497986e-06,
+      "loss": 0.4211,
+      "step": 36075
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.662789550312592,
+      "learning_rate": 2.845381050465248e-06,
+      "loss": 0.429,
+      "step": 36080
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.526399649423495,
+      "learning_rate": 2.8439021409558277e-06,
+      "loss": 0.4188,
+      "step": 36085
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.279901147187739,
+      "learning_rate": 2.8424234631286353e-06,
+      "loss": 0.4272,
+      "step": 36090
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.2157143690530186,
+      "learning_rate": 2.8409450171425644e-06,
+      "loss": 0.4148,
+      "step": 36095
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.022163582817568,
+      "learning_rate": 2.8394668031564813e-06,
+      "loss": 0.4071,
+      "step": 36100
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.2367445406550726,
+      "learning_rate": 2.837988821329224e-06,
+      "loss": 0.3917,
+      "step": 36105
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.42671527299746,
+      "learning_rate": 2.8365110718196114e-06,
+      "loss": 0.4413,
+      "step": 36110
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.099769921367586,
+      "learning_rate": 2.8350335547864306e-06,
+      "loss": 0.4097,
+      "step": 36115
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 7.498264773431116,
+      "learning_rate": 2.8335562703884504e-06,
+      "loss": 0.4185,
+      "step": 36120
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.818740377878313,
+      "learning_rate": 2.832079218784411e-06,
+      "loss": 0.418,
+      "step": 36125
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.90611092781113,
+      "learning_rate": 2.8306024001330247e-06,
+      "loss": 0.4109,
+      "step": 36130
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.657551988019308,
+      "learning_rate": 2.8291258145929866e-06,
+      "loss": 0.4104,
+      "step": 36135
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.682180917766779,
+      "learning_rate": 2.8276494623229596e-06,
+      "loss": 0.4345,
+      "step": 36140
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.47116196035868,
+      "learning_rate": 2.826173343481586e-06,
+      "loss": 0.4266,
+      "step": 36145
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.294404794988088,
+      "learning_rate": 2.824697458227481e-06,
+      "loss": 0.4475,
+      "step": 36150
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 8.68025229717112,
+      "learning_rate": 2.823221806719233e-06,
+      "loss": 0.4277,
+      "step": 36155
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 8.643429093035008,
+      "learning_rate": 2.8217463891154096e-06,
+      "loss": 0.4266,
+      "step": 36160
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.703612087835522,
+      "learning_rate": 2.820271205574549e-06,
+      "loss": 0.4008,
+      "step": 36165
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 7.020606870142804,
+      "learning_rate": 2.8187962562551687e-06,
+      "loss": 0.4037,
+      "step": 36170
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.732231467897349,
+      "learning_rate": 2.817321541315755e-06,
+      "loss": 0.4272,
+      "step": 36175
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.801694588023581,
+      "learning_rate": 2.815847060914777e-06,
+      "loss": 0.4154,
+      "step": 36180
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.7854836844884305,
+      "learning_rate": 2.8143728152106697e-06,
+      "loss": 0.4325,
+      "step": 36185
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.9685100375024405,
+      "learning_rate": 2.8128988043618488e-06,
+      "loss": 0.475,
+      "step": 36190
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.297093421081957,
+      "learning_rate": 2.8114250285267063e-06,
+      "loss": 0.4299,
+      "step": 36195
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.360894529551778,
+      "learning_rate": 2.8099514878636014e-06,
+      "loss": 0.4259,
+      "step": 36200
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.01379139379236,
+      "learning_rate": 2.808478182530876e-06,
+      "loss": 0.4363,
+      "step": 36205
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.703639972534212,
+      "learning_rate": 2.8070051126868424e-06,
+      "loss": 0.4129,
+      "step": 36210
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.466941149260079,
+      "learning_rate": 2.8055322784897854e-06,
+      "loss": 0.4035,
+      "step": 36215
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.458612789550459,
+      "learning_rate": 2.8040596800979723e-06,
+      "loss": 0.4301,
+      "step": 36220
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.199020870941774,
+      "learning_rate": 2.802587317669636e-06,
+      "loss": 0.4038,
+      "step": 36225
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.404110959429534,
+      "learning_rate": 2.801115191362992e-06,
+      "loss": 0.4356,
+      "step": 36230
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.252463521404466,
+      "learning_rate": 2.7996433013362256e-06,
+      "loss": 0.4067,
+      "step": 36235
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.8637602019605035,
+      "learning_rate": 2.798171647747495e-06,
+      "loss": 0.4095,
+      "step": 36240
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.712489662144385,
+      "learning_rate": 2.7967002307549406e-06,
+      "loss": 0.4049,
+      "step": 36245
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 7.059442174944944,
+      "learning_rate": 2.7952290505166677e-06,
+      "loss": 0.4298,
+      "step": 36250
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.788027891102387,
+      "learning_rate": 2.793758107190766e-06,
+      "loss": 0.4238,
+      "step": 36255
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.483292394201127,
+      "learning_rate": 2.7922874009352908e-06,
+      "loss": 0.4098,
+      "step": 36260
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.497815878063458,
+      "learning_rate": 2.79081693190828e-06,
+      "loss": 0.4158,
+      "step": 36265
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.125556353120704,
+      "learning_rate": 2.7893467002677375e-06,
+      "loss": 0.4452,
+      "step": 36270
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.554810986836358,
+      "learning_rate": 2.78787670617165e-06,
+      "loss": 0.3933,
+      "step": 36275
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.364484430678394,
+      "learning_rate": 2.7864069497779716e-06,
+      "loss": 0.4755,
+      "step": 36280
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.097687269323784,
+      "learning_rate": 2.784937431244636e-06,
+      "loss": 0.4409,
+      "step": 36285
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.2833162297273555,
+      "learning_rate": 2.7834681507295507e-06,
+      "loss": 0.3931,
+      "step": 36290
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.2170404685208815,
+      "learning_rate": 2.781999108390595e-06,
+      "loss": 0.4616,
+      "step": 36295
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.328508502860927,
+      "learning_rate": 2.7805303043856225e-06,
+      "loss": 0.4197,
+      "step": 36300
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.403230131051744,
+      "learning_rate": 2.7790617388724657e-06,
+      "loss": 0.4085,
+      "step": 36305
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.404895228455706,
+      "learning_rate": 2.7775934120089245e-06,
+      "loss": 0.4304,
+      "step": 36310
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 6.690811282782428,
+      "learning_rate": 2.776125323952782e-06,
+      "loss": 0.439,
+      "step": 36315
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.470672483781449,
+      "learning_rate": 2.774657474861787e-06,
+      "loss": 0.4623,
+      "step": 36320
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.62030239134579,
+      "learning_rate": 2.773189864893667e-06,
+      "loss": 0.411,
+      "step": 36325
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.4739398209745955,
+      "learning_rate": 2.7717224942061248e-06,
+      "loss": 0.413,
+      "step": 36330
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.823506594901438,
+      "learning_rate": 2.770255362956833e-06,
+      "loss": 0.4111,
+      "step": 36335
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.7321082441764535,
+      "learning_rate": 2.768788471303445e-06,
+      "loss": 0.4027,
+      "step": 36340
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.479022310462112,
+      "learning_rate": 2.7673218194035805e-06,
+      "loss": 0.4245,
+      "step": 36345
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.398404333860972,
+      "learning_rate": 2.7658554074148426e-06,
+      "loss": 0.4026,
+      "step": 36350
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 10.521729262212316,
+      "learning_rate": 2.764389235494801e-06,
+      "loss": 0.4111,
+      "step": 36355
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.5946069384225785,
+      "learning_rate": 2.7629233038010007e-06,
+      "loss": 0.3934,
+      "step": 36360
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.948986975965735,
+      "learning_rate": 2.761457612490964e-06,
+      "loss": 0.3975,
+      "step": 36365
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.78800726522519,
+      "learning_rate": 2.759992161722188e-06,
+      "loss": 0.4083,
+      "step": 36370
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.285161015457855,
+      "learning_rate": 2.758526951652139e-06,
+      "loss": 0.4657,
+      "step": 36375
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.789030658528038,
+      "learning_rate": 2.757061982438263e-06,
+      "loss": 0.4205,
+      "step": 36380
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.666308310236844,
+      "learning_rate": 2.755597254237974e-06,
+      "loss": 0.3834,
+      "step": 36385
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.315779074073722,
+      "learning_rate": 2.754132767208667e-06,
+      "loss": 0.394,
+      "step": 36390
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 6.199855913540895,
+      "learning_rate": 2.7526685215077043e-06,
+      "loss": 0.4019,
+      "step": 36395
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.404346693967891,
+      "learning_rate": 2.7512045172924294e-06,
+      "loss": 0.4264,
+      "step": 36400
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 10.028006412437586,
+      "learning_rate": 2.7497407547201516e-06,
+      "loss": 0.4334,
+      "step": 36405
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.097449562431746,
+      "learning_rate": 2.7482772339481622e-06,
+      "loss": 0.3969,
+      "step": 36410
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.02155729977443,
+      "learning_rate": 2.7468139551337224e-06,
+      "loss": 0.387,
+      "step": 36415
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.7772469708864485,
+      "learning_rate": 2.745350918434065e-06,
+      "loss": 0.4279,
+      "step": 36420
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.209891010891415,
+      "learning_rate": 2.743888124006404e-06,
+      "loss": 0.4104,
+      "step": 36425
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.662079746116483,
+      "learning_rate": 2.7424255720079183e-06,
+      "loss": 0.4243,
+      "step": 36430
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.17076897894065,
+      "learning_rate": 2.7409632625957712e-06,
+      "loss": 0.3811,
+      "step": 36435
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.502074816601167,
+      "learning_rate": 2.739501195927091e-06,
+      "loss": 0.4217,
+      "step": 36440
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.316906021402658,
+      "learning_rate": 2.738039372158982e-06,
+      "loss": 0.3941,
+      "step": 36445
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.248481035019215,
+      "learning_rate": 2.736577791448527e-06,
+      "loss": 0.3969,
+      "step": 36450
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.422605647751632,
+      "learning_rate": 2.7351164539527754e-06,
+      "loss": 0.4063,
+      "step": 36455
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.664164693517471,
+      "learning_rate": 2.7336553598287567e-06,
+      "loss": 0.4254,
+      "step": 36460
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.954230322342542,
+      "learning_rate": 2.7321945092334738e-06,
+      "loss": 0.4406,
+      "step": 36465
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.6998602593014915,
+      "learning_rate": 2.7307339023238977e-06,
+      "loss": 0.4037,
+      "step": 36470
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.326544427509304,
+      "learning_rate": 2.72927353925698e-06,
+      "loss": 0.4257,
+      "step": 36475
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.667641852904444,
+      "learning_rate": 2.7278134201896413e-06,
+      "loss": 0.4291,
+      "step": 36480
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.631164303783946,
+      "learning_rate": 2.726353545278779e-06,
+      "loss": 0.3805,
+      "step": 36485
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.366719163763919,
+      "learning_rate": 2.724893914681262e-06,
+      "loss": 0.4486,
+      "step": 36490
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.199164694763938,
+      "learning_rate": 2.7234345285539358e-06,
+      "loss": 0.4313,
+      "step": 36495
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.664584444248726,
+      "learning_rate": 2.721975387053617e-06,
+      "loss": 0.3929,
+      "step": 36500
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.528675809863526,
+      "learning_rate": 2.720516490337095e-06,
+      "loss": 0.4261,
+      "step": 36505
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.344196832492743,
+      "learning_rate": 2.719057838561138e-06,
+      "loss": 0.4205,
+      "step": 36510
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.44046019834484,
+      "learning_rate": 2.7175994318824806e-06,
+      "loss": 0.4308,
+      "step": 36515
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.811771109033519,
+      "learning_rate": 2.716141270457839e-06,
+      "loss": 0.4422,
+      "step": 36520
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.453069511352081,
+      "learning_rate": 2.7146833544438978e-06,
+      "loss": 0.4254,
+      "step": 36525
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.355192848197981,
+      "learning_rate": 2.7132256839973138e-06,
+      "loss": 0.4278,
+      "step": 36530
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.749061141858274,
+      "learning_rate": 2.711768259274724e-06,
+      "loss": 0.3878,
+      "step": 36535
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.146496471026979,
+      "learning_rate": 2.710311080432732e-06,
+      "loss": 0.406,
+      "step": 36540
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.329809819300412,
+      "learning_rate": 2.708854147627922e-06,
+      "loss": 0.4333,
+      "step": 36545
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.2889672549764075,
+      "learning_rate": 2.7073974610168425e-06,
+      "loss": 0.4211,
+      "step": 36550
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.2556861521690115,
+      "learning_rate": 2.7059410207560243e-06,
+      "loss": 0.4092,
+      "step": 36555
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.9084648181916455,
+      "learning_rate": 2.7044848270019696e-06,
+      "loss": 0.403,
+      "step": 36560
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.94150457712519,
+      "learning_rate": 2.703028879911149e-06,
+      "loss": 0.425,
+      "step": 36565
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.661307617645895,
+      "learning_rate": 2.701573179640015e-06,
+      "loss": 0.4346,
+      "step": 36570
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.768069120396337,
+      "learning_rate": 2.700117726344985e-06,
+      "loss": 0.414,
+      "step": 36575
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.625248390632645,
+      "learning_rate": 2.698662520182457e-06,
+      "loss": 0.4168,
+      "step": 36580
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.503170606133523,
+      "learning_rate": 2.6972075613087973e-06,
+      "loss": 0.4097,
+      "step": 36585
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.130839794364503,
+      "learning_rate": 2.6957528498803464e-06,
+      "loss": 0.3994,
+      "step": 36590
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.453089650499444,
+      "learning_rate": 2.6942983860534233e-06,
+      "loss": 0.4343,
+      "step": 36595
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.799107495942954,
+      "learning_rate": 2.6928441699843127e-06,
+      "loss": 0.4211,
+      "step": 36600
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.536252858238333,
+      "learning_rate": 2.6913902018292805e-06,
+      "loss": 0.4248,
+      "step": 36605
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.646444414108311,
+      "learning_rate": 2.689936481744559e-06,
+      "loss": 0.4368,
+      "step": 36610
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.89258333206394,
+      "learning_rate": 2.688483009886356e-06,
+      "loss": 0.3952,
+      "step": 36615
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.0508522257959765,
+      "learning_rate": 2.6870297864108573e-06,
+      "loss": 0.4267,
+      "step": 36620
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.676194204241176,
+      "learning_rate": 2.6855768114742135e-06,
+      "loss": 0.3846,
+      "step": 36625
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.993402856958031,
+      "learning_rate": 2.684124085232558e-06,
+      "loss": 0.4624,
+      "step": 36630
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.859789286184354,
+      "learning_rate": 2.6826716078419878e-06,
+      "loss": 0.4146,
+      "step": 36635
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 10.109306457052867,
+      "learning_rate": 2.6812193794585827e-06,
+      "loss": 0.4041,
+      "step": 36640
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.29970708060734,
+      "learning_rate": 2.6797674002383877e-06,
+      "loss": 0.4212,
+      "step": 36645
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 7.516928644700587,
+      "learning_rate": 2.6783156703374246e-06,
+      "loss": 0.3952,
+      "step": 36650
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 15.648332275307643,
+      "learning_rate": 2.676864189911691e-06,
+      "loss": 0.417,
+      "step": 36655
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.9122091542143425,
+      "learning_rate": 2.675412959117152e-06,
+      "loss": 0.427,
+      "step": 36660
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.645537064434362,
+      "learning_rate": 2.67396197810975e-06,
+      "loss": 0.4124,
+      "step": 36665
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.503849013747944,
+      "learning_rate": 2.6725112470454006e-06,
+      "loss": 0.4048,
+      "step": 36670
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.350843677064242,
+      "learning_rate": 2.671060766079987e-06,
+      "loss": 0.3947,
+      "step": 36675
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.492872940290535,
+      "learning_rate": 2.6696105353693746e-06,
+      "loss": 0.4289,
+      "step": 36680
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.955805009195729,
+      "learning_rate": 2.6681605550693935e-06,
+      "loss": 0.4201,
+      "step": 36685
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.411660138052263,
+      "learning_rate": 2.6667108253358532e-06,
+      "loss": 0.3996,
+      "step": 36690
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.780222465525253,
+      "learning_rate": 2.6652613463245325e-06,
+      "loss": 0.4146,
+      "step": 36695
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 7.6609998734171,
+      "learning_rate": 2.6638121181911825e-06,
+      "loss": 0.3924,
+      "step": 36700
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.0917966203365355,
+      "learning_rate": 2.662363141091532e-06,
+      "loss": 0.4242,
+      "step": 36705
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.000887366140905,
+      "learning_rate": 2.660914415181277e-06,
+      "loss": 0.455,
+      "step": 36710
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.598953488178567,
+      "learning_rate": 2.659465940616094e-06,
+      "loss": 0.4115,
+      "step": 36715
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.2242195689340205,
+      "learning_rate": 2.6580177175516243e-06,
+      "loss": 0.3939,
+      "step": 36720
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.859897513620613,
+      "learning_rate": 2.656569746143485e-06,
+      "loss": 0.3936,
+      "step": 36725
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.996479697641248,
+      "learning_rate": 2.6551220265472695e-06,
+      "loss": 0.4305,
+      "step": 36730
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.850121515215442,
+      "learning_rate": 2.6536745589185424e-06,
+      "loss": 0.3987,
+      "step": 36735
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.570310442510414,
+      "learning_rate": 2.6522273434128375e-06,
+      "loss": 0.4228,
+      "step": 36740
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.537257533717401,
+      "learning_rate": 2.650780380185666e-06,
+      "loss": 0.3883,
+      "step": 36745
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.713636000212951,
+      "learning_rate": 2.649333669392512e-06,
+      "loss": 0.4405,
+      "step": 36750
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.168481118979662,
+      "learning_rate": 2.64788721118883e-06,
+      "loss": 0.41,
+      "step": 36755
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.695367622300172,
+      "learning_rate": 2.6464410057300456e-06,
+      "loss": 0.4189,
+      "step": 36760
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.115583168043013,
+      "learning_rate": 2.6449950531715642e-06,
+      "loss": 0.3903,
+      "step": 36765
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.817883421548684,
+      "learning_rate": 2.6435493536687558e-06,
+      "loss": 0.4011,
+      "step": 36770
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.822750552321125,
+      "learning_rate": 2.642103907376971e-06,
+      "loss": 0.3957,
+      "step": 36775
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 7.088921083813334,
+      "learning_rate": 2.6406587144515278e-06,
+      "loss": 0.427,
+      "step": 36780
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.408250344342435,
+      "learning_rate": 2.639213775047716e-06,
+      "loss": 0.3833,
+      "step": 36785
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.635394882245829,
+      "learning_rate": 2.6377690893208052e-06,
+      "loss": 0.4183,
+      "step": 36790
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.769981995173298,
+      "learning_rate": 2.6363246574260292e-06,
+      "loss": 0.4261,
+      "step": 36795
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.9216904514130055,
+      "learning_rate": 2.6348804795186023e-06,
+      "loss": 0.4237,
+      "step": 36800
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.286144774587911,
+      "learning_rate": 2.633436555753706e-06,
+      "loss": 0.4218,
+      "step": 36805
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.3700808677456395,
+      "learning_rate": 2.6319928862864943e-06,
+      "loss": 0.4194,
+      "step": 36810
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.604255316573414,
+      "learning_rate": 2.6305494712720992e-06,
+      "loss": 0.4314,
+      "step": 36815
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.596443876762925,
+      "learning_rate": 2.6291063108656196e-06,
+      "loss": 0.4069,
+      "step": 36820
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.650878870583402,
+      "learning_rate": 2.6276634052221293e-06,
+      "loss": 0.4245,
+      "step": 36825
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.732009022040906,
+      "learning_rate": 2.626220754496679e-06,
+      "loss": 0.3897,
+      "step": 36830
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.94193684168628,
+      "learning_rate": 2.6247783588442825e-06,
+      "loss": 0.4425,
+      "step": 36835
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.774766278576813,
+      "learning_rate": 2.6233362184199363e-06,
+      "loss": 0.4183,
+      "step": 36840
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.584948319885389,
+      "learning_rate": 2.6218943333786e-06,
+      "loss": 0.4111,
+      "step": 36845
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.931391197413627,
+      "learning_rate": 2.6204527038752152e-06,
+      "loss": 0.3966,
+      "step": 36850
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.240101305107887,
+      "learning_rate": 2.6190113300646867e-06,
+      "loss": 0.46,
+      "step": 36855
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.804345690368732,
+      "learning_rate": 2.617570212101902e-06,
+      "loss": 0.4524,
+      "step": 36860
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.5173813188418865,
+      "learning_rate": 2.6161293501417118e-06,
+      "loss": 0.4205,
+      "step": 36865
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.117152358658561,
+      "learning_rate": 2.6146887443389415e-06,
+      "loss": 0.4095,
+      "step": 36870
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.774960543724538,
+      "learning_rate": 2.613248394848395e-06,
+      "loss": 0.3893,
+      "step": 36875
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.827645382677662,
+      "learning_rate": 2.6118083018248395e-06,
+      "loss": 0.4253,
+      "step": 36880
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.2976688080844445,
+      "learning_rate": 2.610368465423025e-06,
+      "loss": 0.372,
+      "step": 36885
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.370441791084755,
+      "learning_rate": 2.6089288857976626e-06,
+      "loss": 0.446,
+      "step": 36890
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.0315887646897375,
+      "learning_rate": 2.607489563103446e-06,
+      "loss": 0.3992,
+      "step": 36895
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.172398603506704,
+      "learning_rate": 2.606050497495034e-06,
+      "loss": 0.4093,
+      "step": 36900
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.777776860692738,
+      "learning_rate": 2.604611689127061e-06,
+      "loss": 0.4152,
+      "step": 36905
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.475064907492985,
+      "learning_rate": 2.6031731381541343e-06,
+      "loss": 0.396,
+      "step": 36910
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.097735557080617,
+      "learning_rate": 2.60173484473083e-06,
+      "loss": 0.4196,
+      "step": 36915
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.27801905171361,
+      "learning_rate": 2.600296809011702e-06,
+      "loss": 0.4166,
+      "step": 36920
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.261139695024201,
+      "learning_rate": 2.5988590311512737e-06,
+      "loss": 0.4076,
+      "step": 36925
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.1881015452394035,
+      "learning_rate": 2.597421511304038e-06,
+      "loss": 0.4464,
+      "step": 36930
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.988799242420278,
+      "learning_rate": 2.595984249624466e-06,
+      "loss": 0.3836,
+      "step": 36935
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.5124362999881855,
+      "learning_rate": 2.594547246266994e-06,
+      "loss": 0.3913,
+      "step": 36940
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.772434281021417,
+      "learning_rate": 2.5931105013860387e-06,
+      "loss": 0.4292,
+      "step": 36945
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.003772279738226,
+      "learning_rate": 2.5916740151359822e-06,
+      "loss": 0.3996,
+      "step": 36950
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.480029381711743,
+      "learning_rate": 2.5902377876711793e-06,
+      "loss": 0.4008,
+      "step": 36955
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.289919052430077,
+      "learning_rate": 2.588801819145964e-06,
+      "loss": 0.4155,
+      "step": 36960
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.714719928751562,
+      "learning_rate": 2.5873661097146317e-06,
+      "loss": 0.4115,
+      "step": 36965
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.071001850463855,
+      "learning_rate": 2.5859306595314615e-06,
+      "loss": 0.4096,
+      "step": 36970
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.155942230349109,
+      "learning_rate": 2.584495468750694e-06,
+      "loss": 0.4296,
+      "step": 36975
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.429951845304517,
+      "learning_rate": 2.5830605375265506e-06,
+      "loss": 0.4276,
+      "step": 36980
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.646532014474953,
+      "learning_rate": 2.581625866013219e-06,
+      "loss": 0.4019,
+      "step": 36985
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.650484496248696,
+      "learning_rate": 2.58019145436486e-06,
+      "loss": 0.3936,
+      "step": 36990
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.799052121226185,
+      "learning_rate": 2.5787573027356093e-06,
+      "loss": 0.4007,
+      "step": 36995
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.79576543933088,
+      "learning_rate": 2.577323411279571e-06,
+      "loss": 0.4142,
+      "step": 37000
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.246577989215455,
+      "learning_rate": 2.5758897801508265e-06,
+      "loss": 0.429,
+      "step": 37005
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.696300069976304,
+      "learning_rate": 2.5744564095034207e-06,
+      "loss": 0.4128,
+      "step": 37010
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.704977450684182,
+      "learning_rate": 2.5730232994913796e-06,
+      "loss": 0.4211,
+      "step": 37015
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.250893250387021,
+      "learning_rate": 2.5715904502686974e-06,
+      "loss": 0.4333,
+      "step": 37020
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 7.173325980955074,
+      "learning_rate": 2.5701578619893365e-06,
+      "loss": 0.4425,
+      "step": 37025
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.08917208537564,
+      "learning_rate": 2.5687255348072395e-06,
+      "loss": 0.402,
+      "step": 37030
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.392746900117666,
+      "learning_rate": 2.5672934688763127e-06,
+      "loss": 0.4284,
+      "step": 37035
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.975233371641439,
+      "learning_rate": 2.5658616643504384e-06,
+      "loss": 0.3635,
+      "step": 37040
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.500467969019714,
+      "learning_rate": 2.564430121383472e-06,
+      "loss": 0.4297,
+      "step": 37045
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.709001829842048,
+      "learning_rate": 2.562998840129237e-06,
+      "loss": 0.3936,
+      "step": 37050
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.8911817075677675,
+      "learning_rate": 2.5615678207415328e-06,
+      "loss": 0.4346,
+      "step": 37055
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.844222912997217,
+      "learning_rate": 2.560137063374127e-06,
+      "loss": 0.3901,
+      "step": 37060
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.50211780390919,
+      "learning_rate": 2.558706568180763e-06,
+      "loss": 0.3716,
+      "step": 37065
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.859549667242615,
+      "learning_rate": 2.557276335315153e-06,
+      "loss": 0.4038,
+      "step": 37070
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.992798825060795,
+      "learning_rate": 2.5558463649309784e-06,
+      "loss": 0.3991,
+      "step": 37075
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.036823403276468,
+      "learning_rate": 2.554416657181902e-06,
+      "loss": 0.4323,
+      "step": 37080
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.687095730396531,
+      "learning_rate": 2.552987212221546e-06,
+      "loss": 0.423,
+      "step": 37085
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.738133789390756,
+      "learning_rate": 2.551558030203516e-06,
+      "loss": 0.4328,
+      "step": 37090
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.246046285127597,
+      "learning_rate": 2.550129111281379e-06,
+      "loss": 0.4204,
+      "step": 37095
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.6214212706319175,
+      "learning_rate": 2.5487004556086837e-06,
+      "loss": 0.4274,
+      "step": 37100
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.393163629155765,
+      "learning_rate": 2.5472720633389404e-06,
+      "loss": 0.3987,
+      "step": 37105
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.334542286236737,
+      "learning_rate": 2.545843934625639e-06,
+      "loss": 0.4213,
+      "step": 37110
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.812260206538378,
+      "learning_rate": 2.5444160696222396e-06,
+      "loss": 0.4059,
+      "step": 37115
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.436021657861293,
+      "learning_rate": 2.5429884684821703e-06,
+      "loss": 0.398,
+      "step": 37120
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 8.027533356901605,
+      "learning_rate": 2.541561131358833e-06,
+      "loss": 0.4311,
+      "step": 37125
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.786042334603683,
+      "learning_rate": 2.5401340584056027e-06,
+      "loss": 0.4231,
+      "step": 37130
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.065005851699361,
+      "learning_rate": 2.5387072497758223e-06,
+      "loss": 0.428,
+      "step": 37135
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.9724143256521005,
+      "learning_rate": 2.537280705622813e-06,
+      "loss": 0.3846,
+      "step": 37140
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.114992931360732,
+      "learning_rate": 2.5358544260998586e-06,
+      "loss": 0.4211,
+      "step": 37145
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.322275942712595,
+      "learning_rate": 2.534428411360223e-06,
+      "loss": 0.4235,
+      "step": 37150
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.50726036133838,
+      "learning_rate": 2.533002661557136e-06,
+      "loss": 0.3908,
+      "step": 37155
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.274073228679576,
+      "learning_rate": 2.5315771768437992e-06,
+      "loss": 0.3499,
+      "step": 37160
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.31233599695848,
+      "learning_rate": 2.53015195737339e-06,
+      "loss": 0.4161,
+      "step": 37165
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 12.790035088998128,
+      "learning_rate": 2.528727003299051e-06,
+      "loss": 0.4346,
+      "step": 37170
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.223004381797734,
+      "learning_rate": 2.5273023147739053e-06,
+      "loss": 0.4195,
+      "step": 37175
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 8.547919830178392,
+      "learning_rate": 2.5258778919510374e-06,
+      "loss": 0.4159,
+      "step": 37180
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 11.777535512748289,
+      "learning_rate": 2.524453734983508e-06,
+      "loss": 0.3997,
+      "step": 37185
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.944035923509637,
+      "learning_rate": 2.52302984402435e-06,
+      "loss": 0.4036,
+      "step": 37190
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.355378288620898,
+      "learning_rate": 2.521606219226569e-06,
+      "loss": 0.4121,
+      "step": 37195
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.7722295476357255,
+      "learning_rate": 2.5201828607431356e-06,
+      "loss": 0.4185,
+      "step": 37200
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.97315776632713,
+      "learning_rate": 2.5187597687270004e-06,
+      "loss": 0.4339,
+      "step": 37205
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.217279868258032,
+      "learning_rate": 2.517336943331076e-06,
+      "loss": 0.3967,
+      "step": 37210
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.699923750449226,
+      "learning_rate": 2.5159143847082563e-06,
+      "loss": 0.4322,
+      "step": 37215
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 8.261317614190745,
+      "learning_rate": 2.514492093011397e-06,
+      "loss": 0.4378,
+      "step": 37220
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 8.93734667847287,
+      "learning_rate": 2.5130700683933333e-06,
+      "loss": 0.4149,
+      "step": 37225
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.651863631151353,
+      "learning_rate": 2.511648311006865e-06,
+      "loss": 0.4272,
+      "step": 37230
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.689396472362099,
+      "learning_rate": 2.510226821004769e-06,
+      "loss": 0.4158,
+      "step": 37235
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.885267570508475,
+      "learning_rate": 2.5088055985397897e-06,
+      "loss": 0.4226,
+      "step": 37240
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.22027601351842,
+      "learning_rate": 2.5073846437646414e-06,
+      "loss": 0.4306,
+      "step": 37245
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.816526156854411,
+      "learning_rate": 2.505963956832016e-06,
+      "loss": 0.3883,
+      "step": 37250
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.627145852240641,
+      "learning_rate": 2.504543537894568e-06,
+      "loss": 0.4228,
+      "step": 37255
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.039465025094866,
+      "learning_rate": 2.503123387104932e-06,
+      "loss": 0.4077,
+      "step": 37260
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.713766621382293,
+      "learning_rate": 2.5017035046157084e-06,
+      "loss": 0.422,
+      "step": 37265
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.186512509520176,
+      "learning_rate": 2.500283890579467e-06,
+      "loss": 0.4163,
+      "step": 37270
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.724030994072304,
+      "learning_rate": 2.498864545148753e-06,
+      "loss": 0.4196,
+      "step": 37275
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.8461477265876605,
+      "learning_rate": 2.497445468476084e-06,
+      "loss": 0.4408,
+      "step": 37280
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.586900027130816,
+      "learning_rate": 2.4960266607139425e-06,
+      "loss": 0.4095,
+      "step": 37285
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.413684011014718,
+      "learning_rate": 2.4946081220147873e-06,
+      "loss": 0.4099,
+      "step": 37290
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.252677108533727,
+      "learning_rate": 2.493189852531048e-06,
+      "loss": 0.4042,
+      "step": 37295
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.781175451176674,
+      "learning_rate": 2.4917718524151223e-06,
+      "loss": 0.38,
+      "step": 37300
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.735017657140206,
+      "learning_rate": 2.4903541218193793e-06,
+      "loss": 0.416,
+      "step": 37305
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.225304668419313,
+      "learning_rate": 2.4889366608961634e-06,
+      "loss": 0.3999,
+      "step": 37310
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.508559695575597,
+      "learning_rate": 2.487519469797784e-06,
+      "loss": 0.4062,
+      "step": 37315
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.9961693946487875,
+      "learning_rate": 2.486102548676527e-06,
+      "loss": 0.4096,
+      "step": 37320
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.084391144246881,
+      "learning_rate": 2.4846858976846467e-06,
+      "loss": 0.4193,
+      "step": 37325
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.225517806011373,
+      "learning_rate": 2.4832695169743653e-06,
+      "loss": 0.3871,
+      "step": 37330
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.948625096236599,
+      "learning_rate": 2.4818534066978832e-06,
+      "loss": 0.4008,
+      "step": 37335
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.298715459546129,
+      "learning_rate": 2.480437567007365e-06,
+      "loss": 0.4288,
+      "step": 37340
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.883682202729534,
+      "learning_rate": 2.4790219980549507e-06,
+      "loss": 0.4412,
+      "step": 37345
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.0333951597344,
+      "learning_rate": 2.4776066999927494e-06,
+      "loss": 0.4204,
+      "step": 37350
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.564161171379566,
+      "learning_rate": 2.476191672972838e-06,
+      "loss": 0.4022,
+      "step": 37355
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.766537371421515,
+      "learning_rate": 2.474776917147272e-06,
+      "loss": 0.3914,
+      "step": 37360
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.695060175376409,
+      "learning_rate": 2.473362432668069e-06,
+      "loss": 0.4184,
+      "step": 37365
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.598763837265946,
+      "learning_rate": 2.471948219687223e-06,
+      "loss": 0.3979,
+      "step": 37370
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.112981818948408,
+      "learning_rate": 2.4705342783567003e-06,
+      "loss": 0.4181,
+      "step": 37375
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.515737409026771,
+      "learning_rate": 2.469120608828431e-06,
+      "loss": 0.404,
+      "step": 37380
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.525117709831719,
+      "learning_rate": 2.467707211254323e-06,
+      "loss": 0.4391,
+      "step": 37385
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.854523664756024,
+      "learning_rate": 2.46629408578625e-06,
+      "loss": 0.3884,
+      "step": 37390
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.590883321139548,
+      "learning_rate": 2.4648812325760608e-06,
+      "loss": 0.4253,
+      "step": 37395
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.241124085958273,
+      "learning_rate": 2.46346865177557e-06,
+      "loss": 0.4026,
+      "step": 37400
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.855272957871328,
+      "learning_rate": 2.462056343536569e-06,
+      "loss": 0.4068,
+      "step": 37405
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.730198366331148,
+      "learning_rate": 2.460644308010815e-06,
+      "loss": 0.3955,
+      "step": 37410
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.361828360653933,
+      "learning_rate": 2.459232545350035e-06,
+      "loss": 0.3978,
+      "step": 37415
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.1829538351207916,
+      "learning_rate": 2.4578210557059328e-06,
+      "loss": 0.4004,
+      "step": 37420
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.580731984937707,
+      "learning_rate": 2.4564098392301762e-06,
+      "loss": 0.3872,
+      "step": 37425
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.985120921495895,
+      "learning_rate": 2.4549988960744093e-06,
+      "loss": 0.4208,
+      "step": 37430
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.6656444645193265,
+      "learning_rate": 2.4535882263902437e-06,
+      "loss": 0.3997,
+      "step": 37435
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.185878999584126,
+      "learning_rate": 2.4521778303292593e-06,
+      "loss": 0.4112,
+      "step": 37440
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.330915851906834,
+      "learning_rate": 2.4507677080430133e-06,
+      "loss": 0.4037,
+      "step": 37445
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.812815108654893,
+      "learning_rate": 2.4493578596830253e-06,
+      "loss": 0.4251,
+      "step": 37450
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.400125989976663,
+      "learning_rate": 2.447948285400795e-06,
+      "loss": 0.4144,
+      "step": 37455
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.540735588095793,
+      "learning_rate": 2.4465389853477816e-06,
+      "loss": 0.4116,
+      "step": 37460
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.474065047876905,
+      "learning_rate": 2.445129959675424e-06,
+      "loss": 0.3876,
+      "step": 37465
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.283110997189188,
+      "learning_rate": 2.44372120853513e-06,
+      "loss": 0.4236,
+      "step": 37470
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.935817852491185,
+      "learning_rate": 2.4423127320782713e-06,
+      "loss": 0.4063,
+      "step": 37475
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.497913000376273,
+      "learning_rate": 2.4409045304561996e-06,
+      "loss": 0.4135,
+      "step": 37480
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.066740557736832,
+      "learning_rate": 2.4394966038202283e-06,
+      "loss": 0.4357,
+      "step": 37485
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.972230265299325,
+      "learning_rate": 2.4380889523216495e-06,
+      "loss": 0.4158,
+      "step": 37490
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.318528186475073,
+      "learning_rate": 2.436681576111719e-06,
+      "loss": 0.4304,
+      "step": 37495
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.279422060732313,
+      "learning_rate": 2.435274475341664e-06,
+      "loss": 0.4158,
+      "step": 37500
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.031563289288517,
+      "learning_rate": 2.4338676501626878e-06,
+      "loss": 0.4319,
+      "step": 37505
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.361142104309413,
+      "learning_rate": 2.432461100725956e-06,
+      "loss": 0.4223,
+      "step": 37510
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.859435866442603,
+      "learning_rate": 2.431054827182612e-06,
+      "loss": 0.3935,
+      "step": 37515
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.799037485664568,
+      "learning_rate": 2.429648829683764e-06,
+      "loss": 0.4007,
+      "step": 37520
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.379476203172504,
+      "learning_rate": 2.428243108380492e-06,
+      "loss": 0.4124,
+      "step": 37525
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.865293594312809,
+      "learning_rate": 2.4268376634238495e-06,
+      "loss": 0.4071,
+      "step": 37530
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.5979140229018824,
+      "learning_rate": 2.425432494964854e-06,
+      "loss": 0.3967,
+      "step": 37535
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.913608540213429,
+      "learning_rate": 2.4240276031545016e-06,
+      "loss": 0.4019,
+      "step": 37540
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.825015519043568,
+      "learning_rate": 2.4226229881437502e-06,
+      "loss": 0.4032,
+      "step": 37545
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.380107640660564,
+      "learning_rate": 2.4212186500835356e-06,
+      "loss": 0.4239,
+      "step": 37550
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.574154260327579,
+      "learning_rate": 2.419814589124756e-06,
+      "loss": 0.4381,
+      "step": 37555
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.329868686667875,
+      "learning_rate": 2.4184108054182858e-06,
+      "loss": 0.4444,
+      "step": 37560
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.982523434416819,
+      "learning_rate": 2.41700729911497e-06,
+      "loss": 0.4402,
+      "step": 37565
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.712082409711277,
+      "learning_rate": 2.4156040703656175e-06,
+      "loss": 0.3955,
+      "step": 37570
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.05896953060982,
+      "learning_rate": 2.4142011193210158e-06,
+      "loss": 0.4109,
+      "step": 37575
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.295058768864407,
+      "learning_rate": 2.4127984461319153e-06,
+      "loss": 0.4357,
+      "step": 37580
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.528598664034226,
+      "learning_rate": 2.411396050949038e-06,
+      "loss": 0.4261,
+      "step": 37585
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.51267752289972,
+      "learning_rate": 2.4099939339230814e-06,
+      "loss": 0.4417,
+      "step": 37590
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.231566072249203,
+      "learning_rate": 2.4085920952047058e-06,
+      "loss": 0.4009,
+      "step": 37595
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.612495242298953,
+      "learning_rate": 2.407190534944548e-06,
+      "loss": 0.4204,
+      "step": 37600
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.9382712370806785,
+      "learning_rate": 2.40578925329321e-06,
+      "loss": 0.3692,
+      "step": 37605
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.305763602048339,
+      "learning_rate": 2.404388250401264e-06,
+      "loss": 0.4083,
+      "step": 37610
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.619393851912315,
+      "learning_rate": 2.4029875264192576e-06,
+      "loss": 0.4414,
+      "step": 37615
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.159974390985439,
+      "learning_rate": 2.401587081497701e-06,
+      "loss": 0.3798,
+      "step": 37620
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.299737112191288,
+      "learning_rate": 2.400186915787082e-06,
+      "loss": 0.3885,
+      "step": 37625
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 10.129151557162364,
+      "learning_rate": 2.3987870294378513e-06,
+      "loss": 0.4163,
+      "step": 37630
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 12.081109825709918,
+      "learning_rate": 2.3973874226004363e-06,
+      "loss": 0.3947,
+      "step": 37635
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.669906888924857,
+      "learning_rate": 2.3959880954252263e-06,
+      "loss": 0.4058,
+      "step": 37640
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.8184225449772,
+      "learning_rate": 2.3945890480625906e-06,
+      "loss": 0.3989,
+      "step": 37645
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 10.162092643494919,
+      "learning_rate": 2.3931902806628576e-06,
+      "loss": 0.4113,
+      "step": 37650
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.452591396243657,
+      "learning_rate": 2.391791793376334e-06,
+      "loss": 0.4079,
+      "step": 37655
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.010348624722068,
+      "learning_rate": 2.390393586353295e-06,
+      "loss": 0.411,
+      "step": 37660
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.073279424517539,
+      "learning_rate": 2.3889956597439824e-06,
+      "loss": 0.4246,
+      "step": 37665
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.299062686226677,
+      "learning_rate": 2.387598013698607e-06,
+      "loss": 0.4225,
+      "step": 37670
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.254463288894242,
+      "learning_rate": 2.386200648367357e-06,
+      "loss": 0.4212,
+      "step": 37675
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.627641042334762,
+      "learning_rate": 2.3848035639003803e-06,
+      "loss": 0.4103,
+      "step": 37680
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.701986690670523,
+      "learning_rate": 2.3834067604478046e-06,
+      "loss": 0.4254,
+      "step": 37685
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.419882751567113,
+      "learning_rate": 2.3820102381597202e-06,
+      "loss": 0.4507,
+      "step": 37690
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.904287101507349,
+      "learning_rate": 2.3806139971861884e-06,
+      "loss": 0.4177,
+      "step": 37695
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.97503206025527,
+      "learning_rate": 2.3792180376772445e-06,
+      "loss": 0.4325,
+      "step": 37700
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.386308205817909,
+      "learning_rate": 2.377822359782887e-06,
+      "loss": 0.4217,
+      "step": 37705
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.583689489847988,
+      "learning_rate": 2.376426963653091e-06,
+      "loss": 0.4134,
+      "step": 37710
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.735002306111745,
+      "learning_rate": 2.375031849437795e-06,
+      "loss": 0.3793,
+      "step": 37715
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.140189357114928,
+      "learning_rate": 2.3736370172869133e-06,
+      "loss": 0.4252,
+      "step": 37720
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.689969677875854,
+      "learning_rate": 2.372242467350325e-06,
+      "loss": 0.4206,
+      "step": 37725
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.086416613113523,
+      "learning_rate": 2.3708481997778798e-06,
+      "loss": 0.4457,
+      "step": 37730
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.990315329234146,
+      "learning_rate": 2.369454214719398e-06,
+      "loss": 0.4203,
+      "step": 37735
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.4633217080904215,
+      "learning_rate": 2.3680605123246725e-06,
+      "loss": 0.389,
+      "step": 37740
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.746868226179556,
+      "learning_rate": 2.3666670927434584e-06,
+      "loss": 0.3885,
+      "step": 37745
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.7063579459136555,
+      "learning_rate": 2.3652739561254888e-06,
+      "loss": 0.3877,
+      "step": 37750
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.896519934834272,
+      "learning_rate": 2.363881102620459e-06,
+      "loss": 0.3972,
+      "step": 37755
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.913498064137382,
+      "learning_rate": 2.3624885323780407e-06,
+      "loss": 0.4095,
+      "step": 37760
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.108919903583556,
+      "learning_rate": 2.3610962455478677e-06,
+      "loss": 0.4111,
+      "step": 37765
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.921840185704436,
+      "learning_rate": 2.3597042422795513e-06,
+      "loss": 0.4306,
+      "step": 37770
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.852995035964516,
+      "learning_rate": 2.358312522722665e-06,
+      "loss": 0.4043,
+      "step": 37775
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.529711926116011,
+      "learning_rate": 2.3569210870267584e-06,
+      "loss": 0.4305,
+      "step": 37780
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.657574902923141,
+      "learning_rate": 2.3555299353413462e-06,
+      "loss": 0.3874,
+      "step": 37785
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.905875732974656,
+      "learning_rate": 2.354139067815911e-06,
+      "loss": 0.3995,
+      "step": 37790
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 6.364117491954523,
+      "learning_rate": 2.352748484599912e-06,
+      "loss": 0.4365,
+      "step": 37795
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 6.89228524091132,
+      "learning_rate": 2.35135818584277e-06,
+      "loss": 0.3904,
+      "step": 37800
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.604604486773298,
+      "learning_rate": 2.349968171693882e-06,
+      "loss": 0.3895,
+      "step": 37805
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.547158382550993,
+      "learning_rate": 2.34857844230261e-06,
+      "loss": 0.4219,
+      "step": 37810
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.232739738429124,
+      "learning_rate": 2.3471889978182833e-06,
+      "loss": 0.4453,
+      "step": 37815
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.120494392345778,
+      "learning_rate": 2.3457998383902087e-06,
+      "loss": 0.4093,
+      "step": 37820
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.932578816784359,
+      "learning_rate": 2.344410964167653e-06,
+      "loss": 0.4396,
+      "step": 37825
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.588241513071872,
+      "learning_rate": 2.34302237529986e-06,
+      "loss": 0.4199,
+      "step": 37830
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.584702044689125,
+      "learning_rate": 2.34163407193604e-06,
+      "loss": 0.434,
+      "step": 37835
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.187374255593626,
+      "learning_rate": 2.3402460542253697e-06,
+      "loss": 0.4314,
+      "step": 37840
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.8905638875035145,
+      "learning_rate": 2.338858322317001e-06,
+      "loss": 0.3942,
+      "step": 37845
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.839459242498384,
+      "learning_rate": 2.3374708763600486e-06,
+      "loss": 0.3866,
+      "step": 37850
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.604624014402693,
+      "learning_rate": 2.336083716503603e-06,
+      "loss": 0.4483,
+      "step": 37855
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.494063249026857,
+      "learning_rate": 2.3346968428967166e-06,
+      "loss": 0.4276,
+      "step": 37860
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.29113972614164,
+      "learning_rate": 2.3333102556884195e-06,
+      "loss": 0.4001,
+      "step": 37865
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.6989887823760075,
+      "learning_rate": 2.3319239550277046e-06,
+      "loss": 0.4072,
+      "step": 37870
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 7.588016207966241,
+      "learning_rate": 2.3305379410635332e-06,
+      "loss": 0.3998,
+      "step": 37875
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.386087370158286,
+      "learning_rate": 2.3291522139448435e-06,
+      "loss": 0.4152,
+      "step": 37880
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.108323281361963,
+      "learning_rate": 2.3277667738205344e-06,
+      "loss": 0.3894,
+      "step": 37885
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 7.355453885605402,
+      "learning_rate": 2.32638162083948e-06,
+      "loss": 0.4276,
+      "step": 37890
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.577862098778375,
+      "learning_rate": 2.324996755150521e-06,
+      "loss": 0.3728,
+      "step": 37895
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.876721549951402,
+      "learning_rate": 2.323612176902464e-06,
+      "loss": 0.4082,
+      "step": 37900
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.893429653420833,
+      "learning_rate": 2.3222278862440917e-06,
+      "loss": 0.3795,
+      "step": 37905
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.113507778468923,
+      "learning_rate": 2.3208438833241497e-06,
+      "loss": 0.4241,
+      "step": 37910
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 79.98111007846417,
+      "learning_rate": 2.319460168291358e-06,
+      "loss": 0.4638,
+      "step": 37915
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.9487472433427975,
+      "learning_rate": 2.3180767412943995e-06,
+      "loss": 0.4241,
+      "step": 37920
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.226125973952726,
+      "learning_rate": 2.3166936024819313e-06,
+      "loss": 0.422,
+      "step": 37925
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.940809355740361,
+      "learning_rate": 2.31531075200258e-06,
+      "loss": 0.4458,
+      "step": 37930
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.0268930430563445,
+      "learning_rate": 2.313928190004935e-06,
+      "loss": 0.3701,
+      "step": 37935
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.58558284411666,
+      "learning_rate": 2.3125459166375625e-06,
+      "loss": 0.3853,
+      "step": 37940
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.522188671731066,
+      "learning_rate": 2.31116393204899e-06,
+      "loss": 0.3954,
+      "step": 37945
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.406528873180831,
+      "learning_rate": 2.309782236387721e-06,
+      "loss": 0.4125,
+      "step": 37950
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.178343992922706,
+      "learning_rate": 2.308400829802224e-06,
+      "loss": 0.396,
+      "step": 37955
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.665358440445545,
+      "learning_rate": 2.307019712440935e-06,
+      "loss": 0.4077,
+      "step": 37960
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.9265159269255525,
+      "learning_rate": 2.305638884452265e-06,
+      "loss": 0.4206,
+      "step": 37965
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.94031396378375,
+      "learning_rate": 2.3042583459845853e-06,
+      "loss": 0.3825,
+      "step": 37970
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.476643028450257,
+      "learning_rate": 2.3028780971862457e-06,
+      "loss": 0.3737,
+      "step": 37975
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 12.275128561514984,
+      "learning_rate": 2.3014981382055585e-06,
+      "loss": 0.4332,
+      "step": 37980
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.856065342077004,
+      "learning_rate": 2.300118469190803e-06,
+      "loss": 0.4023,
+      "step": 37985
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.753232569364006,
+      "learning_rate": 2.2987390902902357e-06,
+      "loss": 0.4239,
+      "step": 37990
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.524916146054429,
+      "learning_rate": 2.297360001652073e-06,
+      "loss": 0.4018,
+      "step": 37995
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.344325463129725,
+      "learning_rate": 2.2959812034245077e-06,
+      "loss": 0.4064,
+      "step": 38000
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.907803185467804,
+      "learning_rate": 2.2946026957556937e-06,
+      "loss": 0.4312,
+      "step": 38005
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.594221586335419,
+      "learning_rate": 2.2932244787937618e-06,
+      "loss": 0.4144,
+      "step": 38010
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.99973523736737,
+      "learning_rate": 2.291846552686804e-06,
+      "loss": 0.3782,
+      "step": 38015
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.396141243222192,
+      "learning_rate": 2.290468917582886e-06,
+      "loss": 0.4327,
+      "step": 38020
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.567389639433534,
+      "learning_rate": 2.2890915736300422e-06,
+      "loss": 0.4231,
+      "step": 38025
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.541973240712317,
+      "learning_rate": 2.2877145209762714e-06,
+      "loss": 0.4373,
+      "step": 38030
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.229987450128883,
+      "learning_rate": 2.286337759769548e-06,
+      "loss": 0.3926,
+      "step": 38035
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.870012224100013,
+      "learning_rate": 2.284961290157808e-06,
+      "loss": 0.3947,
+      "step": 38040
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.342811809778449,
+      "learning_rate": 2.283585112288958e-06,
+      "loss": 0.3772,
+      "step": 38045
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.333846513686606,
+      "learning_rate": 2.282209226310878e-06,
+      "loss": 0.4065,
+      "step": 38050
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.0244839834440835,
+      "learning_rate": 2.2808336323714094e-06,
+      "loss": 0.4091,
+      "step": 38055
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.890421913453521,
+      "learning_rate": 2.2794583306183694e-06,
+      "loss": 0.4033,
+      "step": 38060
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.805186999635017,
+      "learning_rate": 2.2780833211995383e-06,
+      "loss": 0.4142,
+      "step": 38065
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.541539644382564,
+      "learning_rate": 2.276708604262665e-06,
+      "loss": 0.3957,
+      "step": 38070
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.7224896999990165,
+      "learning_rate": 2.2753341799554735e-06,
+      "loss": 0.4235,
+      "step": 38075
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.557678874844924,
+      "learning_rate": 2.273960048425647e-06,
+      "loss": 0.3909,
+      "step": 38080
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.86070320469673,
+      "learning_rate": 2.2725862098208464e-06,
+      "loss": 0.3897,
+      "step": 38085
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.325590840359363,
+      "learning_rate": 2.271212664288694e-06,
+      "loss": 0.4168,
+      "step": 38090
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.7631955298398,
+      "learning_rate": 2.2698394119767825e-06,
+      "loss": 0.3726,
+      "step": 38095
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.743947556770906,
+      "learning_rate": 2.268466453032675e-06,
+      "loss": 0.3956,
+      "step": 38100
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.578967483277369,
+      "learning_rate": 2.267093787603904e-06,
+      "loss": 0.4061,
+      "step": 38105
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.226628945326729,
+      "learning_rate": 2.2657214158379655e-06,
+      "loss": 0.4034,
+      "step": 38110
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.70623435647588,
+      "learning_rate": 2.2643493378823278e-06,
+      "loss": 0.4067,
+      "step": 38115
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.903708262756084,
+      "learning_rate": 2.262977553884429e-06,
+      "loss": 0.3828,
+      "step": 38120
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.185004150810348,
+      "learning_rate": 2.2616060639916708e-06,
+      "loss": 0.4381,
+      "step": 38125
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 7.827506693720943,
+      "learning_rate": 2.2602348683514258e-06,
+      "loss": 0.4166,
+      "step": 38130
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.383455925629179,
+      "learning_rate": 2.2588639671110366e-06,
+      "loss": 0.3959,
+      "step": 38135
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.104712434454658,
+      "learning_rate": 2.2574933604178105e-06,
+      "loss": 0.402,
+      "step": 38140
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.536926853638174,
+      "learning_rate": 2.2561230484190277e-06,
+      "loss": 0.3932,
+      "step": 38145
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.1498435429993314,
+      "learning_rate": 2.2547530312619335e-06,
+      "loss": 0.4081,
+      "step": 38150
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.455705215270543,
+      "learning_rate": 2.2533833090937397e-06,
+      "loss": 0.3903,
+      "step": 38155
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.053294158079101,
+      "learning_rate": 2.2520138820616326e-06,
+      "loss": 0.4306,
+      "step": 38160
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.471603332557932,
+      "learning_rate": 2.2506447503127606e-06,
+      "loss": 0.4069,
+      "step": 38165
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.981212242396687,
+      "learning_rate": 2.2492759139942456e-06,
+      "loss": 0.3882,
+      "step": 38170
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.909309713283266,
+      "learning_rate": 2.247907373253173e-06,
+      "loss": 0.4151,
+      "step": 38175
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.317573791265963,
+      "learning_rate": 2.246539128236598e-06,
+      "loss": 0.4009,
+      "step": 38180
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.548642843410977,
+      "learning_rate": 2.2451711790915463e-06,
+      "loss": 0.3946,
+      "step": 38185
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.573241544890438,
+      "learning_rate": 2.2438035259650085e-06,
+      "loss": 0.3996,
+      "step": 38190
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.84357767238359,
+      "learning_rate": 2.242436169003946e-06,
+      "loss": 0.4086,
+      "step": 38195
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.789318406802224,
+      "learning_rate": 2.241069108355288e-06,
+      "loss": 0.4105,
+      "step": 38200
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.725955913285168,
+      "learning_rate": 2.23970234416593e-06,
+      "loss": 0.3935,
+      "step": 38205
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.014557303941419,
+      "learning_rate": 2.238335876582738e-06,
+      "loss": 0.4217,
+      "step": 38210
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.219457643630035,
+      "learning_rate": 2.2369697057525427e-06,
+      "loss": 0.3927,
+      "step": 38215
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.8677932223357745,
+      "learning_rate": 2.2356038318221487e-06,
+      "loss": 0.4337,
+      "step": 38220
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.530386893227267,
+      "learning_rate": 2.2342382549383213e-06,
+      "loss": 0.4395,
+      "step": 38225
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 7.585153410202333,
+      "learning_rate": 2.2328729752478013e-06,
+      "loss": 0.421,
+      "step": 38230
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 13.276570454775175,
+      "learning_rate": 2.231507992897292e-06,
+      "loss": 0.4166,
+      "step": 38235
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.083582448661602,
+      "learning_rate": 2.2301433080334663e-06,
+      "loss": 0.397,
+      "step": 38240
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.868368137671233,
+      "learning_rate": 2.2287789208029677e-06,
+      "loss": 0.4214,
+      "step": 38245
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 7.5182045707673195,
+      "learning_rate": 2.2274148313524023e-06,
+      "loss": 0.3842,
+      "step": 38250
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.555929393056832,
+      "learning_rate": 2.2260510398283518e-06,
+      "loss": 0.4197,
+      "step": 38255
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.5329968735218875,
+      "learning_rate": 2.224687546377358e-06,
+      "loss": 0.4124,
+      "step": 38260
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.009927636139155,
+      "learning_rate": 2.223324351145937e-06,
+      "loss": 0.3979,
+      "step": 38265
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.8892108371488465,
+      "learning_rate": 2.2219614542805684e-06,
+      "loss": 0.4366,
+      "step": 38270
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.718257033879936,
+      "learning_rate": 2.2205988559277007e-06,
+      "loss": 0.4097,
+      "step": 38275
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.958336357175654,
+      "learning_rate": 2.2192365562337543e-06,
+      "loss": 0.3917,
+      "step": 38280
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.3575227125026,
+      "learning_rate": 2.21787455534511e-06,
+      "loss": 0.3902,
+      "step": 38285
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.308724346255989,
+      "learning_rate": 2.2165128534081244e-06,
+      "loss": 0.3604,
+      "step": 38290
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.464676475254124,
+      "learning_rate": 2.215151450569118e-06,
+      "loss": 0.3967,
+      "step": 38295
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.891423501714995,
+      "learning_rate": 2.213790346974377e-06,
+      "loss": 0.4005,
+      "step": 38300
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.067428752177997,
+      "learning_rate": 2.212429542770162e-06,
+      "loss": 0.4208,
+      "step": 38305
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.156959061512414,
+      "learning_rate": 2.2110690381026926e-06,
+      "loss": 0.4012,
+      "step": 38310
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.822972155335179,
+      "learning_rate": 2.209708833118166e-06,
+      "loss": 0.3864,
+      "step": 38315
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.006381694705248,
+      "learning_rate": 2.208348927962739e-06,
+      "loss": 0.4321,
+      "step": 38320
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.754068529708678,
+      "learning_rate": 2.206989322782539e-06,
+      "loss": 0.4047,
+      "step": 38325
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 7.63250621373145,
+      "learning_rate": 2.2056300177236643e-06,
+      "loss": 0.3918,
+      "step": 38330
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 7.839191277671578,
+      "learning_rate": 2.2042710129321755e-06,
+      "loss": 0.3878,
+      "step": 38335
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.964535428910256,
+      "learning_rate": 2.202912308554106e-06,
+      "loss": 0.3931,
+      "step": 38340
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.291096054045648,
+      "learning_rate": 2.201553904735452e-06,
+      "loss": 0.4286,
+      "step": 38345
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.955057309545511,
+      "learning_rate": 2.2001958016221826e-06,
+      "loss": 0.407,
+      "step": 38350
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.822559744762919,
+      "learning_rate": 2.1988379993602316e-06,
+      "loss": 0.4114,
+      "step": 38355
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.308815657077496,
+      "learning_rate": 2.197480498095498e-06,
+      "loss": 0.3546,
+      "step": 38360
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.701002282985683,
+      "learning_rate": 2.1961232979738547e-06,
+      "loss": 0.3973,
+      "step": 38365
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.888752659921957,
+      "learning_rate": 2.194766399141136e-06,
+      "loss": 0.3993,
+      "step": 38370
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.080663918185043,
+      "learning_rate": 2.1934098017431498e-06,
+      "loss": 0.3855,
+      "step": 38375
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.684344995501928,
+      "learning_rate": 2.1920535059256655e-06,
+      "loss": 0.4106,
+      "step": 38380
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.901557550814901,
+      "learning_rate": 2.1906975118344236e-06,
+      "loss": 0.4305,
+      "step": 38385
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.495169066401919,
+      "learning_rate": 2.189341819615135e-06,
+      "loss": 0.3851,
+      "step": 38390
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.78500350171946,
+      "learning_rate": 2.18798642941347e-06,
+      "loss": 0.4048,
+      "step": 38395
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.248591999953721,
+      "learning_rate": 2.1866313413750755e-06,
+      "loss": 0.4129,
+      "step": 38400
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.7044268015767985,
+      "learning_rate": 2.1852765556455596e-06,
+      "loss": 0.3839,
+      "step": 38405
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 7.188892401512902,
+      "learning_rate": 2.1839220723704985e-06,
+      "loss": 0.4118,
+      "step": 38410
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.137718121287876,
+      "learning_rate": 2.182567891695441e-06,
+      "loss": 0.4012,
+      "step": 38415
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.67217502753716,
+      "learning_rate": 2.1812140137658955e-06,
+      "loss": 0.4112,
+      "step": 38420
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.455173055441711,
+      "learning_rate": 2.1798604387273463e-06,
+      "loss": 0.3584,
+      "step": 38425
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.803369904154309,
+      "learning_rate": 2.1785071667252376e-06,
+      "loss": 0.3814,
+      "step": 38430
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.908988811537963,
+      "learning_rate": 2.177154197904988e-06,
+      "loss": 0.4184,
+      "step": 38435
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.9731338705417105,
+      "learning_rate": 2.175801532411978e-06,
+      "loss": 0.3926,
+      "step": 38440
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.799991686734702,
+      "learning_rate": 2.1744491703915555e-06,
+      "loss": 0.357,
+      "step": 38445
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.4978227725044455,
+      "learning_rate": 2.1730971119890416e-06,
+      "loss": 0.4002,
+      "step": 38450
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.250610716393017,
+      "learning_rate": 2.171745357349717e-06,
+      "loss": 0.3939,
+      "step": 38455
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.452189601655305,
+      "learning_rate": 2.1703939066188386e-06,
+      "loss": 0.3933,
+      "step": 38460
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.626129117239306,
+      "learning_rate": 2.1690427599416195e-06,
+      "loss": 0.3735,
+      "step": 38465
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.1507599224321305,
+      "learning_rate": 2.1676919174632527e-06,
+      "loss": 0.4189,
+      "step": 38470
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.675997257923313,
+      "learning_rate": 2.1663413793288874e-06,
+      "loss": 0.4034,
+      "step": 38475
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.615745239943095,
+      "learning_rate": 2.164991145683647e-06,
+      "loss": 0.3866,
+      "step": 38480
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.973203534109425,
+      "learning_rate": 2.163641216672621e-06,
+      "loss": 0.4113,
+      "step": 38485
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.184493107621935,
+      "learning_rate": 2.1622915924408646e-06,
+      "loss": 0.4085,
+      "step": 38490
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 11.086631146869495,
+      "learning_rate": 2.1609422731333986e-06,
+      "loss": 0.4083,
+      "step": 38495
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.913882429785973,
+      "learning_rate": 2.1595932588952164e-06,
+      "loss": 0.4175,
+      "step": 38500
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.813323665594315,
+      "learning_rate": 2.158244549871273e-06,
+      "loss": 0.3822,
+      "step": 38505
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.628538006571762,
+      "learning_rate": 2.156896146206495e-06,
+      "loss": 0.404,
+      "step": 38510
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.013120678653632,
+      "learning_rate": 2.155548048045773e-06,
+      "loss": 0.4125,
+      "step": 38515
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.113861884290618,
+      "learning_rate": 2.154200255533968e-06,
+      "loss": 0.3699,
+      "step": 38520
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.887863434683157,
+      "learning_rate": 2.1528527688159057e-06,
+      "loss": 0.4111,
+      "step": 38525
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.481691056906396,
+      "learning_rate": 2.1515055880363763e-06,
+      "loss": 0.4025,
+      "step": 38530
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.395663483698197,
+      "learning_rate": 2.1501587133401445e-06,
+      "loss": 0.394,
+      "step": 38535
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.6222518839369995,
+      "learning_rate": 2.148812144871935e-06,
+      "loss": 0.4309,
+      "step": 38540
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.745201229958961,
+      "learning_rate": 2.1474658827764453e-06,
+      "loss": 0.3957,
+      "step": 38545
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.477977586552221,
+      "learning_rate": 2.146119927198336e-06,
+      "loss": 0.3897,
+      "step": 38550
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.428390864853101,
+      "learning_rate": 2.1447742782822338e-06,
+      "loss": 0.3829,
+      "step": 38555
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.545247530486125,
+      "learning_rate": 2.1434289361727363e-06,
+      "loss": 0.41,
+      "step": 38560
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.19207433195976,
+      "learning_rate": 2.1420839010144086e-06,
+      "loss": 0.413,
+      "step": 38565
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.647612086719953,
+      "learning_rate": 2.140739172951777e-06,
+      "loss": 0.4293,
+      "step": 38570
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.410536160373945,
+      "learning_rate": 2.1393947521293413e-06,
+      "loss": 0.3921,
+      "step": 38575
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.049584885388937,
+      "learning_rate": 2.138050638691563e-06,
+      "loss": 0.4028,
+      "step": 38580
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.769364112906789,
+      "learning_rate": 2.136706832782876e-06,
+      "loss": 0.398,
+      "step": 38585
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.76686710250167,
+      "learning_rate": 2.1353633345476742e-06,
+      "loss": 0.3892,
+      "step": 38590
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.290813454338231,
+      "learning_rate": 2.134020144130327e-06,
+      "loss": 0.387,
+      "step": 38595
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.534713155982688,
+      "learning_rate": 2.1326772616751623e-06,
+      "loss": 0.4065,
+      "step": 38600
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.9098049548268055,
+      "learning_rate": 2.1313346873264816e-06,
+      "loss": 0.3835,
+      "step": 38605
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.695743105993168,
+      "learning_rate": 2.1299924212285493e-06,
+      "loss": 0.399,
+      "step": 38610
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.555631939857085,
+      "learning_rate": 2.1286504635255963e-06,
+      "loss": 0.3786,
+      "step": 38615
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.358899041852783,
+      "learning_rate": 2.127308814361825e-06,
+      "loss": 0.3922,
+      "step": 38620
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.6645020119128695,
+      "learning_rate": 2.125967473881397e-06,
+      "loss": 0.3971,
+      "step": 38625
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.924011309498007,
+      "learning_rate": 2.1246264422284507e-06,
+      "loss": 0.3702,
+      "step": 38630
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.433465364614165,
+      "learning_rate": 2.1232857195470825e-06,
+      "loss": 0.3822,
+      "step": 38635
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.765182060416091,
+      "learning_rate": 2.1219453059813583e-06,
+      "loss": 0.4016,
+      "step": 38640
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.181989455395341,
+      "learning_rate": 2.120605201675315e-06,
+      "loss": 0.3979,
+      "step": 38645
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.00143295704809,
+      "learning_rate": 2.119265406772948e-06,
+      "loss": 0.4022,
+      "step": 38650
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.117214685595613,
+      "learning_rate": 2.1179259214182267e-06,
+      "loss": 0.4112,
+      "step": 38655
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.8399465080894135,
+      "learning_rate": 2.1165867457550865e-06,
+      "loss": 0.3927,
+      "step": 38660
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.807864459149209,
+      "learning_rate": 2.1152478799274247e-06,
+      "loss": 0.3951,
+      "step": 38665
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.6053442197291155,
+      "learning_rate": 2.113909324079111e-06,
+      "loss": 0.4068,
+      "step": 38670
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.50249582365174,
+      "learning_rate": 2.112571078353976e-06,
+      "loss": 0.4235,
+      "step": 38675
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.3688072795808734,
+      "learning_rate": 2.1112331428958233e-06,
+      "loss": 0.3739,
+      "step": 38680
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.292860673413531,
+      "learning_rate": 2.109895517848417e-06,
+      "loss": 0.4191,
+      "step": 38685
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.89379053199037,
+      "learning_rate": 2.108558203355494e-06,
+      "loss": 0.3884,
+      "step": 38690
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.765328355765687,
+      "learning_rate": 2.1072211995607534e-06,
+      "loss": 0.3926,
+      "step": 38695
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.6605360820933175,
+      "learning_rate": 2.105884506607859e-06,
+      "loss": 0.4142,
+      "step": 38700
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.239545475837199,
+      "learning_rate": 2.1045481246404497e-06,
+      "loss": 0.4105,
+      "step": 38705
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.203008940340655,
+      "learning_rate": 2.103212053802121e-06,
+      "loss": 0.3764,
+      "step": 38710
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.426397186096734,
+      "learning_rate": 2.1018762942364437e-06,
+      "loss": 0.3862,
+      "step": 38715
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.306825500480146,
+      "learning_rate": 2.1005408460869486e-06,
+      "loss": 0.4287,
+      "step": 38720
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.715503113905795,
+      "learning_rate": 2.0992057094971342e-06,
+      "loss": 0.3981,
+      "step": 38725
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 8.98365887608179,
+      "learning_rate": 2.09787088461047e-06,
+      "loss": 0.3854,
+      "step": 38730
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.227889910904786,
+      "learning_rate": 2.0965363715703857e-06,
+      "loss": 0.4156,
+      "step": 38735
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.7393308788034645,
+      "learning_rate": 2.095202170520284e-06,
+      "loss": 0.4218,
+      "step": 38740
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.292223087645373,
+      "learning_rate": 2.093868281603527e-06,
+      "loss": 0.3964,
+      "step": 38745
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.139624667569269,
+      "learning_rate": 2.092534704963449e-06,
+      "loss": 0.3898,
+      "step": 38750
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.737970076906442,
+      "learning_rate": 2.0912014407433507e-06,
+      "loss": 0.3972,
+      "step": 38755
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.5147933770327064,
+      "learning_rate": 2.0898684890864935e-06,
+      "loss": 0.3883,
+      "step": 38760
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 7.4293994473746325,
+      "learning_rate": 2.088535850136112e-06,
+      "loss": 0.3997,
+      "step": 38765
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.845198988293049,
+      "learning_rate": 2.0872035240354007e-06,
+      "loss": 0.3963,
+      "step": 38770
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.704743055052215,
+      "learning_rate": 2.0858715109275273e-06,
+      "loss": 0.3881,
+      "step": 38775
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.986911366585578,
+      "learning_rate": 2.084539810955622e-06,
+      "loss": 0.3865,
+      "step": 38780
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.115826486255224,
+      "learning_rate": 2.0832084242627783e-06,
+      "loss": 0.421,
+      "step": 38785
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.602096951745674,
+      "learning_rate": 2.081877350992064e-06,
+      "loss": 0.3861,
+      "step": 38790
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.256387121064838,
+      "learning_rate": 2.080546591286506e-06,
+      "loss": 0.4074,
+      "step": 38795
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.451157458162828,
+      "learning_rate": 2.0792161452891025e-06,
+      "loss": 0.3878,
+      "step": 38800
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.966273887512294,
+      "learning_rate": 2.077886013142815e-06,
+      "loss": 0.4146,
+      "step": 38805
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.5331927343856355,
+      "learning_rate": 2.0765561949905693e-06,
+      "loss": 0.3943,
+      "step": 38810
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.609237308128622,
+      "learning_rate": 2.075226690975265e-06,
+      "loss": 0.4112,
+      "step": 38815
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.112732306750538,
+      "learning_rate": 2.0738975012397585e-06,
+      "loss": 0.3876,
+      "step": 38820
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.8975734575630705,
+      "learning_rate": 2.0725686259268813e-06,
+      "loss": 0.4287,
+      "step": 38825
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.401027482380962,
+      "learning_rate": 2.0712400651794233e-06,
+      "loss": 0.4038,
+      "step": 38830
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.435267720334556,
+      "learning_rate": 2.069911819140147e-06,
+      "loss": 0.3815,
+      "step": 38835
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.553894346786587,
+      "learning_rate": 2.0685838879517767e-06,
+      "loss": 0.3821,
+      "step": 38840
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.713386969070219,
+      "learning_rate": 2.0672562717570043e-06,
+      "loss": 0.3907,
+      "step": 38845
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.167098796179337,
+      "learning_rate": 2.0659289706984903e-06,
+      "loss": 0.3949,
+      "step": 38850
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.344477417274741,
+      "learning_rate": 2.064601984918856e-06,
+      "loss": 0.4008,
+      "step": 38855
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.603400445501257,
+      "learning_rate": 2.063275314560695e-06,
+      "loss": 0.4309,
+      "step": 38860
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.237801260847373,
+      "learning_rate": 2.0619489597665622e-06,
+      "loss": 0.3989,
+      "step": 38865
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.878509897529411,
+      "learning_rate": 2.060622920678978e-06,
+      "loss": 0.4162,
+      "step": 38870
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.413789470759705,
+      "learning_rate": 2.0592971974404358e-06,
+      "loss": 0.3897,
+      "step": 38875
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.142975068918899,
+      "learning_rate": 2.057971790193386e-06,
+      "loss": 0.3487,
+      "step": 38880
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.3539662990290084,
+      "learning_rate": 2.0566466990802534e-06,
+      "loss": 0.404,
+      "step": 38885
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.64585346887782,
+      "learning_rate": 2.055321924243423e-06,
+      "loss": 0.3867,
+      "step": 38890
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.234958381467139,
+      "learning_rate": 2.053997465825246e-06,
+      "loss": 0.3826,
+      "step": 38895
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.568749889569857,
+      "learning_rate": 2.0526733239680445e-06,
+      "loss": 0.3982,
+      "step": 38900
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.955018933633224,
+      "learning_rate": 2.0513494988141002e-06,
+      "loss": 0.439,
+      "step": 38905
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.116998926552031,
+      "learning_rate": 2.0500259905056676e-06,
+      "loss": 0.4243,
+      "step": 38910
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.268386539561251,
+      "learning_rate": 2.0487027991849594e-06,
+      "loss": 0.4411,
+      "step": 38915
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.201328055758747,
+      "learning_rate": 2.0473799249941627e-06,
+      "loss": 0.3836,
+      "step": 38920
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.970729449189368,
+      "learning_rate": 2.046057368075422e-06,
+      "loss": 0.4083,
+      "step": 38925
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.8883094970171515,
+      "learning_rate": 2.0447351285708554e-06,
+      "loss": 0.4092,
+      "step": 38930
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.49127190563385,
+      "learning_rate": 2.0434132066225405e-06,
+      "loss": 0.4042,
+      "step": 38935
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.313038945917286,
+      "learning_rate": 2.0420916023725244e-06,
+      "loss": 0.3974,
+      "step": 38940
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.701695780740329,
+      "learning_rate": 2.0407703159628224e-06,
+      "loss": 0.3653,
+      "step": 38945
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.363619762417488,
+      "learning_rate": 2.0394493475354094e-06,
+      "loss": 0.4098,
+      "step": 38950
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.816465055287263,
+      "learning_rate": 2.0381286972322283e-06,
+      "loss": 0.3834,
+      "step": 38955
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.314960838394469,
+      "learning_rate": 2.0368083651951927e-06,
+      "loss": 0.4023,
+      "step": 38960
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.389517851042226,
+      "learning_rate": 2.0354883515661734e-06,
+      "loss": 0.4249,
+      "step": 38965
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.597833283499703,
+      "learning_rate": 2.034168656487016e-06,
+      "loss": 0.4006,
+      "step": 38970
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.730014782523671,
+      "learning_rate": 2.032849280099526e-06,
+      "loss": 0.3893,
+      "step": 38975
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.6965649993995875,
+      "learning_rate": 2.0315302225454736e-06,
+      "loss": 0.3662,
+      "step": 38980
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.46922335185519,
+      "learning_rate": 2.0302114839666017e-06,
+      "loss": 0.4162,
+      "step": 38985
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.213379806054316,
+      "learning_rate": 2.0288930645046106e-06,
+      "loss": 0.3651,
+      "step": 38990
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.147714807298288,
+      "learning_rate": 2.0275749643011745e-06,
+      "loss": 0.3979,
+      "step": 38995
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.725775462814913,
+      "learning_rate": 2.0262571834979243e-06,
+      "loss": 0.3938,
+      "step": 39000
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.445150688922724,
+      "learning_rate": 2.0249397222364664e-06,
+      "loss": 0.3984,
+      "step": 39005
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.795208073400065,
+      "learning_rate": 2.023622580658365e-06,
+      "loss": 0.3909,
+      "step": 39010
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.817298566269459,
+      "learning_rate": 2.022305758905151e-06,
+      "loss": 0.372,
+      "step": 39015
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.78488919786896,
+      "learning_rate": 2.0209892571183247e-06,
+      "loss": 0.3942,
+      "step": 39020
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.852040691074928,
+      "learning_rate": 2.0196730754393524e-06,
+      "loss": 0.408,
+      "step": 39025
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.757677745792287,
+      "learning_rate": 2.018357214009659e-06,
+      "loss": 0.418,
+      "step": 39030
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.603541620603779,
+      "learning_rate": 2.0170416729706437e-06,
+      "loss": 0.3866,
+      "step": 39035
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.59621435926493,
+      "learning_rate": 2.0157264524636644e-06,
+      "loss": 0.4072,
+      "step": 39040
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.7671967765864895,
+      "learning_rate": 2.0144115526300495e-06,
+      "loss": 0.3887,
+      "step": 39045
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.680602077720724,
+      "learning_rate": 2.013096973611088e-06,
+      "loss": 0.4106,
+      "step": 39050
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.867675568728056,
+      "learning_rate": 2.0117827155480405e-06,
+      "loss": 0.3577,
+      "step": 39055
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.239896507137009,
+      "learning_rate": 2.0104687785821286e-06,
+      "loss": 0.375,
+      "step": 39060
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.2848515735994335,
+      "learning_rate": 2.0091551628545385e-06,
+      "loss": 0.4027,
+      "step": 39065
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.980979996629347,
+      "learning_rate": 2.0078418685064278e-06,
+      "loss": 0.3909,
+      "step": 39070
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.828427743870696,
+      "learning_rate": 2.0065288956789114e-06,
+      "loss": 0.4115,
+      "step": 39075
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.874373314519098,
+      "learning_rate": 2.005216244513078e-06,
+      "loss": 0.4032,
+      "step": 39080
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.500648376458064,
+      "learning_rate": 2.0039039151499747e-06,
+      "loss": 0.3872,
+      "step": 39085
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.178901616250842,
+      "learning_rate": 2.0025919077306206e-06,
+      "loss": 0.4075,
+      "step": 39090
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.387145842189551,
+      "learning_rate": 2.0012802223959937e-06,
+      "loss": 0.4047,
+      "step": 39095
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.60506482068075,
+      "learning_rate": 1.99996885928704e-06,
+      "loss": 0.3839,
+      "step": 39100
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.5306047194833114,
+      "learning_rate": 1.998657818544674e-06,
+      "loss": 0.425,
+      "step": 39105
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.825596400925807,
+      "learning_rate": 1.9973471003097698e-06,
+      "loss": 0.3949,
+      "step": 39110
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.922033121496603,
+      "learning_rate": 1.9960367047231715e-06,
+      "loss": 0.4018,
+      "step": 39115
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.597560061172928,
+      "learning_rate": 1.9947266319256887e-06,
+      "loss": 0.4021,
+      "step": 39120
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.657990797184312,
+      "learning_rate": 1.9934168820580907e-06,
+      "loss": 0.3968,
+      "step": 39125
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.209202164190039,
+      "learning_rate": 1.992107455261119e-06,
+      "loss": 0.407,
+      "step": 39130
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.386116825533306,
+      "learning_rate": 1.9907983516754754e-06,
+      "loss": 0.3963,
+      "step": 39135
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 8.269427638207159,
+      "learning_rate": 1.9894895714418303e-06,
+      "loss": 0.3869,
+      "step": 39140
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 6.174116633717544,
+      "learning_rate": 1.9881811147008158e-06,
+      "loss": 0.3919,
+      "step": 39145
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.734472965468408,
+      "learning_rate": 1.986872981593034e-06,
+      "loss": 0.4254,
+      "step": 39150
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.022932641200943,
+      "learning_rate": 1.985565172259049e-06,
+      "loss": 0.3946,
+      "step": 39155
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.498077637297552,
+      "learning_rate": 1.9842576868393883e-06,
+      "loss": 0.3717,
+      "step": 39160
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 6.5332225905143355,
+      "learning_rate": 1.9829505254745497e-06,
+      "loss": 0.3831,
+      "step": 39165
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.93601416254683,
+      "learning_rate": 1.9816436883049906e-06,
+      "loss": 0.4047,
+      "step": 39170
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.733251523796037,
+      "learning_rate": 1.9803371754711402e-06,
+      "loss": 0.397,
+      "step": 39175
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.712771353063706,
+      "learning_rate": 1.979030987113387e-06,
+      "loss": 0.4272,
+      "step": 39180
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 8.283336273144208,
+      "learning_rate": 1.977725123372085e-06,
+      "loss": 0.3841,
+      "step": 39185
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.748921865335893,
+      "learning_rate": 1.976419584387559e-06,
+      "loss": 0.4207,
+      "step": 39190
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.062050962654206,
+      "learning_rate": 1.975114370300091e-06,
+      "loss": 0.392,
+      "step": 39195
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.014145429863991,
+      "learning_rate": 1.973809481249935e-06,
+      "loss": 0.3957,
+      "step": 39200
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.318200049638917,
+      "learning_rate": 1.972504917377304e-06,
+      "loss": 0.3921,
+      "step": 39205
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.624013880590756,
+      "learning_rate": 1.9712006788223813e-06,
+      "loss": 0.393,
+      "step": 39210
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.339503008943358,
+      "learning_rate": 1.9698967657253147e-06,
+      "loss": 0.3549,
+      "step": 39215
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.19103963618126,
+      "learning_rate": 1.9685931782262113e-06,
+      "loss": 0.4061,
+      "step": 39220
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.515879538966008,
+      "learning_rate": 1.9672899164651516e-06,
+      "loss": 0.3626,
+      "step": 39225
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.761942211722023,
+      "learning_rate": 1.965986980582173e-06,
+      "loss": 0.4116,
+      "step": 39230
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.459228425857306,
+      "learning_rate": 1.964684370717285e-06,
+      "loss": 0.4066,
+      "step": 39235
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.445562264940093,
+      "learning_rate": 1.9633820870104575e-06,
+      "loss": 0.4106,
+      "step": 39240
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.682108648769833,
+      "learning_rate": 1.9620801296016247e-06,
+      "loss": 0.3848,
+      "step": 39245
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 6.744454442004688,
+      "learning_rate": 1.9607784986306906e-06,
+      "loss": 0.3727,
+      "step": 39250
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.539855465939607,
+      "learning_rate": 1.9594771942375183e-06,
+      "loss": 0.3694,
+      "step": 39255
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.9894757678328965,
+      "learning_rate": 1.9581762165619423e-06,
+      "loss": 0.3753,
+      "step": 39260
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.5534286960075,
+      "learning_rate": 1.9568755657437564e-06,
+      "loss": 0.3911,
+      "step": 39265
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.659752904131898,
+      "learning_rate": 1.95557524192272e-06,
+      "loss": 0.3972,
+      "step": 39270
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.808294937638788,
+      "learning_rate": 1.9542752452385604e-06,
+      "loss": 0.3711,
+      "step": 39275
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.816358919335911,
+      "learning_rate": 1.9529755758309666e-06,
+      "loss": 0.3919,
+      "step": 39280
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.6226762383187365,
+      "learning_rate": 1.951676233839597e-06,
+      "loss": 0.4161,
+      "step": 39285
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.479114808354138,
+      "learning_rate": 1.9503772194040665e-06,
+      "loss": 0.3592,
+      "step": 39290
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.579325120042945,
+      "learning_rate": 1.949078532663965e-06,
+      "loss": 0.3956,
+      "step": 39295
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.889692048547527,
+      "learning_rate": 1.947780173758838e-06,
+      "loss": 0.413,
+      "step": 39300
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.612972680628571,
+      "learning_rate": 1.9464821428282023e-06,
+      "loss": 0.3659,
+      "step": 39305
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.312394924222412,
+      "learning_rate": 1.9451844400115378e-06,
+      "loss": 0.3766,
+      "step": 39310
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.0878461852833485,
+      "learning_rate": 1.9438870654482855e-06,
+      "loss": 0.3724,
+      "step": 39315
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.310007001105772,
+      "learning_rate": 1.9425900192778575e-06,
+      "loss": 0.3477,
+      "step": 39320
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.810854392040784,
+      "learning_rate": 1.9412933016396256e-06,
+      "loss": 0.4027,
+      "step": 39325
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.403281496748654,
+      "learning_rate": 1.939996912672925e-06,
+      "loss": 0.3941,
+      "step": 39330
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.529455231183697,
+      "learning_rate": 1.9387008525170635e-06,
+      "loss": 0.4147,
+      "step": 39335
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.626997511243248,
+      "learning_rate": 1.937405121311304e-06,
+      "loss": 0.3761,
+      "step": 39340
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.929562626534522,
+      "learning_rate": 1.9361097191948817e-06,
+      "loss": 0.3798,
+      "step": 39345
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.040053761044108,
+      "learning_rate": 1.9348146463069928e-06,
+      "loss": 0.3693,
+      "step": 39350
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.377633274374966,
+      "learning_rate": 1.9335199027867955e-06,
+      "loss": 0.4228,
+      "step": 39355
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.109222863158795,
+      "learning_rate": 1.9322254887734203e-06,
+      "loss": 0.392,
+      "step": 39360
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.132264102947876,
+      "learning_rate": 1.9309314044059537e-06,
+      "loss": 0.3789,
+      "step": 39365
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.976285053201243,
+      "learning_rate": 1.9296376498234543e-06,
+      "loss": 0.4071,
+      "step": 39370
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.992502688260796,
+      "learning_rate": 1.9283442251649394e-06,
+      "loss": 0.3753,
+      "step": 39375
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.075911079635343,
+      "learning_rate": 1.927051130569392e-06,
+      "loss": 0.4151,
+      "step": 39380
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.429071718373526,
+      "learning_rate": 1.9257583661757634e-06,
+      "loss": 0.3825,
+      "step": 39385
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.5806253644574175,
+      "learning_rate": 1.9244659321229668e-06,
+      "loss": 0.3985,
+      "step": 39390
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.494931489782704,
+      "learning_rate": 1.9231738285498785e-06,
+      "loss": 0.3722,
+      "step": 39395
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.260413983614203,
+      "learning_rate": 1.9218820555953405e-06,
+      "loss": 0.4215,
+      "step": 39400
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.897041511698496,
+      "learning_rate": 1.9205906133981627e-06,
+      "loss": 0.3941,
+      "step": 39405
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.925661136097314,
+      "learning_rate": 1.9192995020971133e-06,
+      "loss": 0.4121,
+      "step": 39410
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.636157699110875,
+      "learning_rate": 1.9180087218309273e-06,
+      "loss": 0.3852,
+      "step": 39415
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.63613934585706,
+      "learning_rate": 1.916718272738308e-06,
+      "loss": 0.3939,
+      "step": 39420
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.925446808807623,
+      "learning_rate": 1.9154281549579157e-06,
+      "loss": 0.3614,
+      "step": 39425
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.436743271862936,
+      "learning_rate": 1.914138368628383e-06,
+      "loss": 0.374,
+      "step": 39430
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.576896601885174,
+      "learning_rate": 1.9128489138883017e-06,
+      "loss": 0.4257,
+      "step": 39435
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.723575563308978,
+      "learning_rate": 1.9115597908762277e-06,
+      "loss": 0.3908,
+      "step": 39440
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.41529628446949,
+      "learning_rate": 1.9102709997306863e-06,
+      "loss": 0.3761,
+      "step": 39445
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.758141671306461,
+      "learning_rate": 1.9089825405901597e-06,
+      "loss": 0.3785,
+      "step": 39450
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.413224050824788,
+      "learning_rate": 1.907694413593103e-06,
+      "loss": 0.4214,
+      "step": 39455
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.798702141896009,
+      "learning_rate": 1.906406618877929e-06,
+      "loss": 0.3903,
+      "step": 39460
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.287535595220807,
+      "learning_rate": 1.9051191565830147e-06,
+      "loss": 0.3838,
+      "step": 39465
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.940844969553668,
+      "learning_rate": 1.9038320268467081e-06,
+      "loss": 0.3555,
+      "step": 39470
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.763730501188215,
+      "learning_rate": 1.9025452298073127e-06,
+      "loss": 0.3911,
+      "step": 39475
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.867809024857266,
+      "learning_rate": 1.9012587656031022e-06,
+      "loss": 0.4029,
+      "step": 39480
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.606378094814525,
+      "learning_rate": 1.899972634372315e-06,
+      "loss": 0.4101,
+      "step": 39485
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.950571708424779,
+      "learning_rate": 1.8986868362531479e-06,
+      "loss": 0.3846,
+      "step": 39490
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.639142141166102,
+      "learning_rate": 1.897401371383769e-06,
+      "loss": 0.4156,
+      "step": 39495
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 7.521940871616371,
+      "learning_rate": 1.8961162399023043e-06,
+      "loss": 0.4028,
+      "step": 39500
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.5318394373804045,
+      "learning_rate": 1.8948314419468495e-06,
+      "loss": 0.4242,
+      "step": 39505
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 7.476824102881447,
+      "learning_rate": 1.8935469776554589e-06,
+      "loss": 0.4121,
+      "step": 39510
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.751126129917067,
+      "learning_rate": 1.892262847166157e-06,
+      "loss": 0.3759,
+      "step": 39515
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.399312116418375,
+      "learning_rate": 1.8909790506169273e-06,
+      "loss": 0.3744,
+      "step": 39520
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.369512931193964,
+      "learning_rate": 1.8896955881457184e-06,
+      "loss": 0.4174,
+      "step": 39525
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.939418429983109,
+      "learning_rate": 1.8884124598904474e-06,
+      "loss": 0.3882,
+      "step": 39530
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.672525721857139,
+      "learning_rate": 1.8871296659889877e-06,
+      "loss": 0.3809,
+      "step": 39535
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.671526954543785,
+      "learning_rate": 1.8858472065791856e-06,
+      "loss": 0.3968,
+      "step": 39540
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.412224228577191,
+      "learning_rate": 1.8845650817988454e-06,
+      "loss": 0.3654,
+      "step": 39545
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.236915588437215,
+      "learning_rate": 1.8832832917857341e-06,
+      "loss": 0.4079,
+      "step": 39550
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.086324031331009,
+      "learning_rate": 1.8820018366775906e-06,
+      "loss": 0.3886,
+      "step": 39555
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.207902869035254,
+      "learning_rate": 1.8807207166121084e-06,
+      "loss": 0.4152,
+      "step": 39560
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.762951690033797,
+      "learning_rate": 1.879439931726953e-06,
+      "loss": 0.3658,
+      "step": 39565
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.518356343370002,
+      "learning_rate": 1.8781594821597477e-06,
+      "loss": 0.3787,
+      "step": 39570
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.805396373492158,
+      "learning_rate": 1.876879368048084e-06,
+      "loss": 0.366,
+      "step": 39575
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.932263168030357,
+      "learning_rate": 1.8755995895295171e-06,
+      "loss": 0.4302,
+      "step": 39580
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.785857077956581,
+      "learning_rate": 1.8743201467415617e-06,
+      "loss": 0.4024,
+      "step": 39585
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.184166770390915,
+      "learning_rate": 1.8730410398217024e-06,
+      "loss": 0.3935,
+      "step": 39590
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.5109971914732245,
+      "learning_rate": 1.8717622689073827e-06,
+      "loss": 0.3899,
+      "step": 39595
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.398428167430032,
+      "learning_rate": 1.8704838341360148e-06,
+      "loss": 0.3577,
+      "step": 39600
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.951762431716239,
+      "learning_rate": 1.86920573564497e-06,
+      "loss": 0.3954,
+      "step": 39605
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 7.120061651291709,
+      "learning_rate": 1.867927973571585e-06,
+      "loss": 0.402,
+      "step": 39610
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.376852122443278,
+      "learning_rate": 1.8666505480531638e-06,
+      "loss": 0.4175,
+      "step": 39615
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.659641246982101,
+      "learning_rate": 1.8653734592269673e-06,
+      "loss": 0.3508,
+      "step": 39620
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.020145632739526,
+      "learning_rate": 1.8640967072302292e-06,
+      "loss": 0.4005,
+      "step": 39625
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.11957725384284,
+      "learning_rate": 1.8628202922001381e-06,
+      "loss": 0.3824,
+      "step": 39630
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.660902042465461,
+      "learning_rate": 1.8615442142738531e-06,
+      "loss": 0.3362,
+      "step": 39635
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.924188483670731,
+      "learning_rate": 1.860268473588494e-06,
+      "loss": 0.3803,
+      "step": 39640
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.923638465380285,
+      "learning_rate": 1.858993070281142e-06,
+      "loss": 0.3908,
+      "step": 39645
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.373037236508251,
+      "learning_rate": 1.8577180044888487e-06,
+      "loss": 0.3496,
+      "step": 39650
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.310046001357836,
+      "learning_rate": 1.8564432763486222e-06,
+      "loss": 0.3661,
+      "step": 39655
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.196622318158273,
+      "learning_rate": 1.855168885997441e-06,
+      "loss": 0.3816,
+      "step": 39660
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.044666895592206,
+      "learning_rate": 1.8538948335722402e-06,
+      "loss": 0.369,
+      "step": 39665
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.962374557992139,
+      "learning_rate": 1.852621119209924e-06,
+      "loss": 0.3959,
+      "step": 39670
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.746987427245066,
+      "learning_rate": 1.8513477430473615e-06,
+      "loss": 0.4152,
+      "step": 39675
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.162547842833919,
+      "learning_rate": 1.850074705221378e-06,
+      "loss": 0.3811,
+      "step": 39680
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.004663789773719,
+      "learning_rate": 1.8488020058687705e-06,
+      "loss": 0.3638,
+      "step": 39685
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.321604927292,
+      "learning_rate": 1.8475296451262953e-06,
+      "loss": 0.3911,
+      "step": 39690
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.805710004138573,
+      "learning_rate": 1.8462576231306706e-06,
+      "loss": 0.41,
+      "step": 39695
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.547511515704962,
+      "learning_rate": 1.844985940018585e-06,
+      "loss": 0.3666,
+      "step": 39700
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.961191279493139,
+      "learning_rate": 1.8437145959266821e-06,
+      "loss": 0.3927,
+      "step": 39705
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.787963243117133,
+      "learning_rate": 1.8424435909915773e-06,
+      "loss": 0.4067,
+      "step": 39710
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.272288553057736,
+      "learning_rate": 1.8411729253498428e-06,
+      "loss": 0.3916,
+      "step": 39715
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.020562475059103,
+      "learning_rate": 1.8399025991380197e-06,
+      "loss": 0.4038,
+      "step": 39720
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.656818523628417,
+      "learning_rate": 1.838632612492609e-06,
+      "loss": 0.3904,
+      "step": 39725
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.092253368498932,
+      "learning_rate": 1.8373629655500746e-06,
+      "loss": 0.3799,
+      "step": 39730
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.8502567763066775,
+      "learning_rate": 1.8360936584468486e-06,
+      "loss": 0.3908,
+      "step": 39735
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.287050953157093,
+      "learning_rate": 1.8348246913193213e-06,
+      "loss": 0.3703,
+      "step": 39740
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.095888948582343,
+      "learning_rate": 1.8335560643038514e-06,
+      "loss": 0.3924,
+      "step": 39745
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.388213094768144,
+      "learning_rate": 1.8322877775367553e-06,
+      "loss": 0.391,
+      "step": 39750
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.940983560238933,
+      "learning_rate": 1.8310198311543192e-06,
+      "loss": 0.3825,
+      "step": 39755
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.368845982457075,
+      "learning_rate": 1.8297522252927863e-06,
+      "loss": 0.403,
+      "step": 39760
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.800167607118869,
+      "learning_rate": 1.8284849600883686e-06,
+      "loss": 0.3962,
+      "step": 39765
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.612145761309282,
+      "learning_rate": 1.82721803567724e-06,
+      "loss": 0.3682,
+      "step": 39770
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.812702118365965,
+      "learning_rate": 1.8259514521955368e-06,
+      "loss": 0.4102,
+      "step": 39775
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.891077334074156,
+      "learning_rate": 1.824685209779356e-06,
+      "loss": 0.4098,
+      "step": 39780
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.45152695790426,
+      "learning_rate": 1.823419308564765e-06,
+      "loss": 0.4017,
+      "step": 39785
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.7630537946305225,
+      "learning_rate": 1.8221537486877866e-06,
+      "loss": 0.4222,
+      "step": 39790
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.317864713299714,
+      "learning_rate": 1.8208885302844147e-06,
+      "loss": 0.4157,
+      "step": 39795
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.761185300543549,
+      "learning_rate": 1.8196236534905992e-06,
+      "loss": 0.3916,
+      "step": 39800
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.30928314552267,
+      "learning_rate": 1.8183591184422594e-06,
+      "loss": 0.364,
+      "step": 39805
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.9886560228569525,
+      "learning_rate": 1.8170949252752745e-06,
+      "loss": 0.3872,
+      "step": 39810
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.847642331118586,
+      "learning_rate": 1.8158310741254853e-06,
+      "loss": 0.3695,
+      "step": 39815
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.640814823407414,
+      "learning_rate": 1.8145675651287015e-06,
+      "loss": 0.3861,
+      "step": 39820
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.283218923150907,
+      "learning_rate": 1.8133043984206894e-06,
+      "loss": 0.3582,
+      "step": 39825
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.65729786494483,
+      "learning_rate": 1.8120415741371855e-06,
+      "loss": 0.3759,
+      "step": 39830
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.579537865520433,
+      "learning_rate": 1.8107790924138847e-06,
+      "loss": 0.3697,
+      "step": 39835
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.80671901708008,
+      "learning_rate": 1.8095169533864427e-06,
+      "loss": 0.3932,
+      "step": 39840
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.830477689387796,
+      "learning_rate": 1.8082551571904854e-06,
+      "loss": 0.3897,
+      "step": 39845
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.8268395670951865,
+      "learning_rate": 1.8069937039615992e-06,
+      "loss": 0.4089,
+      "step": 39850
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.91057321248474,
+      "learning_rate": 1.8057325938353298e-06,
+      "loss": 0.3883,
+      "step": 39855
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.285469458486904,
+      "learning_rate": 1.8044718269471918e-06,
+      "loss": 0.3751,
+      "step": 39860
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.551259040139519,
+      "learning_rate": 1.8032114034326576e-06,
+      "loss": 0.396,
+      "step": 39865
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.692882712259712,
+      "learning_rate": 1.8019513234271684e-06,
+      "loss": 0.3856,
+      "step": 39870
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.600972094497757,
+      "learning_rate": 1.8006915870661218e-06,
+      "loss": 0.3806,
+      "step": 39875
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.49988349118192,
+      "learning_rate": 1.7994321944848853e-06,
+      "loss": 0.3649,
+      "step": 39880
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.118171304183476,
+      "learning_rate": 1.7981731458187828e-06,
+      "loss": 0.3672,
+      "step": 39885
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.591487130933895,
+      "learning_rate": 1.7969144412031086e-06,
+      "loss": 0.3517,
+      "step": 39890
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.145874648692256,
+      "learning_rate": 1.7956560807731144e-06,
+      "loss": 0.3779,
+      "step": 39895
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 9.145632912209214,
+      "learning_rate": 1.7943980646640136e-06,
+      "loss": 0.3989,
+      "step": 39900
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.613315290941895,
+      "learning_rate": 1.7931403930109904e-06,
+      "loss": 0.385,
+      "step": 39905
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 8.404404999385008,
+      "learning_rate": 1.7918830659491832e-06,
+      "loss": 0.4078,
+      "step": 39910
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.832867129164264,
+      "learning_rate": 1.7906260836137e-06,
+      "loss": 0.4082,
+      "step": 39915
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.84273079348836,
+      "learning_rate": 1.7893694461396093e-06,
+      "loss": 0.4031,
+      "step": 39920
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.085281413321037,
+      "learning_rate": 1.788113153661939e-06,
+      "loss": 0.3423,
+      "step": 39925
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.2583943463115554,
+      "learning_rate": 1.7868572063156868e-06,
+      "loss": 0.3832,
+      "step": 39930
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 8.184451483121519,
+      "learning_rate": 1.7856016042358076e-06,
+      "loss": 0.3908,
+      "step": 39935
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.637568268935216,
+      "learning_rate": 1.7843463475572216e-06,
+      "loss": 0.3583,
+      "step": 39940
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.988223285001286,
+      "learning_rate": 1.7830914364148145e-06,
+      "loss": 0.3629,
+      "step": 39945
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.474123226405208,
+      "learning_rate": 1.7818368709434276e-06,
+      "loss": 0.3732,
+      "step": 39950
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.610931757106105,
+      "learning_rate": 1.7805826512778734e-06,
+      "loss": 0.3866,
+      "step": 39955
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.387604416512119,
+      "learning_rate": 1.7793287775529206e-06,
+      "loss": 0.3664,
+      "step": 39960
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.3282718625370515,
+      "learning_rate": 1.778075249903306e-06,
+      "loss": 0.3921,
+      "step": 39965
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.732838549904333,
+      "learning_rate": 1.7768220684637233e-06,
+      "loss": 0.3841,
+      "step": 39970
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.480958165378972,
+      "learning_rate": 1.7755692333688363e-06,
+      "loss": 0.4094,
+      "step": 39975
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.8594509812252955,
+      "learning_rate": 1.7743167447532656e-06,
+      "loss": 0.3912,
+      "step": 39980
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.201267026690023,
+      "learning_rate": 1.7730646027515947e-06,
+      "loss": 0.3791,
+      "step": 39985
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.904683434394065,
+      "learning_rate": 1.7718128074983759e-06,
+      "loss": 0.357,
+      "step": 39990
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.614484975725764,
+      "learning_rate": 1.7705613591281162e-06,
+      "loss": 0.3913,
+      "step": 39995
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.903514375355792,
+      "learning_rate": 1.769310257775293e-06,
+      "loss": 0.3841,
+      "step": 40000
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.101166798534954,
+      "learning_rate": 1.7680595035743402e-06,
+      "loss": 0.3933,
+      "step": 40005
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.626978306924325,
+      "learning_rate": 1.7668090966596556e-06,
+      "loss": 0.3687,
+      "step": 40010
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.118185491288194,
+      "learning_rate": 1.7655590371656051e-06,
+      "loss": 0.396,
+      "step": 40015
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.51995297695505,
+      "learning_rate": 1.7643093252265087e-06,
+      "loss": 0.3408,
+      "step": 40020
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.585735743445754,
+      "learning_rate": 1.7630599609766575e-06,
+      "loss": 0.3786,
+      "step": 40025
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.597941801697267,
+      "learning_rate": 1.7618109445502969e-06,
+      "loss": 0.3905,
+      "step": 40030
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.291746850834901,
+      "learning_rate": 1.7605622760816416e-06,
+      "loss": 0.3717,
+      "step": 40035
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.183254091595897,
+      "learning_rate": 1.7593139557048688e-06,
+      "loss": 0.3818,
+      "step": 40040
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.698888169204857,
+      "learning_rate": 1.7580659835541115e-06,
+      "loss": 0.3953,
+      "step": 40045
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.08503488443472,
+      "learning_rate": 1.7568183597634736e-06,
+      "loss": 0.3478,
+      "step": 40050
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 6.575952842973665,
+      "learning_rate": 1.755571084467015e-06,
+      "loss": 0.4021,
+      "step": 40055
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.717343849842922,
+      "learning_rate": 1.7543241577987634e-06,
+      "loss": 0.3807,
+      "step": 40060
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.896224140805982,
+      "learning_rate": 1.7530775798927053e-06,
+      "loss": 0.3848,
+      "step": 40065
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 6.206786661431114,
+      "learning_rate": 1.7518313508827895e-06,
+      "loss": 0.3833,
+      "step": 40070
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.457832631442598,
+      "learning_rate": 1.7505854709029319e-06,
+      "loss": 0.3935,
+      "step": 40075
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.77053742441305,
+      "learning_rate": 1.7493399400870042e-06,
+      "loss": 0.3876,
+      "step": 40080
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.315225688429048,
+      "learning_rate": 1.7480947585688486e-06,
+      "loss": 0.3804,
+      "step": 40085
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.007213698493856,
+      "learning_rate": 1.7468499264822624e-06,
+      "loss": 0.3972,
+      "step": 40090
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.255648013712933,
+      "learning_rate": 1.7456054439610076e-06,
+      "loss": 0.3585,
+      "step": 40095
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.616146107390057,
+      "learning_rate": 1.7443613111388119e-06,
+      "loss": 0.3908,
+      "step": 40100
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.7905542605923,
+      "learning_rate": 1.74311752814936e-06,
+      "loss": 0.4015,
+      "step": 40105
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.456087471308075,
+      "learning_rate": 1.7418740951263053e-06,
+      "loss": 0.3468,
+      "step": 40110
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.476592988351036,
+      "learning_rate": 1.740631012203256e-06,
+      "loss": 0.4064,
+      "step": 40115
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.131544441389975,
+      "learning_rate": 1.739388279513791e-06,
+      "loss": 0.3988,
+      "step": 40120
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.603534983768999,
+      "learning_rate": 1.7381458971914444e-06,
+      "loss": 0.3639,
+      "step": 40125
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 8.43381883975848,
+      "learning_rate": 1.7369038653697163e-06,
+      "loss": 0.3781,
+      "step": 40130
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.003059387400197,
+      "learning_rate": 1.7356621841820708e-06,
+      "loss": 0.3895,
+      "step": 40135
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.452544345035951,
+      "learning_rate": 1.734420853761929e-06,
+      "loss": 0.4036,
+      "step": 40140
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.69387442137243,
+      "learning_rate": 1.7331798742426798e-06,
+      "loss": 0.4047,
+      "step": 40145
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.121479283529866,
+      "learning_rate": 1.7319392457576706e-06,
+      "loss": 0.3916,
+      "step": 40150
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.633748456271346,
+      "learning_rate": 1.730698968440211e-06,
+      "loss": 0.3559,
+      "step": 40155
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.2209201655120925,
+      "learning_rate": 1.7294590424235775e-06,
+      "loss": 0.3842,
+      "step": 40160
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.787937888618413,
+      "learning_rate": 1.7282194678410019e-06,
+      "loss": 0.3768,
+      "step": 40165
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.726384762671759,
+      "learning_rate": 1.7269802448256857e-06,
+      "loss": 0.3633,
+      "step": 40170
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.48174191295697,
+      "learning_rate": 1.7257413735107874e-06,
+      "loss": 0.3756,
+      "step": 40175
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.32187970316261,
+      "learning_rate": 1.7245028540294273e-06,
+      "loss": 0.4005,
+      "step": 40180
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.091410517401763,
+      "learning_rate": 1.7232646865146925e-06,
+      "loss": 0.3681,
+      "step": 40185
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 9.237693585398954,
+      "learning_rate": 1.7220268710996273e-06,
+      "loss": 0.402,
+      "step": 40190
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.939114523363182,
+      "learning_rate": 1.7207894079172432e-06,
+      "loss": 0.4014,
+      "step": 40195
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 7.504750750160518,
+      "learning_rate": 1.719552297100508e-06,
+      "loss": 0.4257,
+      "step": 40200
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.6011876625089485,
+      "learning_rate": 1.7183155387823575e-06,
+      "loss": 0.4136,
+      "step": 40205
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.318831549647662,
+      "learning_rate": 1.7170791330956836e-06,
+      "loss": 0.3744,
+      "step": 40210
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.512806244539759,
+      "learning_rate": 1.7158430801733478e-06,
+      "loss": 0.3936,
+      "step": 40215
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.398991443550248,
+      "learning_rate": 1.714607380148165e-06,
+      "loss": 0.3674,
+      "step": 40220
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.834808978448424,
+      "learning_rate": 1.7133720331529192e-06,
+      "loss": 0.4031,
+      "step": 40225
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.234458990469325,
+      "learning_rate": 1.7121370393203545e-06,
+      "loss": 0.3526,
+      "step": 40230
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.230986911149694,
+      "learning_rate": 1.7109023987831758e-06,
+      "loss": 0.4306,
+      "step": 40235
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.260164425173618,
+      "learning_rate": 1.7096681116740489e-06,
+      "loss": 0.3653,
+      "step": 40240
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.09656629929404,
+      "learning_rate": 1.7084341781256058e-06,
+      "loss": 0.3697,
+      "step": 40245
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.9899251099614945,
+      "learning_rate": 1.7072005982704354e-06,
+      "loss": 0.3697,
+      "step": 40250
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.608311515844059,
+      "learning_rate": 1.7059673722410953e-06,
+      "loss": 0.3504,
+      "step": 40255
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.661292840743603,
+      "learning_rate": 1.7047345001700981e-06,
+      "loss": 0.4,
+      "step": 40260
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.65632882069433,
+      "learning_rate": 1.70350198218992e-06,
+      "loss": 0.3672,
+      "step": 40265
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.380331718138243,
+      "learning_rate": 1.702269818433005e-06,
+      "loss": 0.3983,
+      "step": 40270
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.4660712596552745,
+      "learning_rate": 1.7010380090317496e-06,
+      "loss": 0.3802,
+      "step": 40275
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.887172784336666,
+      "learning_rate": 1.6998065541185216e-06,
+      "loss": 0.3907,
+      "step": 40280
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.2312738967975285,
+      "learning_rate": 1.6985754538256427e-06,
+      "loss": 0.4181,
+      "step": 40285
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.17169900427142,
+      "learning_rate": 1.6973447082854027e-06,
+      "loss": 0.363,
+      "step": 40290
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.041953468166664,
+      "learning_rate": 1.6961143176300503e-06,
+      "loss": 0.3733,
+      "step": 40295
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.670575023396616,
+      "learning_rate": 1.6948842819917933e-06,
+      "loss": 0.3909,
+      "step": 40300
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.05039389259463,
+      "learning_rate": 1.6936546015028072e-06,
+      "loss": 0.4076,
+      "step": 40305
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.382956696895827,
+      "learning_rate": 1.6924252762952275e-06,
+      "loss": 0.3801,
+      "step": 40310
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.530599516911307,
+      "learning_rate": 1.6911963065011478e-06,
+      "loss": 0.4104,
+      "step": 40315
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.374977071160649,
+      "learning_rate": 1.6899676922526297e-06,
+      "loss": 0.3685,
+      "step": 40320
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.856868607579105,
+      "learning_rate": 1.6887394336816892e-06,
+      "loss": 0.3841,
+      "step": 40325
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.822420778666742,
+      "learning_rate": 1.6875115309203128e-06,
+      "loss": 0.3677,
+      "step": 40330
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.980018474505903,
+      "learning_rate": 1.6862839841004392e-06,
+      "loss": 0.3813,
+      "step": 40335
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.318868293061312,
+      "learning_rate": 1.6850567933539774e-06,
+      "loss": 0.3865,
+      "step": 40340
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.28277277721484,
+      "learning_rate": 1.6838299588127937e-06,
+      "loss": 0.403,
+      "step": 40345
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.37803981519213,
+      "learning_rate": 1.6826034806087144e-06,
+      "loss": 0.3591,
+      "step": 40350
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.519348203250091,
+      "learning_rate": 1.6813773588735338e-06,
+      "loss": 0.4053,
+      "step": 40355
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.767719750048162,
+      "learning_rate": 1.6801515937390001e-06,
+      "loss": 0.3545,
+      "step": 40360
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.166497253582355,
+      "learning_rate": 1.678926185336831e-06,
+      "loss": 0.3967,
+      "step": 40365
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.3982206883441135,
+      "learning_rate": 1.6777011337986992e-06,
+      "loss": 0.3432,
+      "step": 40370
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.561414973283749,
+      "learning_rate": 1.676476439256244e-06,
+      "loss": 0.3918,
+      "step": 40375
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.428004784875688,
+      "learning_rate": 1.6752521018410633e-06,
+      "loss": 0.3739,
+      "step": 40380
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.195053332195572,
+      "learning_rate": 1.6740281216847165e-06,
+      "loss": 0.365,
+      "step": 40385
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.921958038547423,
+      "learning_rate": 1.6728044989187276e-06,
+      "loss": 0.3986,
+      "step": 40390
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.7015340087411674,
+      "learning_rate": 1.671581233674578e-06,
+      "loss": 0.4132,
+      "step": 40395
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.311419329575143,
+      "learning_rate": 1.6703583260837141e-06,
+      "loss": 0.3884,
+      "step": 40400
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.4625246607491045,
+      "learning_rate": 1.6691357762775445e-06,
+      "loss": 0.3902,
+      "step": 40405
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.81466550448462,
+      "learning_rate": 1.6679135843874344e-06,
+      "loss": 0.3725,
+      "step": 40410
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.390794897561129,
+      "learning_rate": 1.6666917505447171e-06,
+      "loss": 0.3771,
+      "step": 40415
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.0075798180616005,
+      "learning_rate": 1.6654702748806806e-06,
+      "loss": 0.4033,
+      "step": 40420
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.067239046390207,
+      "learning_rate": 1.6642491575265807e-06,
+      "loss": 0.3808,
+      "step": 40425
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.249901217365233,
+      "learning_rate": 1.6630283986136308e-06,
+      "loss": 0.3533,
+      "step": 40430
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.202734373070795,
+      "learning_rate": 1.6618079982730052e-06,
+      "loss": 0.3619,
+      "step": 40435
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.502412149561305,
+      "learning_rate": 1.660587956635844e-06,
+      "loss": 0.3837,
+      "step": 40440
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.431563390945679,
+      "learning_rate": 1.6593682738332434e-06,
+      "loss": 0.4127,
+      "step": 40445
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.680340853441706,
+      "learning_rate": 1.658148949996267e-06,
+      "loss": 0.3825,
+      "step": 40450
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.629858676737288,
+      "learning_rate": 1.656929985255933e-06,
+      "loss": 0.3564,
+      "step": 40455
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.03217003073883,
+      "learning_rate": 1.6557113797432273e-06,
+      "loss": 0.3768,
+      "step": 40460
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.5766269217099955,
+      "learning_rate": 1.6544931335890934e-06,
+      "loss": 0.3908,
+      "step": 40465
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.338633797693593,
+      "learning_rate": 1.6532752469244362e-06,
+      "loss": 0.361,
+      "step": 40470
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.592202192053285,
+      "learning_rate": 1.6520577198801247e-06,
+      "loss": 0.3881,
+      "step": 40475
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.821523306890911,
+      "learning_rate": 1.6508405525869859e-06,
+      "loss": 0.3807,
+      "step": 40480
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.025014604605802,
+      "learning_rate": 1.6496237451758118e-06,
+      "loss": 0.3658,
+      "step": 40485
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.756637352965077,
+      "learning_rate": 1.6484072977773512e-06,
+      "loss": 0.3922,
+      "step": 40490
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.713603424803598,
+      "learning_rate": 1.6471912105223182e-06,
+      "loss": 0.3752,
+      "step": 40495
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.496059990610466,
+      "learning_rate": 1.6459754835413882e-06,
+      "loss": 0.3894,
+      "step": 40500
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.663798283636861,
+      "learning_rate": 1.6447601169651933e-06,
+      "loss": 0.3922,
+      "step": 40505
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.76114830557741,
+      "learning_rate": 1.6435451109243334e-06,
+      "loss": 0.3568,
+      "step": 40510
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.470802302673014,
+      "learning_rate": 1.6423304655493627e-06,
+      "loss": 0.3777,
+      "step": 40515
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.110184425128485,
+      "learning_rate": 1.6411161809708038e-06,
+      "loss": 0.3803,
+      "step": 40520
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.355958891754689,
+      "learning_rate": 1.6399022573191347e-06,
+      "loss": 0.3721,
+      "step": 40525
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.7045848819087475,
+      "learning_rate": 1.6386886947247959e-06,
+      "loss": 0.3805,
+      "step": 40530
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.500835312947065,
+      "learning_rate": 1.6374754933181936e-06,
+      "loss": 0.3665,
+      "step": 40535
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.791410238145022,
+      "learning_rate": 1.6362626532296872e-06,
+      "loss": 0.3907,
+      "step": 40540
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.969446936783075,
+      "learning_rate": 1.6350501745896064e-06,
+      "loss": 0.406,
+      "step": 40545
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.809957590921652,
+      "learning_rate": 1.6338380575282341e-06,
+      "loss": 0.3676,
+      "step": 40550
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.84334907773835,
+      "learning_rate": 1.6326263021758177e-06,
+      "loss": 0.4029,
+      "step": 40555
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.860346619578493,
+      "learning_rate": 1.6314149086625674e-06,
+      "loss": 0.3833,
+      "step": 40560
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.672873076175617,
+      "learning_rate": 1.6302038771186507e-06,
+      "loss": 0.3573,
+      "step": 40565
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.565323847713768,
+      "learning_rate": 1.628993207674201e-06,
+      "loss": 0.3729,
+      "step": 40570
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.871387321383843,
+      "learning_rate": 1.6277829004593065e-06,
+      "loss": 0.3661,
+      "step": 40575
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.921764024629117,
+      "learning_rate": 1.6265729556040243e-06,
+      "loss": 0.3792,
+      "step": 40580
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.628195277449639,
+      "learning_rate": 1.625363373238364e-06,
+      "loss": 0.4596,
+      "step": 40585
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.241764352139462,
+      "learning_rate": 1.6241541534923028e-06,
+      "loss": 0.3558,
+      "step": 40590
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.467257929690298,
+      "learning_rate": 1.622945296495778e-06,
+      "loss": 0.386,
+      "step": 40595
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.809227856666526,
+      "learning_rate": 1.6217368023786838e-06,
+      "loss": 0.4057,
+      "step": 40600
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.708594995182603,
+      "learning_rate": 1.6205286712708813e-06,
+      "loss": 0.3903,
+      "step": 40605
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.8796229505159925,
+      "learning_rate": 1.619320903302188e-06,
+      "loss": 0.3808,
+      "step": 40610
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.678629484760158,
+      "learning_rate": 1.6181134986023817e-06,
+      "loss": 0.3798,
+      "step": 40615
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.720973732849028,
+      "learning_rate": 1.6169064573012072e-06,
+      "loss": 0.3715,
+      "step": 40620
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.814377236137956,
+      "learning_rate": 1.6156997795283624e-06,
+      "loss": 0.3783,
+      "step": 40625
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.417103814417879,
+      "learning_rate": 1.6144934654135148e-06,
+      "loss": 0.3753,
+      "step": 40630
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.0065188606912745,
+      "learning_rate": 1.6132875150862853e-06,
+      "loss": 0.4103,
+      "step": 40635
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.9714511946192355,
+      "learning_rate": 1.6120819286762574e-06,
+      "loss": 0.4156,
+      "step": 40640
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.317206691917036,
+      "learning_rate": 1.6108767063129798e-06,
+      "loss": 0.3897,
+      "step": 40645
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.861857406802999,
+      "learning_rate": 1.609671848125956e-06,
+      "loss": 0.3758,
+      "step": 40650
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.81094751787445,
+      "learning_rate": 1.6084673542446567e-06,
+      "loss": 0.3742,
+      "step": 40655
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.436022098390143,
+      "learning_rate": 1.6072632247985075e-06,
+      "loss": 0.3881,
+      "step": 40660
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.927519085257424,
+      "learning_rate": 1.6060594599168972e-06,
+      "loss": 0.4152,
+      "step": 40665
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.802895990350653,
+      "learning_rate": 1.6048560597291758e-06,
+      "loss": 0.3883,
+      "step": 40670
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.997356432757642,
+      "learning_rate": 1.6036530243646564e-06,
+      "loss": 0.3541,
+      "step": 40675
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.178252585774607,
+      "learning_rate": 1.6024503539526077e-06,
+      "loss": 0.3643,
+      "step": 40680
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.81546052738333,
+      "learning_rate": 1.6012480486222626e-06,
+      "loss": 0.4504,
+      "step": 40685
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.6443873418273425,
+      "learning_rate": 1.6000461085028162e-06,
+      "loss": 0.3774,
+      "step": 40690
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.20862116748695,
+      "learning_rate": 1.5988445337234205e-06,
+      "loss": 0.3922,
+      "step": 40695
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.090724390447631,
+      "learning_rate": 1.597643324413189e-06,
+      "loss": 0.3798,
+      "step": 40700
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.972254039083642,
+      "learning_rate": 1.596442480701199e-06,
+      "loss": 0.3624,
+      "step": 40705
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.308436788062177,
+      "learning_rate": 1.5952420027164839e-06,
+      "loss": 0.3603,
+      "step": 40710
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.839966926446589,
+      "learning_rate": 1.594041890588044e-06,
+      "loss": 0.4028,
+      "step": 40715
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.426533284941413,
+      "learning_rate": 1.5928421444448344e-06,
+      "loss": 0.3984,
+      "step": 40720
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.334377695764412,
+      "learning_rate": 1.5916427644157717e-06,
+      "loss": 0.3652,
+      "step": 40725
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.13341820801301,
+      "learning_rate": 1.5904437506297376e-06,
+      "loss": 0.3659,
+      "step": 40730
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.811878173929772,
+      "learning_rate": 1.5892451032155686e-06,
+      "loss": 0.3971,
+      "step": 40735
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.839613876293462,
+      "learning_rate": 1.588046822302068e-06,
+      "loss": 0.3931,
+      "step": 40740
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.641267564857314,
+      "learning_rate": 1.586848908017994e-06,
+      "loss": 0.3594,
+      "step": 40745
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.265078255817545,
+      "learning_rate": 1.5856513604920665e-06,
+      "loss": 0.3887,
+      "step": 40750
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.924287364070042,
+      "learning_rate": 1.5844541798529706e-06,
+      "loss": 0.3897,
+      "step": 40755
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.418766995878838,
+      "learning_rate": 1.5832573662293455e-06,
+      "loss": 0.3786,
+      "step": 40760
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.247511210258513,
+      "learning_rate": 1.582060919749796e-06,
+      "loss": 0.3551,
+      "step": 40765
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.783076149553892,
+      "learning_rate": 1.5808648405428866e-06,
+      "loss": 0.3794,
+      "step": 40770
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 6.568669354521603,
+      "learning_rate": 1.579669128737138e-06,
+      "loss": 0.3932,
+      "step": 40775
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.404243627253454,
+      "learning_rate": 1.5784737844610388e-06,
+      "loss": 0.3793,
+      "step": 40780
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.471633109198061,
+      "learning_rate": 1.5772788078430295e-06,
+      "loss": 0.4197,
+      "step": 40785
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.17267264392894,
+      "learning_rate": 1.57608419901152e-06,
+      "loss": 0.379,
+      "step": 40790
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.8909752234693284,
+      "learning_rate": 1.5748899580948723e-06,
+      "loss": 0.3922,
+      "step": 40795
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.530214688417191,
+      "learning_rate": 1.5736960852214167e-06,
+      "loss": 0.3865,
+      "step": 40800
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.536300365596513,
+      "learning_rate": 1.5725025805194377e-06,
+      "loss": 0.3733,
+      "step": 40805
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.738979810237718,
+      "learning_rate": 1.5713094441171818e-06,
+      "loss": 0.4215,
+      "step": 40810
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.6112891500415,
+      "learning_rate": 1.5701166761428599e-06,
+      "loss": 0.3837,
+      "step": 40815
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.267322599042935,
+      "learning_rate": 1.5689242767246366e-06,
+      "loss": 0.3814,
+      "step": 40820
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.715174760088558,
+      "learning_rate": 1.5677322459906436e-06,
+      "loss": 0.3578,
+      "step": 40825
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.234044556174818,
+      "learning_rate": 1.566540584068969e-06,
+      "loss": 0.3827,
+      "step": 40830
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.479917105729541,
+      "learning_rate": 1.5653492910876595e-06,
+      "loss": 0.3746,
+      "step": 40835
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.679943290035979,
+      "learning_rate": 1.564158367174729e-06,
+      "loss": 0.3906,
+      "step": 40840
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.136481668956996,
+      "learning_rate": 1.5629678124581438e-06,
+      "loss": 0.3901,
+      "step": 40845
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.376873173911454,
+      "learning_rate": 1.5617776270658374e-06,
+      "loss": 0.3857,
+      "step": 40850
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.394554983598227,
+      "learning_rate": 1.5605878111256972e-06,
+      "loss": 0.3566,
+      "step": 40855
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.481350604069281,
+      "learning_rate": 1.5593983647655764e-06,
+      "loss": 0.35,
+      "step": 40860
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.353970597670246,
+      "learning_rate": 1.558209288113287e-06,
+      "loss": 0.3553,
+      "step": 40865
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.064622350006163,
+      "learning_rate": 1.5570205812965982e-06,
+      "loss": 0.3855,
+      "step": 40870
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.966188123128139,
+      "learning_rate": 1.5558322444432444e-06,
+      "loss": 0.3888,
+      "step": 40875
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.345661301466502,
+      "learning_rate": 1.554644277680915e-06,
+      "loss": 0.3775,
+      "step": 40880
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.868666959598315,
+      "learning_rate": 1.5534566811372648e-06,
+      "loss": 0.3879,
+      "step": 40885
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 6.154470452715487,
+      "learning_rate": 1.5522694549399047e-06,
+      "loss": 0.3642,
+      "step": 40890
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.184623228605581,
+      "learning_rate": 1.5510825992164063e-06,
+      "loss": 0.3871,
+      "step": 40895
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.153310343958727,
+      "learning_rate": 1.5498961140943053e-06,
+      "loss": 0.3916,
+      "step": 40900
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.562091939341356,
+      "learning_rate": 1.5487099997010918e-06,
+      "loss": 0.3909,
+      "step": 40905
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.3833323057626945,
+      "learning_rate": 1.5475242561642218e-06,
+      "loss": 0.3969,
+      "step": 40910
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.311173049896008,
+      "learning_rate": 1.5463388836111083e-06,
+      "loss": 0.3805,
+      "step": 40915
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.621440572540954,
+      "learning_rate": 1.5451538821691214e-06,
+      "loss": 0.3626,
+      "step": 40920
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.674777468191818,
+      "learning_rate": 1.543969251965599e-06,
+      "loss": 0.4025,
+      "step": 40925
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.557421883043057,
+      "learning_rate": 1.542784993127832e-06,
+      "loss": 0.3675,
+      "step": 40930
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.129566341297116,
+      "learning_rate": 1.5416011057830765e-06,
+      "loss": 0.3873,
+      "step": 40935
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.563078388464786,
+      "learning_rate": 1.540417590058544e-06,
+      "loss": 0.3591,
+      "step": 40940
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.244213722893947,
+      "learning_rate": 1.5392344460814113e-06,
+      "loss": 0.4273,
+      "step": 40945
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.094560947769992,
+      "learning_rate": 1.5380516739788087e-06,
+      "loss": 0.3682,
+      "step": 40950
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.505086647240551,
+      "learning_rate": 1.5368692738778335e-06,
+      "loss": 0.387,
+      "step": 40955
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.498789918590851,
+      "learning_rate": 1.5356872459055394e-06,
+      "loss": 0.3465,
+      "step": 40960
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.820769572094417,
+      "learning_rate": 1.5345055901889388e-06,
+      "loss": 0.394,
+      "step": 40965
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.791439790016614,
+      "learning_rate": 1.5333243068550085e-06,
+      "loss": 0.3751,
+      "step": 40970
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.55250492826794,
+      "learning_rate": 1.5321433960306808e-06,
+      "loss": 0.3693,
+      "step": 40975
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.270833146308453,
+      "learning_rate": 1.5309628578428487e-06,
+      "loss": 0.3558,
+      "step": 40980
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.659859310376631,
+      "learning_rate": 1.529782692418369e-06,
+      "loss": 0.3842,
+      "step": 40985
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.083723396185362,
+      "learning_rate": 1.528602899884053e-06,
+      "loss": 0.3744,
+      "step": 40990
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 8.007805627671026,
+      "learning_rate": 1.527423480366677e-06,
+      "loss": 0.3797,
+      "step": 40995
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.629745728060397,
+      "learning_rate": 1.526244433992972e-06,
+      "loss": 0.3707,
+      "step": 41000
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.29122520084207,
+      "learning_rate": 1.5250657608896359e-06,
+      "loss": 0.369,
+      "step": 41005
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.0311020264316575,
+      "learning_rate": 1.5238874611833188e-06,
+      "loss": 0.3878,
+      "step": 41010
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.058390944753647,
+      "learning_rate": 1.5227095350006344e-06,
+      "loss": 0.3781,
+      "step": 41015
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.338325784886404,
+      "learning_rate": 1.521531982468158e-06,
+      "loss": 0.3644,
+      "step": 41020
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.424131997517405,
+      "learning_rate": 1.5203548037124199e-06,
+      "loss": 0.3848,
+      "step": 41025
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.815794483256725,
+      "learning_rate": 1.5191779988599176e-06,
+      "loss": 0.3626,
+      "step": 41030
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.924553387563179,
+      "learning_rate": 1.518001568037099e-06,
+      "loss": 0.3728,
+      "step": 41035
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.723624819856848,
+      "learning_rate": 1.5168255113703805e-06,
+      "loss": 0.3686,
+      "step": 41040
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.432214576467207,
+      "learning_rate": 1.5156498289861321e-06,
+      "loss": 0.3797,
+      "step": 41045
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.246535250718635,
+      "learning_rate": 1.514474521010687e-06,
+      "loss": 0.3598,
+      "step": 41050
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.387679548552412,
+      "learning_rate": 1.5132995875703393e-06,
+      "loss": 0.3706,
+      "step": 41055
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.845500059135731,
+      "learning_rate": 1.5121250287913387e-06,
+      "loss": 0.3693,
+      "step": 41060
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 9.057007563879125,
+      "learning_rate": 1.5109508447998956e-06,
+      "loss": 0.3936,
+      "step": 41065
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.706049512572617,
+      "learning_rate": 1.5097770357221836e-06,
+      "loss": 0.3863,
+      "step": 41070
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.334496043402186,
+      "learning_rate": 1.5086036016843313e-06,
+      "loss": 0.4276,
+      "step": 41075
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.243426822776101,
+      "learning_rate": 1.5074305428124325e-06,
+      "loss": 0.3561,
+      "step": 41080
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 11.798026476982267,
+      "learning_rate": 1.5062578592325338e-06,
+      "loss": 0.3621,
+      "step": 41085
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.113419671835875,
+      "learning_rate": 1.5050855510706491e-06,
+      "loss": 0.3567,
+      "step": 41090
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.536698788021902,
+      "learning_rate": 1.5039136184527465e-06,
+      "loss": 0.3819,
+      "step": 41095
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.159698410576232,
+      "learning_rate": 1.5027420615047527e-06,
+      "loss": 0.3852,
+      "step": 41100
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.483275684812258,
+      "learning_rate": 1.5015708803525607e-06,
+      "loss": 0.352,
+      "step": 41105
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.409720247496013,
+      "learning_rate": 1.500400075122016e-06,
+      "loss": 0.3293,
+      "step": 41110
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.864562323156177,
+      "learning_rate": 1.49922964593893e-06,
+      "loss": 0.3775,
+      "step": 41115
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.582645043870312,
+      "learning_rate": 1.4980595929290682e-06,
+      "loss": 0.3388,
+      "step": 41120
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.971510177781197,
+      "learning_rate": 1.4968899162181566e-06,
+      "loss": 0.3746,
+      "step": 41125
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.342454429663956,
+      "learning_rate": 1.4957206159318843e-06,
+      "loss": 0.4003,
+      "step": 41130
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.386339371066969,
+      "learning_rate": 1.4945516921958985e-06,
+      "loss": 0.381,
+      "step": 41135
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 8.40558156905113,
+      "learning_rate": 1.4933831451358032e-06,
+      "loss": 0.3786,
+      "step": 41140
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.6527546428144415,
+      "learning_rate": 1.492214974877166e-06,
+      "loss": 0.3573,
+      "step": 41145
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.008328683240061,
+      "learning_rate": 1.491047181545509e-06,
+      "loss": 0.3496,
+      "step": 41150
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.137348978051636,
+      "learning_rate": 1.4898797652663206e-06,
+      "loss": 0.3832,
+      "step": 41155
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.402391843206946,
+      "learning_rate": 1.4887127261650408e-06,
+      "loss": 0.3586,
+      "step": 41160
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.877030220146141,
+      "learning_rate": 1.487546064367077e-06,
+      "loss": 0.3727,
+      "step": 41165
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.451313273696703,
+      "learning_rate": 1.4863797799977886e-06,
+      "loss": 0.4203,
+      "step": 41170
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.544209830360189,
+      "learning_rate": 1.485213873182501e-06,
+      "loss": 0.3709,
+      "step": 41175
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 8.802021736841098,
+      "learning_rate": 1.4840483440464948e-06,
+      "loss": 0.4226,
+      "step": 41180
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.014701088904308,
+      "learning_rate": 1.4828831927150096e-06,
+      "loss": 0.4191,
+      "step": 41185
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.589328705932698,
+      "learning_rate": 1.4817184193132488e-06,
+      "loss": 0.3706,
+      "step": 41190
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.8671270281107475,
+      "learning_rate": 1.4805540239663696e-06,
+      "loss": 0.3833,
+      "step": 41195
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.750281417746645,
+      "learning_rate": 1.4793900067994948e-06,
+      "loss": 0.3768,
+      "step": 41200
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.852647510417684,
+      "learning_rate": 1.4782263679377018e-06,
+      "loss": 0.3667,
+      "step": 41205
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.639245774739903,
+      "learning_rate": 1.477063107506026e-06,
+      "loss": 0.3696,
+      "step": 41210
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.678349940400766,
+      "learning_rate": 1.4759002256294691e-06,
+      "loss": 0.3928,
+      "step": 41215
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.682189264635429,
+      "learning_rate": 1.4747377224329845e-06,
+      "loss": 0.3745,
+      "step": 41220
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.026459060219476,
+      "learning_rate": 1.4735755980414902e-06,
+      "loss": 0.4051,
+      "step": 41225
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.962382841294382,
+      "learning_rate": 1.472413852579862e-06,
+      "loss": 0.361,
+      "step": 41230
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.432993333532744,
+      "learning_rate": 1.4712524861729328e-06,
+      "loss": 0.38,
+      "step": 41235
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.84267554570037,
+      "learning_rate": 1.470091498945499e-06,
+      "loss": 0.3699,
+      "step": 41240
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.795298784613299,
+      "learning_rate": 1.4689308910223104e-06,
+      "loss": 0.3812,
+      "step": 41245
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 7.465642466091052,
+      "learning_rate": 1.467770662528083e-06,
+      "loss": 0.3631,
+      "step": 41250
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.545525192552546,
+      "learning_rate": 1.4666108135874851e-06,
+      "loss": 0.3592,
+      "step": 41255
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.242304682894049,
+      "learning_rate": 1.465451344325151e-06,
+      "loss": 0.3602,
+      "step": 41260
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.713111656574008,
+      "learning_rate": 1.4642922548656695e-06,
+      "loss": 0.3545,
+      "step": 41265
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.309981303294159,
+      "learning_rate": 1.4631335453335877e-06,
+      "loss": 0.3438,
+      "step": 41270
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.687516209370226,
+      "learning_rate": 1.4619752158534174e-06,
+      "loss": 0.3656,
+      "step": 41275
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.786537559134805,
+      "learning_rate": 1.460817266549623e-06,
+      "loss": 0.3797,
+      "step": 41280
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.569979725037328,
+      "learning_rate": 1.4596596975466342e-06,
+      "loss": 0.374,
+      "step": 41285
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.478371640377967,
+      "learning_rate": 1.4585025089688366e-06,
+      "loss": 0.3376,
+      "step": 41290
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 7.089197001967615,
+      "learning_rate": 1.4573457009405716e-06,
+      "loss": 0.4028,
+      "step": 41295
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.8026464978663075,
+      "learning_rate": 1.4561892735861482e-06,
+      "loss": 0.3503,
+      "step": 41300
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 6.550690911035344,
+      "learning_rate": 1.455033227029825e-06,
+      "loss": 0.3742,
+      "step": 41305
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.776837430310707,
+      "learning_rate": 1.4538775613958289e-06,
+      "loss": 0.3591,
+      "step": 41310
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 6.620114119199254,
+      "learning_rate": 1.4527222768083377e-06,
+      "loss": 0.3714,
+      "step": 41315
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.524204452624631,
+      "learning_rate": 1.4515673733914926e-06,
+      "loss": 0.3547,
+      "step": 41320
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.215791621823948,
+      "learning_rate": 1.4504128512693955e-06,
+      "loss": 0.381,
+      "step": 41325
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.630587571043699,
+      "learning_rate": 1.4492587105661015e-06,
+      "loss": 0.3636,
+      "step": 41330
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.358019452151076,
+      "learning_rate": 1.4481049514056316e-06,
+      "loss": 0.3746,
+      "step": 41335
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.4237005159013565,
+      "learning_rate": 1.4469515739119583e-06,
+      "loss": 0.3608,
+      "step": 41340
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.7422580007666,
+      "learning_rate": 1.4457985782090205e-06,
+      "loss": 0.3603,
+      "step": 41345
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.715583950427968,
+      "learning_rate": 1.4446459644207124e-06,
+      "loss": 0.3657,
+      "step": 41350
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.698222572958804,
+      "learning_rate": 1.4434937326708837e-06,
+      "loss": 0.3499,
+      "step": 41355
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.201807539958746,
+      "learning_rate": 1.4423418830833518e-06,
+      "loss": 0.3744,
+      "step": 41360
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 6.3381142221138305,
+      "learning_rate": 1.4411904157818835e-06,
+      "loss": 0.4119,
+      "step": 41365
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.969531506545102,
+      "learning_rate": 1.4400393308902133e-06,
+      "loss": 0.391,
+      "step": 41370
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.737253772410425,
+      "learning_rate": 1.4388886285320281e-06,
+      "loss": 0.3734,
+      "step": 41375
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.717335145515903,
+      "learning_rate": 1.437738308830975e-06,
+      "loss": 0.3761,
+      "step": 41380
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.156294678426971,
+      "learning_rate": 1.4365883719106633e-06,
+      "loss": 0.3746,
+      "step": 41385
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.3731152947031,
+      "learning_rate": 1.435438817894656e-06,
+      "loss": 0.3878,
+      "step": 41390
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 6.342736889782638,
+      "learning_rate": 1.434289646906481e-06,
+      "loss": 0.3422,
+      "step": 41395
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.5571586886572755,
+      "learning_rate": 1.4331408590696182e-06,
+      "loss": 0.3709,
+      "step": 41400
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 7.298640611879569,
+      "learning_rate": 1.4319924545075137e-06,
+      "loss": 0.3907,
+      "step": 41405
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.493926763384604,
+      "learning_rate": 1.430844433343565e-06,
+      "loss": 0.3889,
+      "step": 41410
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.963244884218475,
+      "learning_rate": 1.4296967957011342e-06,
+      "loss": 0.3839,
+      "step": 41415
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.535090327382197,
+      "learning_rate": 1.4285495417035411e-06,
+      "loss": 0.3572,
+      "step": 41420
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.426326852781701,
+      "learning_rate": 1.4274026714740601e-06,
+      "loss": 0.3797,
+      "step": 41425
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.7193713564748005,
+      "learning_rate": 1.4262561851359308e-06,
+      "loss": 0.3802,
+      "step": 41430
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.6806080297893615,
+      "learning_rate": 1.425110082812346e-06,
+      "loss": 0.3822,
+      "step": 41435
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.334496719452402,
+      "learning_rate": 1.423964364626459e-06,
+      "loss": 0.3882,
+      "step": 41440
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.7277683705123525,
+      "learning_rate": 1.4228190307013845e-06,
+      "loss": 0.388,
+      "step": 41445
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.897797861548727,
+      "learning_rate": 1.4216740811601915e-06,
+      "loss": 0.3658,
+      "step": 41450
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.040401471109189,
+      "learning_rate": 1.420529516125912e-06,
+      "loss": 0.3491,
+      "step": 41455
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.665000912640641,
+      "learning_rate": 1.4193853357215331e-06,
+      "loss": 0.3443,
+      "step": 41460
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 7.097019856409633,
+      "learning_rate": 1.418241540070001e-06,
+      "loss": 0.3956,
+      "step": 41465
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.362517713058213,
+      "learning_rate": 1.4170981292942248e-06,
+      "loss": 0.3568,
+      "step": 41470
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.506807738596816,
+      "learning_rate": 1.4159551035170655e-06,
+      "loss": 0.3554,
+      "step": 41475
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.9957157494220095,
+      "learning_rate": 1.4148124628613492e-06,
+      "loss": 0.4186,
+      "step": 41480
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.587804492736562,
+      "learning_rate": 1.413670207449855e-06,
+      "loss": 0.3728,
+      "step": 41485
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.705940252804655,
+      "learning_rate": 1.412528337405326e-06,
+      "loss": 0.3681,
+      "step": 41490
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.83490997795641,
+      "learning_rate": 1.4113868528504587e-06,
+      "loss": 0.3701,
+      "step": 41495
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.60418297060718,
+      "learning_rate": 1.410245753907913e-06,
+      "loss": 0.3548,
+      "step": 41500
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.825715126402115,
+      "learning_rate": 1.4091050407003025e-06,
+      "loss": 0.339,
+      "step": 41505
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.057393212416262,
+      "learning_rate": 1.4079647133502033e-06,
+      "loss": 0.3736,
+      "step": 41510
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.4553613565875505,
+      "learning_rate": 1.4068247719801497e-06,
+      "loss": 0.3824,
+      "step": 41515
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.533306894948522,
+      "learning_rate": 1.4056852167126328e-06,
+      "loss": 0.351,
+      "step": 41520
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.715541749177123,
+      "learning_rate": 1.4045460476701006e-06,
+      "loss": 0.3851,
+      "step": 41525
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.603344085842678,
+      "learning_rate": 1.4034072649749646e-06,
+      "loss": 0.3716,
+      "step": 41530
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.403149084826972,
+      "learning_rate": 1.4022688687495901e-06,
+      "loss": 0.3641,
+      "step": 41535
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.895832540303241,
+      "learning_rate": 1.4011308591163047e-06,
+      "loss": 0.4189,
+      "step": 41540
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.169701701755714,
+      "learning_rate": 1.3999932361973917e-06,
+      "loss": 0.3623,
+      "step": 41545
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.516106282935076,
+      "learning_rate": 1.398856000115092e-06,
+      "loss": 0.3828,
+      "step": 41550
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.4977017376881125,
+      "learning_rate": 1.3977191509916094e-06,
+      "loss": 0.3622,
+      "step": 41555
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.846541545398783,
+      "learning_rate": 1.3965826889491008e-06,
+      "loss": 0.3712,
+      "step": 41560
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.629694430996607,
+      "learning_rate": 1.3954466141096867e-06,
+      "loss": 0.3738,
+      "step": 41565
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.315466466695631,
+      "learning_rate": 1.3943109265954408e-06,
+      "loss": 0.359,
+      "step": 41570
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.559927034532552,
+      "learning_rate": 1.3931756265284003e-06,
+      "loss": 0.3858,
+      "step": 41575
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.773281738613931,
+      "learning_rate": 1.392040714030557e-06,
+      "loss": 0.3779,
+      "step": 41580
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.5318569459830655,
+      "learning_rate": 1.3909061892238602e-06,
+      "loss": 0.3428,
+      "step": 41585
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.394046415543049,
+      "learning_rate": 1.3897720522302215e-06,
+      "loss": 0.3631,
+      "step": 41590
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.794106531694925,
+      "learning_rate": 1.3886383031715111e-06,
+      "loss": 0.381,
+      "step": 41595
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.5524649568518125,
+      "learning_rate": 1.3875049421695514e-06,
+      "loss": 0.3627,
+      "step": 41600
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.6490462069269185,
+      "learning_rate": 1.3863719693461303e-06,
+      "loss": 0.3322,
+      "step": 41605
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.8040181101337005,
+      "learning_rate": 1.3852393848229872e-06,
+      "loss": 0.4208,
+      "step": 41610
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.121062145125661,
+      "learning_rate": 1.3841071887218278e-06,
+      "loss": 0.3813,
+      "step": 41615
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.4776921560623855,
+      "learning_rate": 1.3829753811643071e-06,
+      "loss": 0.3607,
+      "step": 41620
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.479672731586354,
+      "learning_rate": 1.3818439622720465e-06,
+      "loss": 0.3466,
+      "step": 41625
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.910570954227758,
+      "learning_rate": 1.3807129321666201e-06,
+      "loss": 0.3706,
+      "step": 41630
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.818472317059777,
+      "learning_rate": 1.3795822909695605e-06,
+      "loss": 0.3632,
+      "step": 41635
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.126721980157568,
+      "learning_rate": 1.3784520388023641e-06,
+      "loss": 0.3274,
+      "step": 41640
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.6296388314821355,
+      "learning_rate": 1.3773221757864775e-06,
+      "loss": 0.3488,
+      "step": 41645
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.765091515487751,
+      "learning_rate": 1.3761927020433124e-06,
+      "loss": 0.3731,
+      "step": 41650
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.604923886368388,
+      "learning_rate": 1.3750636176942333e-06,
+      "loss": 0.3734,
+      "step": 41655
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.116447850524022,
+      "learning_rate": 1.373934922860568e-06,
+      "loss": 0.346,
+      "step": 41660
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.244332453409364,
+      "learning_rate": 1.372806617663598e-06,
+      "loss": 0.3485,
+      "step": 41665
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.352084726186808,
+      "learning_rate": 1.3716787022245636e-06,
+      "loss": 0.3894,
+      "step": 41670
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.9576958499730415,
+      "learning_rate": 1.3705511766646674e-06,
+      "loss": 0.3756,
+      "step": 41675
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.806080884421692,
+      "learning_rate": 1.3694240411050635e-06,
+      "loss": 0.3728,
+      "step": 41680
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.464414698598367,
+      "learning_rate": 1.3682972956668694e-06,
+      "loss": 0.3868,
+      "step": 41685
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.753367671992034,
+      "learning_rate": 1.3671709404711602e-06,
+      "loss": 0.3959,
+      "step": 41690
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.959482104429004,
+      "learning_rate": 1.3660449756389654e-06,
+      "loss": 0.3848,
+      "step": 41695
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.327636594853386,
+      "learning_rate": 1.364919401291277e-06,
+      "loss": 0.3926,
+      "step": 41700
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.787583437400583,
+      "learning_rate": 1.3637942175490398e-06,
+      "loss": 0.3739,
+      "step": 41705
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.607871008060517,
+      "learning_rate": 1.362669424533164e-06,
+      "loss": 0.3586,
+      "step": 41710
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.217382721853331,
+      "learning_rate": 1.361545022364511e-06,
+      "loss": 0.3853,
+      "step": 41715
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.999931616537445,
+      "learning_rate": 1.360421011163901e-06,
+      "loss": 0.3564,
+      "step": 41720
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 6.009381263935578,
+      "learning_rate": 1.3592973910521184e-06,
+      "loss": 0.3762,
+      "step": 41725
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.060872407200973,
+      "learning_rate": 1.3581741621498967e-06,
+      "loss": 0.3938,
+      "step": 41730
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 7.482587453727809,
+      "learning_rate": 1.357051324577936e-06,
+      "loss": 0.3568,
+      "step": 41735
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.901812743155264,
+      "learning_rate": 1.3559288784568857e-06,
+      "loss": 0.3674,
+      "step": 41740
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.506271378926924,
+      "learning_rate": 1.354806823907362e-06,
+      "loss": 0.3599,
+      "step": 41745
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.988373502308444,
+      "learning_rate": 1.3536851610499325e-06,
+      "loss": 0.3564,
+      "step": 41750
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.160118358919035,
+      "learning_rate": 1.352563890005123e-06,
+      "loss": 0.3844,
+      "step": 41755
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.456918901709329,
+      "learning_rate": 1.3514430108934229e-06,
+      "loss": 0.3634,
+      "step": 41760
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.720142489661282,
+      "learning_rate": 1.3503225238352718e-06,
+      "loss": 0.3621,
+      "step": 41765
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.6730796976704845,
+      "learning_rate": 1.349202428951074e-06,
+      "loss": 0.3849,
+      "step": 41770
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.708039666936624,
+      "learning_rate": 1.3480827263611862e-06,
+      "loss": 0.3784,
+      "step": 41775
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.5775020472229,
+      "learning_rate": 1.346963416185927e-06,
+      "loss": 0.3731,
+      "step": 41780
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.427022226127237,
+      "learning_rate": 1.3458444985455715e-06,
+      "loss": 0.3848,
+      "step": 41785
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.0727974789767805,
+      "learning_rate": 1.3447259735603507e-06,
+      "loss": 0.381,
+      "step": 41790
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.394027541147416,
+      "learning_rate": 1.3436078413504572e-06,
+      "loss": 0.3441,
+      "step": 41795
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.810623074537778,
+      "learning_rate": 1.3424901020360382e-06,
+      "loss": 0.3603,
+      "step": 41800
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.560463698130751,
+      "learning_rate": 1.3413727557371981e-06,
+      "loss": 0.3579,
+      "step": 41805
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.778712992367557,
+      "learning_rate": 1.3402558025740036e-06,
+      "loss": 0.3502,
+      "step": 41810
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.442525354266272,
+      "learning_rate": 1.3391392426664735e-06,
+      "loss": 0.3906,
+      "step": 41815
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.152634030087905,
+      "learning_rate": 1.338023076134589e-06,
+      "loss": 0.3877,
+      "step": 41820
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.66621143335817,
+      "learning_rate": 1.336907303098286e-06,
+      "loss": 0.363,
+      "step": 41825
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.210877694280694,
+      "learning_rate": 1.3357919236774608e-06,
+      "loss": 0.3703,
+      "step": 41830
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.542877284587468,
+      "learning_rate": 1.334676937991965e-06,
+      "loss": 0.3564,
+      "step": 41835
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.586788209250504,
+      "learning_rate": 1.3335623461616071e-06,
+      "loss": 0.3687,
+      "step": 41840
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.78880284551674,
+      "learning_rate": 1.3324481483061574e-06,
+      "loss": 0.3739,
+      "step": 41845
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.514647892992655,
+      "learning_rate": 1.3313343445453392e-06,
+      "loss": 0.3399,
+      "step": 41850
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.693777487639411,
+      "learning_rate": 1.330220934998838e-06,
+      "loss": 0.3743,
+      "step": 41855
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.255976843985964,
+      "learning_rate": 1.3291079197862915e-06,
+      "loss": 0.3721,
+      "step": 41860
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.006070376437593,
+      "learning_rate": 1.327995299027302e-06,
+      "loss": 0.3686,
+      "step": 41865
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.40479363120874,
+      "learning_rate": 1.3268830728414217e-06,
+      "loss": 0.3541,
+      "step": 41870
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.312797149746925,
+      "learning_rate": 1.325771241348166e-06,
+      "loss": 0.3582,
+      "step": 41875
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.891479836302079,
+      "learning_rate": 1.3246598046670072e-06,
+      "loss": 0.379,
+      "step": 41880
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.884039559814739,
+      "learning_rate": 1.3235487629173733e-06,
+      "loss": 0.3524,
+      "step": 41885
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.228085868933899,
+      "learning_rate": 1.3224381162186489e-06,
+      "loss": 0.3638,
+      "step": 41890
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.482258460260676,
+      "learning_rate": 1.32132786469018e-06,
+      "loss": 0.3572,
+      "step": 41895
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.428868936913378,
+      "learning_rate": 1.3202180084512667e-06,
+      "loss": 0.3888,
+      "step": 41900
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.441922147118252,
+      "learning_rate": 1.319108547621169e-06,
+      "loss": 0.3555,
+      "step": 41905
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.682905296226707,
+      "learning_rate": 1.3179994823191022e-06,
+      "loss": 0.3633,
+      "step": 41910
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.472129864605422,
+      "learning_rate": 1.3168908126642421e-06,
+      "loss": 0.3779,
+      "step": 41915
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.180004486006848,
+      "learning_rate": 1.3157825387757189e-06,
+      "loss": 0.3689,
+      "step": 41920
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.778681656948636,
+      "learning_rate": 1.3146746607726201e-06,
+      "loss": 0.3672,
+      "step": 41925
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.808843573270111,
+      "learning_rate": 1.3135671787739952e-06,
+      "loss": 0.3742,
+      "step": 41930
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.802187639254322,
+      "learning_rate": 1.3124600928988445e-06,
+      "loss": 0.3818,
+      "step": 41935
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.360056543214213,
+      "learning_rate": 1.3113534032661334e-06,
+      "loss": 0.3608,
+      "step": 41940
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 15.847509537782647,
+      "learning_rate": 1.3102471099947778e-06,
+      "loss": 0.3572,
+      "step": 41945
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.841931709990855,
+      "learning_rate": 1.309141213203653e-06,
+      "loss": 0.3387,
+      "step": 41950
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.41016421912756,
+      "learning_rate": 1.3080357130115938e-06,
+      "loss": 0.3792,
+      "step": 41955
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.187559739697147,
+      "learning_rate": 1.306930609537393e-06,
+      "loss": 0.3448,
+      "step": 41960
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.495795948164244,
+      "learning_rate": 1.3058259028997948e-06,
+      "loss": 0.3605,
+      "step": 41965
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.382753309694544,
+      "learning_rate": 1.3047215932175077e-06,
+      "loss": 0.3483,
+      "step": 41970
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.5716181643549465,
+      "learning_rate": 1.3036176806091949e-06,
+      "loss": 0.3419,
+      "step": 41975
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.150721114938997,
+      "learning_rate": 1.302514165193476e-06,
+      "loss": 0.396,
+      "step": 41980
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.307574294717373,
+      "learning_rate": 1.301411047088927e-06,
+      "loss": 0.3588,
+      "step": 41985
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.761829431456866,
+      "learning_rate": 1.3003083264140848e-06,
+      "loss": 0.354,
+      "step": 41990
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.2638307749607565,
+      "learning_rate": 1.2992060032874398e-06,
+      "loss": 0.374,
+      "step": 41995
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.443351903348739,
+      "learning_rate": 1.2981040778274444e-06,
+      "loss": 0.3523,
+      "step": 42000
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.314166116795517,
+      "learning_rate": 1.2970025501525036e-06,
+      "loss": 0.3494,
+      "step": 42005
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.805614747845596,
+      "learning_rate": 1.2959014203809793e-06,
+      "loss": 0.3483,
+      "step": 42010
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.563221109534187,
+      "learning_rate": 1.2948006886311965e-06,
+      "loss": 0.3607,
+      "step": 42015
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.639028664449478,
+      "learning_rate": 1.2937003550214305e-06,
+      "loss": 0.3999,
+      "step": 42020
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.933649264612827,
+      "learning_rate": 1.2926004196699205e-06,
+      "loss": 0.3736,
+      "step": 42025
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.600465104816654,
+      "learning_rate": 1.2915008826948566e-06,
+      "loss": 0.3534,
+      "step": 42030
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.642752716311966,
+      "learning_rate": 1.2904017442143884e-06,
+      "loss": 0.382,
+      "step": 42035
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.383477090892518,
+      "learning_rate": 1.2893030043466264e-06,
+      "loss": 0.3316,
+      "step": 42040
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.6475135081138195,
+      "learning_rate": 1.2882046632096318e-06,
+      "loss": 0.3855,
+      "step": 42045
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.308433393615727,
+      "learning_rate": 1.2871067209214272e-06,
+      "loss": 0.3531,
+      "step": 42050
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.649384180267416,
+      "learning_rate": 1.2860091775999928e-06,
+      "loss": 0.3349,
+      "step": 42055
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.66288589554288,
+      "learning_rate": 1.2849120333632625e-06,
+      "loss": 0.3916,
+      "step": 42060
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.964130849795638,
+      "learning_rate": 1.283815288329131e-06,
+      "loss": 0.3871,
+      "step": 42065
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.8521532342133025,
+      "learning_rate": 1.2827189426154468e-06,
+      "loss": 0.3182,
+      "step": 42070
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.67941770433278,
+      "learning_rate": 1.281622996340019e-06,
+      "loss": 0.359,
+      "step": 42075
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.951347083426389,
+      "learning_rate": 1.2805274496206088e-06,
+      "loss": 0.3873,
+      "step": 42080
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.778512172297239,
+      "learning_rate": 1.279432302574941e-06,
+      "loss": 0.3235,
+      "step": 42085
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.637048999459916,
+      "learning_rate": 1.2783375553206929e-06,
+      "loss": 0.3845,
+      "step": 42090
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 13.564738367310575,
+      "learning_rate": 1.2772432079754975e-06,
+      "loss": 0.4089,
+      "step": 42095
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.802261635220572,
+      "learning_rate": 1.2761492606569508e-06,
+      "loss": 0.3561,
+      "step": 42100
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.420038409930823,
+      "learning_rate": 1.2750557134825996e-06,
+      "loss": 0.3524,
+      "step": 42105
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.055511029770161,
+      "learning_rate": 1.2739625665699523e-06,
+      "loss": 0.3548,
+      "step": 42110
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.911610320868343,
+      "learning_rate": 1.2728698200364719e-06,
+      "loss": 0.3913,
+      "step": 42115
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.824163423306069,
+      "learning_rate": 1.271777473999577e-06,
+      "loss": 0.3643,
+      "step": 42120
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.496457440085291,
+      "learning_rate": 1.2706855285766473e-06,
+      "loss": 0.3672,
+      "step": 42125
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.7146890653985585,
+      "learning_rate": 1.2695939838850151e-06,
+      "loss": 0.3628,
+      "step": 42130
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.628729247949617,
+      "learning_rate": 1.2685028400419747e-06,
+      "loss": 0.3682,
+      "step": 42135
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.454871913383951,
+      "learning_rate": 1.2674120971647709e-06,
+      "loss": 0.3438,
+      "step": 42140
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.98463705586402,
+      "learning_rate": 1.2663217553706098e-06,
+      "loss": 0.3374,
+      "step": 42145
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.012009429245551,
+      "learning_rate": 1.2652318147766563e-06,
+      "loss": 0.3671,
+      "step": 42150
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.376290256788171,
+      "learning_rate": 1.264142275500025e-06,
+      "loss": 0.3498,
+      "step": 42155
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 19.3603246234026,
+      "learning_rate": 1.2630531376577954e-06,
+      "loss": 0.4071,
+      "step": 42160
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 7.572495194724955,
+      "learning_rate": 1.2619644013669974e-06,
+      "loss": 0.3641,
+      "step": 42165
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.77345672789725,
+      "learning_rate": 1.260876066744623e-06,
+      "loss": 0.3684,
+      "step": 42170
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.530893686920582,
+      "learning_rate": 1.2597881339076174e-06,
+      "loss": 0.369,
+      "step": 42175
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.588971650663712,
+      "learning_rate": 1.2587006029728815e-06,
+      "loss": 0.3582,
+      "step": 42180
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.895544639275465,
+      "learning_rate": 1.2576134740572798e-06,
+      "loss": 0.3235,
+      "step": 42185
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.902353707970118,
+      "learning_rate": 1.2565267472776249e-06,
+      "loss": 0.3758,
+      "step": 42190
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.275755681550319,
+      "learning_rate": 1.2554404227506933e-06,
+      "loss": 0.3716,
+      "step": 42195
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.474480660736474,
+      "learning_rate": 1.2543545005932146e-06,
+      "loss": 0.3747,
+      "step": 42200
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.420762975329743,
+      "learning_rate": 1.2532689809218745e-06,
+      "loss": 0.378,
+      "step": 42205
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.217549343063801,
+      "learning_rate": 1.252183863853319e-06,
+      "loss": 0.388,
+      "step": 42210
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.467735935719349,
+      "learning_rate": 1.2510991495041463e-06,
+      "loss": 0.3626,
+      "step": 42215
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.850374283615647,
+      "learning_rate": 1.2500148379909166e-06,
+      "loss": 0.3696,
+      "step": 42220
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.496383422687956,
+      "learning_rate": 1.248930929430141e-06,
+      "loss": 0.3719,
+      "step": 42225
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.242307280006708,
+      "learning_rate": 1.2478474239382932e-06,
+      "loss": 0.3088,
+      "step": 42230
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.058077876542059,
+      "learning_rate": 1.2467643216317982e-06,
+      "loss": 0.3775,
+      "step": 42235
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.471919999541183,
+      "learning_rate": 1.2456816226270407e-06,
+      "loss": 0.3331,
+      "step": 42240
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.659201725476392,
+      "learning_rate": 1.2445993270403634e-06,
+      "loss": 0.3609,
+      "step": 42245
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.942208712849299,
+      "learning_rate": 1.2435174349880607e-06,
+      "loss": 0.3591,
+      "step": 42250
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.523505832121196,
+      "learning_rate": 1.24243594658639e-06,
+      "loss": 0.3843,
+      "step": 42255
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.947283233247577,
+      "learning_rate": 1.2413548619515603e-06,
+      "loss": 0.3588,
+      "step": 42260
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.292315925839984,
+      "learning_rate": 1.240274181199737e-06,
+      "loss": 0.3512,
+      "step": 42265
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.743756549305206,
+      "learning_rate": 1.2391939044470475e-06,
+      "loss": 0.3754,
+      "step": 42270
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.826997907589478,
+      "learning_rate": 1.2381140318095691e-06,
+      "loss": 0.3956,
+      "step": 42275
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.690877213456252,
+      "learning_rate": 1.2370345634033421e-06,
+      "loss": 0.3751,
+      "step": 42280
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.609682999814088,
+      "learning_rate": 1.2359554993443584e-06,
+      "loss": 0.3395,
+      "step": 42285
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.598513227330498,
+      "learning_rate": 1.2348768397485666e-06,
+      "loss": 0.3683,
+      "step": 42290
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.662917186226731,
+      "learning_rate": 1.2337985847318772e-06,
+      "loss": 0.3472,
+      "step": 42295
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.422588702274017,
+      "learning_rate": 1.2327207344101493e-06,
+      "loss": 0.381,
+      "step": 42300
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.505366246396598,
+      "learning_rate": 1.2316432888992069e-06,
+      "loss": 0.3559,
+      "step": 42305
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.3381788087830255,
+      "learning_rate": 1.2305662483148217e-06,
+      "loss": 0.3448,
+      "step": 42310
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.294779684309665,
+      "learning_rate": 1.2294896127727313e-06,
+      "loss": 0.3684,
+      "step": 42315
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.446079138501898,
+      "learning_rate": 1.2284133823886206e-06,
+      "loss": 0.37,
+      "step": 42320
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.9445942664675675,
+      "learning_rate": 1.2273375572781388e-06,
+      "loss": 0.3701,
+      "step": 42325
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.498016661428263,
+      "learning_rate": 1.226262137556885e-06,
+      "loss": 0.3868,
+      "step": 42330
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.422679423110836,
+      "learning_rate": 1.2251871233404195e-06,
+      "loss": 0.3589,
+      "step": 42335
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.371199858577176,
+      "learning_rate": 1.2241125147442583e-06,
+      "loss": 0.3624,
+      "step": 42340
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.884361530712702,
+      "learning_rate": 1.2230383118838718e-06,
+      "loss": 0.3552,
+      "step": 42345
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.562749827350988,
+      "learning_rate": 1.2219645148746862e-06,
+      "loss": 0.3586,
+      "step": 42350
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.968413991488184,
+      "learning_rate": 1.2208911238320887e-06,
+      "loss": 0.3702,
+      "step": 42355
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.38240302692401,
+      "learning_rate": 1.219818138871417e-06,
+      "loss": 0.3436,
+      "step": 42360
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.866090970297154,
+      "learning_rate": 1.2187455601079706e-06,
+      "loss": 0.3522,
+      "step": 42365
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.71774457805276,
+      "learning_rate": 1.2176733876570008e-06,
+      "loss": 0.3305,
+      "step": 42370
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.715950757583577,
+      "learning_rate": 1.2166016216337195e-06,
+      "loss": 0.3383,
+      "step": 42375
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.780255750910164,
+      "learning_rate": 1.215530262153291e-06,
+      "loss": 0.3639,
+      "step": 42380
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 7.649557384828715,
+      "learning_rate": 1.2144593093308365e-06,
+      "loss": 0.3456,
+      "step": 42385
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.658998179623315,
+      "learning_rate": 1.2133887632814378e-06,
+      "loss": 0.3543,
+      "step": 42390
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.047139095782406,
+      "learning_rate": 1.2123186241201269e-06,
+      "loss": 0.3914,
+      "step": 42395
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.460150286641404,
+      "learning_rate": 1.2112488919618975e-06,
+      "loss": 0.3773,
+      "step": 42400
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.863819087309987,
+      "learning_rate": 1.2101795669216954e-06,
+      "loss": 0.3417,
+      "step": 42405
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.497138328333464,
+      "learning_rate": 1.2091106491144233e-06,
+      "loss": 0.3681,
+      "step": 42410
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.901559440697807,
+      "learning_rate": 1.2080421386549423e-06,
+      "loss": 0.3704,
+      "step": 42415
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.626243185632536,
+      "learning_rate": 1.20697403565807e-06,
+      "loss": 0.3946,
+      "step": 42420
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.25988784988983,
+      "learning_rate": 1.2059063402385762e-06,
+      "loss": 0.3673,
+      "step": 42425
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.092003009439223,
+      "learning_rate": 1.2048390525111925e-06,
+      "loss": 0.3741,
+      "step": 42430
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.992085822155388,
+      "learning_rate": 1.2037721725906e-06,
+      "loss": 0.3554,
+      "step": 42435
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.586761164031409,
+      "learning_rate": 1.2027057005914434e-06,
+      "loss": 0.3566,
+      "step": 42440
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.92405306572773,
+      "learning_rate": 1.2016396366283162e-06,
+      "loss": 0.3755,
+      "step": 42445
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.7923894461247025,
+      "learning_rate": 1.2005739808157751e-06,
+      "loss": 0.3504,
+      "step": 42450
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.044599357681824,
+      "learning_rate": 1.199508733268327e-06,
+      "loss": 0.3821,
+      "step": 42455
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.839607446144359,
+      "learning_rate": 1.1984438941004389e-06,
+      "loss": 0.3442,
+      "step": 42460
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.089035579665456,
+      "learning_rate": 1.1973794634265322e-06,
+      "loss": 0.3651,
+      "step": 42465
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.748395692327808,
+      "learning_rate": 1.1963154413609835e-06,
+      "loss": 0.3343,
+      "step": 42470
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.693467743332488,
+      "learning_rate": 1.1952518280181286e-06,
+      "loss": 0.3694,
+      "step": 42475
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.399528781148906,
+      "learning_rate": 1.1941886235122547e-06,
+      "loss": 0.3695,
+      "step": 42480
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.092356045785655,
+      "learning_rate": 1.1931258279576112e-06,
+      "loss": 0.3758,
+      "step": 42485
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 8.137382948152087,
+      "learning_rate": 1.1920634414683985e-06,
+      "loss": 0.3444,
+      "step": 42490
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.251183036824891,
+      "learning_rate": 1.1910014641587737e-06,
+      "loss": 0.3453,
+      "step": 42495
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.391307992928254,
+      "learning_rate": 1.1899398961428527e-06,
+      "loss": 0.3776,
+      "step": 42500
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.725950315746725,
+      "learning_rate": 1.1888787375347037e-06,
+      "loss": 0.3646,
+      "step": 42505
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.096875154747406,
+      "learning_rate": 1.1878179884483544e-06,
+      "loss": 0.3476,
+      "step": 42510
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.9398674542331715,
+      "learning_rate": 1.1867576489977878e-06,
+      "loss": 0.3553,
+      "step": 42515
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.314723165258239,
+      "learning_rate": 1.18569771929694e-06,
+      "loss": 0.365,
+      "step": 42520
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.729494773691787,
+      "learning_rate": 1.1846381994597073e-06,
+      "loss": 0.3549,
+      "step": 42525
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.281703371643079,
+      "learning_rate": 1.1835790895999365e-06,
+      "loss": 0.3382,
+      "step": 42530
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.897607832496858,
+      "learning_rate": 1.182520389831438e-06,
+      "loss": 0.3529,
+      "step": 42535
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.540451706205725,
+      "learning_rate": 1.1814621002679694e-06,
+      "loss": 0.3613,
+      "step": 42540
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.224883443222026,
+      "learning_rate": 1.1804042210232525e-06,
+      "loss": 0.3803,
+      "step": 42545
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.624150968078994,
+      "learning_rate": 1.1793467522109587e-06,
+      "loss": 0.3707,
+      "step": 42550
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.059456294473843,
+      "learning_rate": 1.178289693944717e-06,
+      "loss": 0.3825,
+      "step": 42555
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.0517341329358665,
+      "learning_rate": 1.1772330463381154e-06,
+      "loss": 0.3544,
+      "step": 42560
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.6496531285395495,
+      "learning_rate": 1.1761768095046933e-06,
+      "loss": 0.3919,
+      "step": 42565
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.2775145645304,
+      "learning_rate": 1.1751209835579497e-06,
+      "loss": 0.3636,
+      "step": 42570
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.936966101273284,
+      "learning_rate": 1.1740655686113373e-06,
+      "loss": 0.3707,
+      "step": 42575
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.170511625514858,
+      "learning_rate": 1.173010564778263e-06,
+      "loss": 0.3662,
+      "step": 42580
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.469860799493985,
+      "learning_rate": 1.1719559721720948e-06,
+      "loss": 0.3556,
+      "step": 42585
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.509544621385367,
+      "learning_rate": 1.1709017909061504e-06,
+      "loss": 0.3343,
+      "step": 42590
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.839685983344473,
+      "learning_rate": 1.1698480210937092e-06,
+      "loss": 0.3686,
+      "step": 42595
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.258152205207002,
+      "learning_rate": 1.168794662848e-06,
+      "loss": 0.3275,
+      "step": 42600
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.706633011505221,
+      "learning_rate": 1.167741716282213e-06,
+      "loss": 0.3826,
+      "step": 42605
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.916007626626051,
+      "learning_rate": 1.1666891815094933e-06,
+      "loss": 0.3816,
+      "step": 42610
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.4945402078139,
+      "learning_rate": 1.1656370586429366e-06,
+      "loss": 0.3805,
+      "step": 42615
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.840101890634254,
+      "learning_rate": 1.1645853477956025e-06,
+      "loss": 0.3644,
+      "step": 42620
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.07723512998222,
+      "learning_rate": 1.1635340490804981e-06,
+      "loss": 0.3758,
+      "step": 42625
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.013491122624717,
+      "learning_rate": 1.1624831626105932e-06,
+      "loss": 0.3643,
+      "step": 42630
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.7561952960882845,
+      "learning_rate": 1.1614326884988086e-06,
+      "loss": 0.3508,
+      "step": 42635
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.28442183659094,
+      "learning_rate": 1.1603826268580209e-06,
+      "loss": 0.3568,
+      "step": 42640
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.271824491240048,
+      "learning_rate": 1.159332977801067e-06,
+      "loss": 0.3935,
+      "step": 42645
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.488640991513405,
+      "learning_rate": 1.1582837414407332e-06,
+      "loss": 0.3315,
+      "step": 42650
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 8.378671217242998,
+      "learning_rate": 1.1572349178897686e-06,
+      "loss": 0.4071,
+      "step": 42655
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.853148764565508,
+      "learning_rate": 1.1561865072608702e-06,
+      "loss": 0.3498,
+      "step": 42660
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.1490141479757385,
+      "learning_rate": 1.155138509666695e-06,
+      "loss": 0.3656,
+      "step": 42665
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.224691882158689,
+      "learning_rate": 1.1540909252198568e-06,
+      "loss": 0.3624,
+      "step": 42670
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.185636218938329,
+      "learning_rate": 1.1530437540329203e-06,
+      "loss": 0.3915,
+      "step": 42675
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.770407006739393,
+      "learning_rate": 1.1519969962184124e-06,
+      "loss": 0.3896,
+      "step": 42680
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.192529260782055,
+      "learning_rate": 1.1509506518888074e-06,
+      "loss": 0.3236,
+      "step": 42685
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.517305542649535,
+      "learning_rate": 1.1499047211565445e-06,
+      "loss": 0.3724,
+      "step": 42690
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.4512329579728265,
+      "learning_rate": 1.148859204134009e-06,
+      "loss": 0.3724,
+      "step": 42695
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.542600508957431,
+      "learning_rate": 1.1478141009335486e-06,
+      "loss": 0.3566,
+      "step": 42700
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.840329621827059,
+      "learning_rate": 1.1467694116674654e-06,
+      "loss": 0.3592,
+      "step": 42705
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.726700809548532,
+      "learning_rate": 1.145725136448013e-06,
+      "loss": 0.3916,
+      "step": 42710
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.823408757694335,
+      "learning_rate": 1.1446812753874066e-06,
+      "loss": 0.3442,
+      "step": 42715
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.880849935296445,
+      "learning_rate": 1.1436378285978117e-06,
+      "loss": 0.3552,
+      "step": 42720
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.581109594625088,
+      "learning_rate": 1.1425947961913503e-06,
+      "loss": 0.3475,
+      "step": 42725
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.275388164638105,
+      "learning_rate": 1.1415521782801036e-06,
+      "loss": 0.3282,
+      "step": 42730
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.677484122563242,
+      "learning_rate": 1.1405099749761022e-06,
+      "loss": 0.3294,
+      "step": 42735
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.947584927636766,
+      "learning_rate": 1.1394681863913388e-06,
+      "loss": 0.3728,
+      "step": 42740
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.102466767534395,
+      "learning_rate": 1.1384268126377568e-06,
+      "loss": 0.3935,
+      "step": 42745
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.320546329072317,
+      "learning_rate": 1.1373858538272547e-06,
+      "loss": 0.3589,
+      "step": 42750
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.789518230043431,
+      "learning_rate": 1.1363453100716908e-06,
+      "loss": 0.3497,
+      "step": 42755
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.505961748227482,
+      "learning_rate": 1.1353051814828736e-06,
+      "loss": 0.3812,
+      "step": 42760
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.541489597493219,
+      "learning_rate": 1.1342654681725724e-06,
+      "loss": 0.3666,
+      "step": 42765
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.79090840679571,
+      "learning_rate": 1.1332261702525072e-06,
+      "loss": 0.3465,
+      "step": 42770
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.566615663460663,
+      "learning_rate": 1.1321872878343542e-06,
+      "loss": 0.3235,
+      "step": 42775
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.302393540097486,
+      "learning_rate": 1.1311488210297472e-06,
+      "loss": 0.3756,
+      "step": 42780
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.282567677613329,
+      "learning_rate": 1.130110769950275e-06,
+      "loss": 0.35,
+      "step": 42785
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.50860637000532,
+      "learning_rate": 1.1290731347074784e-06,
+      "loss": 0.3585,
+      "step": 42790
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.8996798469259755,
+      "learning_rate": 1.1280359154128572e-06,
+      "loss": 0.3395,
+      "step": 42795
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.606964117917527,
+      "learning_rate": 1.1269991121778667e-06,
+      "loss": 0.3148,
+      "step": 42800
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.733248327210414,
+      "learning_rate": 1.1259627251139144e-06,
+      "loss": 0.3563,
+      "step": 42805
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.947891824404803,
+      "learning_rate": 1.1249267543323633e-06,
+      "loss": 0.35,
+      "step": 42810
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.490153591450526,
+      "learning_rate": 1.1238911999445356e-06,
+      "loss": 0.3395,
+      "step": 42815
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.539950352977154,
+      "learning_rate": 1.122856062061704e-06,
+      "loss": 0.3653,
+      "step": 42820
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.6516196029149315,
+      "learning_rate": 1.1218213407951006e-06,
+      "loss": 0.3541,
+      "step": 42825
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.396347971692736,
+      "learning_rate": 1.1207870362559104e-06,
+      "loss": 0.3231,
+      "step": 42830
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.001218368092357,
+      "learning_rate": 1.119753148555272e-06,
+      "loss": 0.3597,
+      "step": 42835
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.1681320005852225,
+      "learning_rate": 1.1187196778042835e-06,
+      "loss": 0.3669,
+      "step": 42840
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.593220039690747,
+      "learning_rate": 1.1176866241139938e-06,
+      "loss": 0.3589,
+      "step": 42845
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.039168848888276,
+      "learning_rate": 1.116653987595412e-06,
+      "loss": 0.3171,
+      "step": 42850
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.952587473166305,
+      "learning_rate": 1.1156217683594962e-06,
+      "loss": 0.3621,
+      "step": 42855
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.259426043426545,
+      "learning_rate": 1.114589966517166e-06,
+      "loss": 0.3396,
+      "step": 42860
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.18436919600694,
+      "learning_rate": 1.113558582179291e-06,
+      "loss": 0.3954,
+      "step": 42865
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.869072471229186,
+      "learning_rate": 1.1125276154566978e-06,
+      "loss": 0.357,
+      "step": 42870
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.491125796378438,
+      "learning_rate": 1.1114970664601687e-06,
+      "loss": 0.3881,
+      "step": 42875
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.614701961113036,
+      "learning_rate": 1.1104669353004426e-06,
+      "loss": 0.3985,
+      "step": 42880
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.400489516627011,
+      "learning_rate": 1.1094372220882082e-06,
+      "loss": 0.3709,
+      "step": 42885
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.574379662364799,
+      "learning_rate": 1.1084079269341168e-06,
+      "loss": 0.3359,
+      "step": 42890
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.713964177768655,
+      "learning_rate": 1.1073790499487668e-06,
+      "loss": 0.3641,
+      "step": 42895
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.612098728645514,
+      "learning_rate": 1.1063505912427187e-06,
+      "loss": 0.3511,
+      "step": 42900
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.362508971976225,
+      "learning_rate": 1.105322550926482e-06,
+      "loss": 0.3225,
+      "step": 42905
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.2506411846191146,
+      "learning_rate": 1.1042949291105275e-06,
+      "loss": 0.3573,
+      "step": 42910
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.973191182475268,
+      "learning_rate": 1.103267725905276e-06,
+      "loss": 0.3522,
+      "step": 42915
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.209285931134533,
+      "learning_rate": 1.1022409414211033e-06,
+      "loss": 0.3415,
+      "step": 42920
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.143845947568642,
+      "learning_rate": 1.1012145757683446e-06,
+      "loss": 0.3852,
+      "step": 42925
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.335144342818536,
+      "learning_rate": 1.1001886290572855e-06,
+      "loss": 0.3492,
+      "step": 42930
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.484389042032344,
+      "learning_rate": 1.0991631013981708e-06,
+      "loss": 0.3546,
+      "step": 42935
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.884075111752374,
+      "learning_rate": 1.0981379929011949e-06,
+      "loss": 0.3869,
+      "step": 42940
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.038709618640821,
+      "learning_rate": 1.097113303676513e-06,
+      "loss": 0.3774,
+      "step": 42945
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.694482807404403,
+      "learning_rate": 1.0960890338342323e-06,
+      "loss": 0.3453,
+      "step": 42950
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.90117682929359,
+      "learning_rate": 1.0950651834844123e-06,
+      "loss": 0.3636,
+      "step": 42955
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.266971892132762,
+      "learning_rate": 1.0940417527370733e-06,
+      "loss": 0.3496,
+      "step": 42960
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.075111698053483,
+      "learning_rate": 1.0930187417021849e-06,
+      "loss": 0.3943,
+      "step": 42965
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.645309782701889,
+      "learning_rate": 1.091996150489676e-06,
+      "loss": 0.3845,
+      "step": 42970
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.473858431297266,
+      "learning_rate": 1.0909739792094286e-06,
+      "loss": 0.3612,
+      "step": 42975
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.901206521594583,
+      "learning_rate": 1.0899522279712782e-06,
+      "loss": 0.3569,
+      "step": 42980
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.357958252896021,
+      "learning_rate": 1.0889308968850187e-06,
+      "loss": 0.3739,
+      "step": 42985
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.188909256230952,
+      "learning_rate": 1.0879099860603937e-06,
+      "loss": 0.3809,
+      "step": 42990
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.366863968183281,
+      "learning_rate": 1.0868894956071075e-06,
+      "loss": 0.3083,
+      "step": 42995
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.3366329068196645,
+      "learning_rate": 1.0858694256348152e-06,
+      "loss": 0.3376,
+      "step": 43000
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.066228333290048,
+      "learning_rate": 1.0848497762531251e-06,
+      "loss": 0.3336,
+      "step": 43005
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.498027883393406,
+      "learning_rate": 1.0838305475716078e-06,
+      "loss": 0.3453,
+      "step": 43010
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.302131309750226,
+      "learning_rate": 1.0828117396997794e-06,
+      "loss": 0.382,
+      "step": 43015
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.5950376168562626,
+      "learning_rate": 1.0817933527471197e-06,
+      "loss": 0.36,
+      "step": 43020
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.522134318094178,
+      "learning_rate": 1.0807753868230547e-06,
+      "loss": 0.3649,
+      "step": 43025
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.83539842348374,
+      "learning_rate": 1.0797578420369725e-06,
+      "loss": 0.3991,
+      "step": 43030
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.378807767473629,
+      "learning_rate": 1.078740718498212e-06,
+      "loss": 0.3962,
+      "step": 43035
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.726310380418322,
+      "learning_rate": 1.0777240163160652e-06,
+      "loss": 0.3541,
+      "step": 43040
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.589454089631747,
+      "learning_rate": 1.0767077355997847e-06,
+      "loss": 0.3663,
+      "step": 43045
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.553776969200796,
+      "learning_rate": 1.0756918764585717e-06,
+      "loss": 0.3674,
+      "step": 43050
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.684727431490814,
+      "learning_rate": 1.0746764390015867e-06,
+      "loss": 0.368,
+      "step": 43055
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 7.246213698396501,
+      "learning_rate": 1.073661423337941e-06,
+      "loss": 0.3716,
+      "step": 43060
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.579392698898941,
+      "learning_rate": 1.072646829576704e-06,
+      "loss": 0.3986,
+      "step": 43065
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.698311915325066,
+      "learning_rate": 1.0716326578268988e-06,
+      "loss": 0.3397,
+      "step": 43070
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 7.239743051282497,
+      "learning_rate": 1.0706189081975005e-06,
+      "loss": 0.3646,
+      "step": 43075
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.668251711338613,
+      "learning_rate": 1.0696055807974436e-06,
+      "loss": 0.3549,
+      "step": 43080
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.869417168100637,
+      "learning_rate": 1.0685926757356136e-06,
+      "loss": 0.3339,
+      "step": 43085
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.660513286141073,
+      "learning_rate": 1.067580193120849e-06,
+      "loss": 0.3446,
+      "step": 43090
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.536550325931741,
+      "learning_rate": 1.0665681330619499e-06,
+      "loss": 0.3698,
+      "step": 43095
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.4121793911309295,
+      "learning_rate": 1.0655564956676628e-06,
+      "loss": 0.3514,
+      "step": 43100
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.550953081210822,
+      "learning_rate": 1.0645452810466956e-06,
+      "loss": 0.3483,
+      "step": 43105
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.033125702994512,
+      "learning_rate": 1.063534489307705e-06,
+      "loss": 0.3503,
+      "step": 43110
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.810283334327166,
+      "learning_rate": 1.062524120559308e-06,
+      "loss": 0.3596,
+      "step": 43115
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.9444574105968435,
+      "learning_rate": 1.0615141749100711e-06,
+      "loss": 0.3638,
+      "step": 43120
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.680441825295457,
+      "learning_rate": 1.060504652468517e-06,
+      "loss": 0.341,
+      "step": 43125
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.313650136760696,
+      "learning_rate": 1.059495553343126e-06,
+      "loss": 0.352,
+      "step": 43130
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.542408747405393,
+      "learning_rate": 1.0584868776423267e-06,
+      "loss": 0.3736,
+      "step": 43135
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.532313524525178,
+      "learning_rate": 1.0574786254745095e-06,
+      "loss": 0.3849,
+      "step": 43140
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.9027632410283095,
+      "learning_rate": 1.056470796948012e-06,
+      "loss": 0.3601,
+      "step": 43145
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.405975228332189,
+      "learning_rate": 1.0554633921711327e-06,
+      "loss": 0.3281,
+      "step": 43150
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.789522903631732,
+      "learning_rate": 1.0544564112521193e-06,
+      "loss": 0.3376,
+      "step": 43155
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.0475364087497745,
+      "learning_rate": 1.0534498542991773e-06,
+      "loss": 0.3674,
+      "step": 43160
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.560707161998586,
+      "learning_rate": 1.0524437214204675e-06,
+      "loss": 0.3125,
+      "step": 43165
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.614296182908391,
+      "learning_rate": 1.051438012724102e-06,
+      "loss": 0.3427,
+      "step": 43170
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.488209880985004,
+      "learning_rate": 1.0504327283181465e-06,
+      "loss": 0.3493,
+      "step": 43175
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.8567613946857975,
+      "learning_rate": 1.0494278683106268e-06,
+      "loss": 0.359,
+      "step": 43180
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.397052687821712,
+      "learning_rate": 1.048423432809516e-06,
+      "loss": 0.3337,
+      "step": 43185
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.629733818602774,
+      "learning_rate": 1.047419421922749e-06,
+      "loss": 0.3174,
+      "step": 43190
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.634847597091976,
+      "learning_rate": 1.0464158357582078e-06,
+      "loss": 0.35,
+      "step": 43195
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.876626716703233,
+      "learning_rate": 1.0454126744237342e-06,
+      "loss": 0.3542,
+      "step": 43200
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.902056100936468,
+      "learning_rate": 1.0444099380271222e-06,
+      "loss": 0.3596,
+      "step": 43205
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.506384279963405,
+      "learning_rate": 1.0434076266761184e-06,
+      "loss": 0.3707,
+      "step": 43210
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.040252743715373,
+      "learning_rate": 1.0424057404784282e-06,
+      "loss": 0.3674,
+      "step": 43215
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.047683764095478,
+      "learning_rate": 1.0414042795417056e-06,
+      "loss": 0.3452,
+      "step": 43220
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.0903219258364345,
+      "learning_rate": 1.0404032439735655e-06,
+      "loss": 0.3472,
+      "step": 43225
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.8555499240650635,
+      "learning_rate": 1.0394026338815716e-06,
+      "loss": 0.3105,
+      "step": 43230
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.215914172669765,
+      "learning_rate": 1.0384024493732431e-06,
+      "loss": 0.3762,
+      "step": 43235
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.543103488290535,
+      "learning_rate": 1.0374026905560553e-06,
+      "loss": 0.3377,
+      "step": 43240
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.7548177995340914,
+      "learning_rate": 1.0364033575374377e-06,
+      "loss": 0.3668,
+      "step": 43245
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.892691110099438,
+      "learning_rate": 1.0354044504247713e-06,
+      "loss": 0.3272,
+      "step": 43250
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.67963894133362,
+      "learning_rate": 1.0344059693253944e-06,
+      "loss": 0.327,
+      "step": 43255
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.5946017789824625,
+      "learning_rate": 1.033407914346597e-06,
+      "loss": 0.3309,
+      "step": 43260
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.9379160496499654,
+      "learning_rate": 1.032410285595626e-06,
+      "loss": 0.3529,
+      "step": 43265
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.580929557843822,
+      "learning_rate": 1.0314130831796787e-06,
+      "loss": 0.3578,
+      "step": 43270
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.522582007485534,
+      "learning_rate": 1.0304163072059115e-06,
+      "loss": 0.3405,
+      "step": 43275
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.714784217768867,
+      "learning_rate": 1.0294199577814302e-06,
+      "loss": 0.3355,
+      "step": 43280
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.726055189950499,
+      "learning_rate": 1.028424035013299e-06,
+      "loss": 0.3629,
+      "step": 43285
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.863915755657142,
+      "learning_rate": 1.0274285390085336e-06,
+      "loss": 0.3264,
+      "step": 43290
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.136164250849754,
+      "learning_rate": 1.0264334698741024e-06,
+      "loss": 0.3601,
+      "step": 43295
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.512042974519487,
+      "learning_rate": 1.025438827716933e-06,
+      "loss": 0.3655,
+      "step": 43300
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.788476683195161,
+      "learning_rate": 1.0244446126439012e-06,
+      "loss": 0.364,
+      "step": 43305
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.477570487488166,
+      "learning_rate": 1.0234508247618423e-06,
+      "loss": 0.3542,
+      "step": 43310
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.142502088098034,
+      "learning_rate": 1.0224574641775426e-06,
+      "loss": 0.3746,
+      "step": 43315
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.543256946603856,
+      "learning_rate": 1.021464530997741e-06,
+      "loss": 0.36,
+      "step": 43320
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.352514628843138,
+      "learning_rate": 1.0204720253291357e-06,
+      "loss": 0.3744,
+      "step": 43325
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.53314883612779,
+      "learning_rate": 1.0194799472783729e-06,
+      "loss": 0.3563,
+      "step": 43330
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.780028672157098,
+      "learning_rate": 1.0184882969520565e-06,
+      "loss": 0.3811,
+      "step": 43335
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.707034646085166,
+      "learning_rate": 1.0174970744567464e-06,
+      "loss": 0.3633,
+      "step": 43340
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.552000652056377,
+      "learning_rate": 1.01650627989895e-06,
+      "loss": 0.368,
+      "step": 43345
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.884540408630439,
+      "learning_rate": 1.015515913385136e-06,
+      "loss": 0.3549,
+      "step": 43350
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.836787043370082,
+      "learning_rate": 1.0145259750217208e-06,
+      "loss": 0.3502,
+      "step": 43355
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.974463732444184,
+      "learning_rate": 1.0135364649150798e-06,
+      "loss": 0.3524,
+      "step": 43360
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.463646137280299,
+      "learning_rate": 1.0125473831715382e-06,
+      "loss": 0.3175,
+      "step": 43365
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.77740823679646,
+      "learning_rate": 1.0115587298973794e-06,
+      "loss": 0.3332,
+      "step": 43370
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.965352720357074,
+      "learning_rate": 1.0105705051988373e-06,
+      "loss": 0.3415,
+      "step": 43375
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.1761805683383075,
+      "learning_rate": 1.0095827091820997e-06,
+      "loss": 0.3741,
+      "step": 43380
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.311466247664548,
+      "learning_rate": 1.0085953419533128e-06,
+      "loss": 0.3586,
+      "step": 43385
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.618416334591783,
+      "learning_rate": 1.0076084036185701e-06,
+      "loss": 0.3567,
+      "step": 43390
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.466280404764819,
+      "learning_rate": 1.0066218942839257e-06,
+      "loss": 0.3581,
+      "step": 43395
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.638259497898554,
+      "learning_rate": 1.0056358140553829e-06,
+      "loss": 0.3376,
+      "step": 43400
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.555066038395794,
+      "learning_rate": 1.0046501630388988e-06,
+      "loss": 0.341,
+      "step": 43405
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.192399290954197,
+      "learning_rate": 1.0036649413403893e-06,
+      "loss": 0.3805,
+      "step": 43410
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.6710061054899805,
+      "learning_rate": 1.0026801490657173e-06,
+      "loss": 0.3746,
+      "step": 43415
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.931779119046791,
+      "learning_rate": 1.001695786320706e-06,
+      "loss": 0.3754,
+      "step": 43420
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.848174812797477,
+      "learning_rate": 1.0007118532111275e-06,
+      "loss": 0.3312,
+      "step": 43425
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.034035453384377,
+      "learning_rate": 9.9972834984271e-07,
+      "loss": 0.3609,
+      "step": 43430
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.634366268733935,
+      "learning_rate": 9.987452763211381e-07,
+      "loss": 0.3452,
+      "step": 43435
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.415788569573445,
+      "learning_rate": 9.977626327520428e-07,
+      "loss": 0.363,
+      "step": 43440
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.055785616882905,
+      "learning_rate": 9.96780419241018e-07,
+      "loss": 0.3263,
+      "step": 43445
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.897312973020388,
+      "learning_rate": 9.95798635893603e-07,
+      "loss": 0.3471,
+      "step": 43450
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.506357394633795,
+      "learning_rate": 9.948172828152974e-07,
+      "loss": 0.3296,
+      "step": 43455
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.7375468390477105,
+      "learning_rate": 9.938363601115514e-07,
+      "loss": 0.3517,
+      "step": 43460
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.676075268107502,
+      "learning_rate": 9.928558678877676e-07,
+      "loss": 0.3679,
+      "step": 43465
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.4475776117003,
+      "learning_rate": 9.918758062493073e-07,
+      "loss": 0.348,
+      "step": 43470
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.395459201622998,
+      "learning_rate": 9.90896175301479e-07,
+      "loss": 0.3621,
+      "step": 43475
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.848568076893303,
+      "learning_rate": 9.899169751495518e-07,
+      "loss": 0.3494,
+      "step": 43480
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.583999888517346,
+      "learning_rate": 9.889382058987434e-07,
+      "loss": 0.3519,
+      "step": 43485
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.531939939673199,
+      "learning_rate": 9.879598676542257e-07,
+      "loss": 0.3785,
+      "step": 43490
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.210142755797226,
+      "learning_rate": 9.869819605211277e-07,
+      "loss": 0.3531,
+      "step": 43495
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.396586341580197,
+      "learning_rate": 9.86004484604528e-07,
+      "loss": 0.3488,
+      "step": 43500
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.882581756833667,
+      "learning_rate": 9.850274400094627e-07,
+      "loss": 0.3387,
+      "step": 43505
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.7304096470247705,
+      "learning_rate": 9.840508268409171e-07,
+      "loss": 0.3542,
+      "step": 43510
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.190917960802614,
+      "learning_rate": 9.830746452038353e-07,
+      "loss": 0.3603,
+      "step": 43515
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.0149739070574775,
+      "learning_rate": 9.820988952031092e-07,
+      "loss": 0.3486,
+      "step": 43520
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.474595154363621,
+      "learning_rate": 9.811235769435896e-07,
+      "loss": 0.3539,
+      "step": 43525
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.8841178221262185,
+      "learning_rate": 9.801486905300795e-07,
+      "loss": 0.3623,
+      "step": 43530
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.820060284256322,
+      "learning_rate": 9.791742360673317e-07,
+      "loss": 0.3544,
+      "step": 43535
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.496545564226054,
+      "learning_rate": 9.782002136600594e-07,
+      "loss": 0.3895,
+      "step": 43540
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.524870034904284,
+      "learning_rate": 9.772266234129235e-07,
+      "loss": 0.3465,
+      "step": 43545
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.7122840063230935,
+      "learning_rate": 9.7625346543054e-07,
+      "loss": 0.363,
+      "step": 43550
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.824248073000432,
+      "learning_rate": 9.7528073981748e-07,
+      "loss": 0.3412,
+      "step": 43555
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.7221990418797715,
+      "learning_rate": 9.743084466782666e-07,
+      "loss": 0.3534,
+      "step": 43560
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.7464187992370315,
+      "learning_rate": 9.733365861173782e-07,
+      "loss": 0.3491,
+      "step": 43565
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.658102500981284,
+      "learning_rate": 9.723651582392451e-07,
+      "loss": 0.3698,
+      "step": 43570
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.818336991693816,
+      "learning_rate": 9.713941631482492e-07,
+      "loss": 0.343,
+      "step": 43575
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.260098728332438,
+      "learning_rate": 9.704236009487317e-07,
+      "loss": 0.3443,
+      "step": 43580
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.329658662759802,
+      "learning_rate": 9.694534717449806e-07,
+      "loss": 0.3523,
+      "step": 43585
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.324974830145578,
+      "learning_rate": 9.684837756412436e-07,
+      "loss": 0.3357,
+      "step": 43590
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.2250847620165395,
+      "learning_rate": 9.675145127417162e-07,
+      "loss": 0.3165,
+      "step": 43595
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.629797728082159,
+      "learning_rate": 9.66545683150552e-07,
+      "loss": 0.374,
+      "step": 43600
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.7039192606979325,
+      "learning_rate": 9.655772869718544e-07,
+      "loss": 0.3613,
+      "step": 43605
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.408365979438267,
+      "learning_rate": 9.646093243096833e-07,
+      "loss": 0.3613,
+      "step": 43610
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.898897090792314,
+      "learning_rate": 9.636417952680487e-07,
+      "loss": 0.3187,
+      "step": 43615
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.784863823967556,
+      "learning_rate": 9.626746999509174e-07,
+      "loss": 0.363,
+      "step": 43620
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.4615621138050425,
+      "learning_rate": 9.617080384622085e-07,
+      "loss": 0.3229,
+      "step": 43625
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.79519809907856,
+      "learning_rate": 9.607418109057931e-07,
+      "loss": 0.3611,
+      "step": 43630
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.463654097369577,
+      "learning_rate": 9.59776017385496e-07,
+      "loss": 0.3488,
+      "step": 43635
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.5246864291427515,
+      "learning_rate": 9.588106580050976e-07,
+      "loss": 0.3139,
+      "step": 43640
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.453200078871238,
+      "learning_rate": 9.578457328683276e-07,
+      "loss": 0.3497,
+      "step": 43645
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.539786815358693,
+      "learning_rate": 9.568812420788747e-07,
+      "loss": 0.3647,
+      "step": 43650
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.725255156121965,
+      "learning_rate": 9.55917185740376e-07,
+      "loss": 0.3636,
+      "step": 43655
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.178215742944267,
+      "learning_rate": 9.549535639564223e-07,
+      "loss": 0.3581,
+      "step": 43660
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.845248663914356,
+      "learning_rate": 9.539903768305615e-07,
+      "loss": 0.3925,
+      "step": 43665
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.202138617898354,
+      "learning_rate": 9.530276244662895e-07,
+      "loss": 0.3565,
+      "step": 43670
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.81832249558518,
+      "learning_rate": 9.520653069670615e-07,
+      "loss": 0.3339,
+      "step": 43675
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.623950232561397,
+      "learning_rate": 9.511034244362799e-07,
+      "loss": 0.3553,
+      "step": 43680
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.995389542139731,
+      "learning_rate": 9.501419769773057e-07,
+      "loss": 0.3518,
+      "step": 43685
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.3223302996123385,
+      "learning_rate": 9.491809646934497e-07,
+      "loss": 0.3577,
+      "step": 43690
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.011617296634846,
+      "learning_rate": 9.482203876879748e-07,
+      "loss": 0.322,
+      "step": 43695
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.148526822984768,
+      "learning_rate": 9.472602460641017e-07,
+      "loss": 0.353,
+      "step": 43700
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.801389982432906,
+      "learning_rate": 9.463005399250024e-07,
+      "loss": 0.3303,
+      "step": 43705
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.734765831681855,
+      "learning_rate": 9.45341269373799e-07,
+      "loss": 0.3153,
+      "step": 43710
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.798953351536035,
+      "learning_rate": 9.443824345135721e-07,
+      "loss": 0.3712,
+      "step": 43715
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.577369443166325,
+      "learning_rate": 9.434240354473501e-07,
+      "loss": 0.3678,
+      "step": 43720
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.8039294527983705,
+      "learning_rate": 9.424660722781198e-07,
+      "loss": 0.3272,
+      "step": 43725
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.327270569822577,
+      "learning_rate": 9.415085451088163e-07,
+      "loss": 0.3406,
+      "step": 43730
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.402462600693131,
+      "learning_rate": 9.405514540423322e-07,
+      "loss": 0.3424,
+      "step": 43735
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.453115794429658,
+      "learning_rate": 9.395947991815085e-07,
+      "loss": 0.353,
+      "step": 43740
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.687035856849718,
+      "learning_rate": 9.386385806291453e-07,
+      "loss": 0.3757,
+      "step": 43745
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.541579300240977,
+      "learning_rate": 9.376827984879905e-07,
+      "loss": 0.3504,
+      "step": 43750
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.024964366744509,
+      "learning_rate": 9.367274528607462e-07,
+      "loss": 0.3591,
+      "step": 43755
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.805346917594688,
+      "learning_rate": 9.357725438500714e-07,
+      "loss": 0.356,
+      "step": 43760
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.4208678309441,
+      "learning_rate": 9.348180715585714e-07,
+      "loss": 0.3488,
+      "step": 43765
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.99069494095584,
+      "learning_rate": 9.338640360888124e-07,
+      "loss": 0.336,
+      "step": 43770
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.563176922512066,
+      "learning_rate": 9.32910437543308e-07,
+      "loss": 0.3356,
+      "step": 43775
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.340789279829952,
+      "learning_rate": 9.319572760245255e-07,
+      "loss": 0.3628,
+      "step": 43780
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.775398930130766,
+      "learning_rate": 9.310045516348881e-07,
+      "loss": 0.3846,
+      "step": 43785
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.451004191355735,
+      "learning_rate": 9.300522644767684e-07,
+      "loss": 0.3944,
+      "step": 43790
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.889612762272115,
+      "learning_rate": 9.291004146524945e-07,
+      "loss": 0.3807,
+      "step": 43795
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 17.221536250743267,
+      "learning_rate": 9.281490022643491e-07,
+      "loss": 0.3498,
+      "step": 43800
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.738300720330171,
+      "learning_rate": 9.271980274145625e-07,
+      "loss": 0.3382,
+      "step": 43805
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.644940660433678,
+      "learning_rate": 9.262474902053237e-07,
+      "loss": 0.3232,
+      "step": 43810
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.737625667943532,
+      "learning_rate": 9.252973907387697e-07,
+      "loss": 0.3598,
+      "step": 43815
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 9.977930268125835,
+      "learning_rate": 9.24347729116995e-07,
+      "loss": 0.3683,
+      "step": 43820
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.329929468250905,
+      "learning_rate": 9.233985054420435e-07,
+      "loss": 0.3515,
+      "step": 43825
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.791546485178549,
+      "learning_rate": 9.22449719815915e-07,
+      "loss": 0.3869,
+      "step": 43830
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.678452082940854,
+      "learning_rate": 9.215013723405591e-07,
+      "loss": 0.3919,
+      "step": 43835
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.55156797863798,
+      "learning_rate": 9.205534631178798e-07,
+      "loss": 0.3425,
+      "step": 43840
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.507942225396343,
+      "learning_rate": 9.196059922497363e-07,
+      "loss": 0.3336,
+      "step": 43845
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 9.686081776468555,
+      "learning_rate": 9.186589598379353e-07,
+      "loss": 0.3502,
+      "step": 43850
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 8.422982866386748,
+      "learning_rate": 9.17712365984243e-07,
+      "loss": 0.3744,
+      "step": 43855
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.827116610585081,
+      "learning_rate": 9.167662107903735e-07,
+      "loss": 0.3364,
+      "step": 43860
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.033505566965495,
+      "learning_rate": 9.15820494357994e-07,
+      "loss": 0.3962,
+      "step": 43865
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.539734137289886,
+      "learning_rate": 9.148752167887287e-07,
+      "loss": 0.3561,
+      "step": 43870
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.110277854754629,
+      "learning_rate": 9.139303781841486e-07,
+      "loss": 0.3245,
+      "step": 43875
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.7338676785625475,
+      "learning_rate": 9.12985978645784e-07,
+      "loss": 0.3622,
+      "step": 43880
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.4537484441987,
+      "learning_rate": 9.120420182751128e-07,
+      "loss": 0.3496,
+      "step": 43885
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.760842851071135,
+      "learning_rate": 9.110984971735675e-07,
+      "loss": 0.3346,
+      "step": 43890
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.29210792447318,
+      "learning_rate": 9.101554154425357e-07,
+      "loss": 0.3356,
+      "step": 43895
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.277067057868431,
+      "learning_rate": 9.092127731833533e-07,
+      "loss": 0.3405,
+      "step": 43900
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.983623319622154,
+      "learning_rate": 9.082705704973138e-07,
+      "loss": 0.3732,
+      "step": 43905
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.998600476909276,
+      "learning_rate": 9.073288074856584e-07,
+      "loss": 0.3385,
+      "step": 43910
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.397915903607824,
+      "learning_rate": 9.063874842495857e-07,
+      "loss": 0.3521,
+      "step": 43915
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.399127011999149,
+      "learning_rate": 9.054466008902452e-07,
+      "loss": 0.3443,
+      "step": 43920
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.076140472896828,
+      "learning_rate": 9.04506157508736e-07,
+      "loss": 0.3478,
+      "step": 43925
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.771040874518689,
+      "learning_rate": 9.035661542061164e-07,
+      "loss": 0.3282,
+      "step": 43930
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.764495449884524,
+      "learning_rate": 9.026265910833915e-07,
+      "loss": 0.3307,
+      "step": 43935
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.894289259401423,
+      "learning_rate": 9.016874682415239e-07,
+      "loss": 0.343,
+      "step": 43940
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.6520390140199215,
+      "learning_rate": 9.007487857814246e-07,
+      "loss": 0.3389,
+      "step": 43945
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.605151237190835,
+      "learning_rate": 8.998105438039584e-07,
+      "loss": 0.3908,
+      "step": 43950
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.787140948496046,
+      "learning_rate": 8.988727424099464e-07,
+      "loss": 0.3341,
+      "step": 43955
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.053377741736455,
+      "learning_rate": 8.979353817001563e-07,
+      "loss": 0.369,
+      "step": 43960
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.875304711403996,
+      "learning_rate": 8.969984617753146e-07,
+      "loss": 0.3375,
+      "step": 43965
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.005540197045185,
+      "learning_rate": 8.960619827360951e-07,
+      "loss": 0.3607,
+      "step": 43970
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.7801687630338945,
+      "learning_rate": 8.951259446831284e-07,
+      "loss": 0.3476,
+      "step": 43975
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.530451062249858,
+      "learning_rate": 8.941903477169944e-07,
+      "loss": 0.3494,
+      "step": 43980
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.664961379976749,
+      "learning_rate": 8.932551919382277e-07,
+      "loss": 0.3471,
+      "step": 43985
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.196541035106309,
+      "learning_rate": 8.92320477447316e-07,
+      "loss": 0.3411,
+      "step": 43990
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.69161782574326,
+      "learning_rate": 8.913862043446963e-07,
+      "loss": 0.3694,
+      "step": 43995
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.813935139318951,
+      "learning_rate": 8.904523727307635e-07,
+      "loss": 0.3726,
+      "step": 44000
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.449693510106746,
+      "learning_rate": 8.895189827058598e-07,
+      "loss": 0.3252,
+      "step": 44005
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.144595709635144,
+      "learning_rate": 8.885860343702807e-07,
+      "loss": 0.3758,
+      "step": 44010
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.499951839589321,
+      "learning_rate": 8.876535278242788e-07,
+      "loss": 0.369,
+      "step": 44015
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.64946762625513,
+      "learning_rate": 8.867214631680532e-07,
+      "loss": 0.3473,
+      "step": 44020
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.440712374897632,
+      "learning_rate": 8.857898405017612e-07,
+      "loss": 0.36,
+      "step": 44025
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.373268530531762,
+      "learning_rate": 8.848586599255076e-07,
+      "loss": 0.3513,
+      "step": 44030
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.732680096831105,
+      "learning_rate": 8.83927921539352e-07,
+      "loss": 0.3456,
+      "step": 44035
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.303962304466678,
+      "learning_rate": 8.829976254433076e-07,
+      "loss": 0.3408,
+      "step": 44040
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.901822957771553,
+      "learning_rate": 8.820677717373372e-07,
+      "loss": 0.366,
+      "step": 44045
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.761613390083313,
+      "learning_rate": 8.811383605213591e-07,
+      "loss": 0.3643,
+      "step": 44050
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.090882910730315,
+      "learning_rate": 8.802093918952426e-07,
+      "loss": 0.3384,
+      "step": 44055
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.560597953357254,
+      "learning_rate": 8.792808659588076e-07,
+      "loss": 0.358,
+      "step": 44060
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 8.106390771733917,
+      "learning_rate": 8.783527828118299e-07,
+      "loss": 0.3393,
+      "step": 44065
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.213727936338028,
+      "learning_rate": 8.774251425540369e-07,
+      "loss": 0.3574,
+      "step": 44070
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.538798588504677,
+      "learning_rate": 8.764979452851058e-07,
+      "loss": 0.3298,
+      "step": 44075
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.664709588768496,
+      "learning_rate": 8.755711911046682e-07,
+      "loss": 0.3632,
+      "step": 44080
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.620657667558471,
+      "learning_rate": 8.746448801123103e-07,
+      "loss": 0.3218,
+      "step": 44085
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.057409884261421,
+      "learning_rate": 8.737190124075668e-07,
+      "loss": 0.3441,
+      "step": 44090
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.105871557587499,
+      "learning_rate": 8.72793588089924e-07,
+      "loss": 0.3323,
+      "step": 44095
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.995818818511657,
+      "learning_rate": 8.718686072588272e-07,
+      "loss": 0.3182,
+      "step": 44100
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.697528507573346,
+      "learning_rate": 8.709440700136657e-07,
+      "loss": 0.3698,
+      "step": 44105
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.827599048255153,
+      "learning_rate": 8.700199764537876e-07,
+      "loss": 0.3809,
+      "step": 44110
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.558622120621011,
+      "learning_rate": 8.690963266784907e-07,
+      "loss": 0.3368,
+      "step": 44115
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.491501574733111,
+      "learning_rate": 8.681731207870231e-07,
+      "loss": 0.3596,
+      "step": 44120
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.766531291782006,
+      "learning_rate": 8.672503588785897e-07,
+      "loss": 0.3342,
+      "step": 44125
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.626581899254411,
+      "learning_rate": 8.663280410523434e-07,
+      "loss": 0.3654,
+      "step": 44130
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.2621047605774915,
+      "learning_rate": 8.654061674073938e-07,
+      "loss": 0.3356,
+      "step": 44135
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.749185535655736,
+      "learning_rate": 8.644847380427989e-07,
+      "loss": 0.3313,
+      "step": 44140
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.8056168870327065,
+      "learning_rate": 8.635637530575691e-07,
+      "loss": 0.352,
+      "step": 44145
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.676457696084985,
+      "learning_rate": 8.626432125506706e-07,
+      "loss": 0.3579,
+      "step": 44150
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.7567903095705235,
+      "learning_rate": 8.617231166210172e-07,
+      "loss": 0.3419,
+      "step": 44155
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.889250251464815,
+      "learning_rate": 8.608034653674785e-07,
+      "loss": 0.338,
+      "step": 44160
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.603592082293575,
+      "learning_rate": 8.598842588888761e-07,
+      "loss": 0.3331,
+      "step": 44165
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.202106032016938,
+      "learning_rate": 8.589654972839812e-07,
+      "loss": 0.3673,
+      "step": 44170
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.103395208552594,
+      "learning_rate": 8.580471806515201e-07,
+      "loss": 0.3186,
+      "step": 44175
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.630736793045195,
+      "learning_rate": 8.571293090901684e-07,
+      "loss": 0.3619,
+      "step": 44180
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.541640109764103,
+      "learning_rate": 8.562118826985571e-07,
+      "loss": 0.3229,
+      "step": 44185
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.441121404906713,
+      "learning_rate": 8.552949015752665e-07,
+      "loss": 0.3524,
+      "step": 44190
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.957926598428815,
+      "learning_rate": 8.543783658188315e-07,
+      "loss": 0.3456,
+      "step": 44195
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.335333781799094,
+      "learning_rate": 8.534622755277366e-07,
+      "loss": 0.3566,
+      "step": 44200
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.55555271440869,
+      "learning_rate": 8.525466308004199e-07,
+      "loss": 0.3391,
+      "step": 44205
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.929445293932128,
+      "learning_rate": 8.516314317352725e-07,
+      "loss": 0.3436,
+      "step": 44210
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.190184008181839,
+      "learning_rate": 8.507166784306347e-07,
+      "loss": 0.3461,
+      "step": 44215
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.66077158709162,
+      "learning_rate": 8.498023709848035e-07,
+      "loss": 0.3665,
+      "step": 44220
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.250552212017445,
+      "learning_rate": 8.488885094960226e-07,
+      "loss": 0.3729,
+      "step": 44225
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.953751773413122,
+      "learning_rate": 8.479750940624925e-07,
+      "loss": 0.3484,
+      "step": 44230
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.580191792202555,
+      "learning_rate": 8.47062124782363e-07,
+      "loss": 0.3594,
+      "step": 44235
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.372847104021162,
+      "learning_rate": 8.461496017537352e-07,
+      "loss": 0.3262,
+      "step": 44240
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.4175859248973435,
+      "learning_rate": 8.452375250746658e-07,
+      "loss": 0.3549,
+      "step": 44245
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.072502246571446,
+      "learning_rate": 8.443258948431604e-07,
+      "loss": 0.3317,
+      "step": 44250
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.925112630434505,
+      "learning_rate": 8.434147111571772e-07,
+      "loss": 0.3783,
+      "step": 44255
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.795085259124459,
+      "learning_rate": 8.425039741146296e-07,
+      "loss": 0.345,
+      "step": 44260
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.693133935750688,
+      "learning_rate": 8.415936838133765e-07,
+      "loss": 0.3455,
+      "step": 44265
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.667041642741462,
+      "learning_rate": 8.406838403512363e-07,
+      "loss": 0.3606,
+      "step": 44270
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.55027899950803,
+      "learning_rate": 8.397744438259725e-07,
+      "loss": 0.3062,
+      "step": 44275
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.587479111705073,
+      "learning_rate": 8.388654943353064e-07,
+      "loss": 0.3793,
+      "step": 44280
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.4443166706002355,
+      "learning_rate": 8.37956991976907e-07,
+      "loss": 0.3528,
+      "step": 44285
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.750046861597566,
+      "learning_rate": 8.370489368483969e-07,
+      "loss": 0.3295,
+      "step": 44290
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.2278896348195545,
+      "learning_rate": 8.361413290473519e-07,
+      "loss": 0.3173,
+      "step": 44295
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.352502786694757,
+      "learning_rate": 8.352341686712962e-07,
+      "loss": 0.3529,
+      "step": 44300
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.331734081914186,
+      "learning_rate": 8.343274558177105e-07,
+      "loss": 0.3595,
+      "step": 44305
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.723387677421115,
+      "learning_rate": 8.33421190584024e-07,
+      "loss": 0.3273,
+      "step": 44310
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.64324574200726,
+      "learning_rate": 8.325153730676194e-07,
+      "loss": 0.3286,
+      "step": 44315
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.404223866217227,
+      "learning_rate": 8.316100033658309e-07,
+      "loss": 0.3163,
+      "step": 44320
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.65372281204677,
+      "learning_rate": 8.307050815759426e-07,
+      "loss": 0.3414,
+      "step": 44325
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.729993214905292,
+      "learning_rate": 8.298006077951953e-07,
+      "loss": 0.359,
+      "step": 44330
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.269714409863494,
+      "learning_rate": 8.288965821207756e-07,
+      "loss": 0.3457,
+      "step": 44335
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.542042151469218,
+      "learning_rate": 8.27993004649828e-07,
+      "loss": 0.3673,
+      "step": 44340
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.7856233782340345,
+      "learning_rate": 8.270898754794432e-07,
+      "loss": 0.3841,
+      "step": 44345
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.061868233838201,
+      "learning_rate": 8.26187194706668e-07,
+      "loss": 0.3367,
+      "step": 44350
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.294338768524923,
+      "learning_rate": 8.252849624285004e-07,
+      "loss": 0.3368,
+      "step": 44355
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.725613968867999,
+      "learning_rate": 8.243831787418866e-07,
+      "loss": 0.339,
+      "step": 44360
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.93660765090371,
+      "learning_rate": 8.234818437437298e-07,
+      "loss": 0.3973,
+      "step": 44365
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.914833726032178,
+      "learning_rate": 8.225809575308818e-07,
+      "loss": 0.3575,
+      "step": 44370
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.60223427380691,
+      "learning_rate": 8.216805202001443e-07,
+      "loss": 0.3312,
+      "step": 44375
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.865653198684339,
+      "learning_rate": 8.20780531848277e-07,
+      "loss": 0.3528,
+      "step": 44380
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.373911080635319,
+      "learning_rate": 8.198809925719841e-07,
+      "loss": 0.378,
+      "step": 44385
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.762112520158262,
+      "learning_rate": 8.189819024679285e-07,
+      "loss": 0.3459,
+      "step": 44390
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.567877158922223,
+      "learning_rate": 8.180832616327184e-07,
+      "loss": 0.3714,
+      "step": 44395
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.805688174382899,
+      "learning_rate": 8.171850701629191e-07,
+      "loss": 0.318,
+      "step": 44400
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.475666983940325,
+      "learning_rate": 8.162873281550448e-07,
+      "loss": 0.337,
+      "step": 44405
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.543551938126569,
+      "learning_rate": 8.153900357055594e-07,
+      "loss": 0.3198,
+      "step": 44410
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.541585476119786,
+      "learning_rate": 8.144931929108841e-07,
+      "loss": 0.3159,
+      "step": 44415
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.953033082317601,
+      "learning_rate": 8.13596799867386e-07,
+      "loss": 0.3171,
+      "step": 44420
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.845871126148948,
+      "learning_rate": 8.12700856671389e-07,
+      "loss": 0.3366,
+      "step": 44425
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.487990737133694,
+      "learning_rate": 8.118053634191637e-07,
+      "loss": 0.3813,
+      "step": 44430
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.829448555182738,
+      "learning_rate": 8.109103202069373e-07,
+      "loss": 0.3039,
+      "step": 44435
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.185810975738731,
+      "learning_rate": 8.100157271308834e-07,
+      "loss": 0.3346,
+      "step": 44440
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.312857772746139,
+      "learning_rate": 8.09121584287132e-07,
+      "loss": 0.3331,
+      "step": 44445
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.0701202927130815,
+      "learning_rate": 8.082278917717629e-07,
+      "loss": 0.3664,
+      "step": 44450
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.357899393176421,
+      "learning_rate": 8.073346496808065e-07,
+      "loss": 0.3089,
+      "step": 44455
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.1138905776558925,
+      "learning_rate": 8.064418581102445e-07,
+      "loss": 0.3335,
+      "step": 44460
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.849758504003736,
+      "learning_rate": 8.055495171560135e-07,
+      "loss": 0.3541,
+      "step": 44465
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.7556108709580585,
+      "learning_rate": 8.046576269139977e-07,
+      "loss": 0.3219,
+      "step": 44470
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.328591524710588,
+      "learning_rate": 8.037661874800362e-07,
+      "loss": 0.3348,
+      "step": 44475
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.198723147603352,
+      "learning_rate": 8.028751989499162e-07,
+      "loss": 0.328,
+      "step": 44480
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.840623655667464,
+      "learning_rate": 8.019846614193799e-07,
+      "loss": 0.3743,
+      "step": 44485
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.873340058137853,
+      "learning_rate": 8.010945749841198e-07,
+      "loss": 0.3417,
+      "step": 44490
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.588646299405075,
+      "learning_rate": 8.002049397397776e-07,
+      "loss": 0.3209,
+      "step": 44495
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.345110140095459,
+      "learning_rate": 7.993157557819503e-07,
+      "loss": 0.3321,
+      "step": 44500
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.517240457719273,
+      "learning_rate": 7.984270232061836e-07,
+      "loss": 0.3311,
+      "step": 44505
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.426640436410661,
+      "learning_rate": 7.975387421079767e-07,
+      "loss": 0.3442,
+      "step": 44510
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.566775949798058,
+      "learning_rate": 7.966509125827793e-07,
+      "loss": 0.3243,
+      "step": 44515
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.8314126750494415,
+      "learning_rate": 7.957635347259907e-07,
+      "loss": 0.3396,
+      "step": 44520
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.074945417485195,
+      "learning_rate": 7.948766086329651e-07,
+      "loss": 0.3391,
+      "step": 44525
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 6.219482845702791,
+      "learning_rate": 7.93990134399008e-07,
+      "loss": 0.3647,
+      "step": 44530
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.562744980177607,
+      "learning_rate": 7.931041121193722e-07,
+      "loss": 0.3357,
+      "step": 44535
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.787700461660074,
+      "learning_rate": 7.922185418892669e-07,
+      "loss": 0.3676,
+      "step": 44540
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.402010946003515,
+      "learning_rate": 7.913334238038484e-07,
+      "loss": 0.3563,
+      "step": 44545
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.21840161300093,
+      "learning_rate": 7.904487579582292e-07,
+      "loss": 0.3512,
+      "step": 44550
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.4085572660437915,
+      "learning_rate": 7.895645444474681e-07,
+      "loss": 0.3544,
+      "step": 44555
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.494106842057671,
+      "learning_rate": 7.886807833665799e-07,
+      "loss": 0.3337,
+      "step": 44560
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.7858159887065765,
+      "learning_rate": 7.877974748105261e-07,
+      "loss": 0.3295,
+      "step": 44565
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.846625073743843,
+      "learning_rate": 7.869146188742244e-07,
+      "loss": 0.3348,
+      "step": 44570
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.9417794468001235,
+      "learning_rate": 7.860322156525413e-07,
+      "loss": 0.3383,
+      "step": 44575
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.562123002643166,
+      "learning_rate": 7.851502652402926e-07,
+      "loss": 0.331,
+      "step": 44580
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.828027079116812,
+      "learning_rate": 7.842687677322508e-07,
+      "loss": 0.3243,
+      "step": 44585
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.518395277358132,
+      "learning_rate": 7.833877232231347e-07,
+      "loss": 0.3342,
+      "step": 44590
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.133451535257724,
+      "learning_rate": 7.825071318076177e-07,
+      "loss": 0.3567,
+      "step": 44595
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.636806447381211,
+      "learning_rate": 7.816269935803222e-07,
+      "loss": 0.3658,
+      "step": 44600
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.569085031167726,
+      "learning_rate": 7.80747308635823e-07,
+      "loss": 0.3542,
+      "step": 44605
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.529129237943756,
+      "learning_rate": 7.798680770686473e-07,
+      "loss": 0.3382,
+      "step": 44610
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.081965625152124,
+      "learning_rate": 7.7898929897327e-07,
+      "loss": 0.347,
+      "step": 44615
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.3032046904492685,
+      "learning_rate": 7.781109744441217e-07,
+      "loss": 0.3072,
+      "step": 44620
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.243785155354974,
+      "learning_rate": 7.772331035755831e-07,
+      "loss": 0.347,
+      "step": 44625
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.241055760001078,
+      "learning_rate": 7.763556864619825e-07,
+      "loss": 0.3502,
+      "step": 44630
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.356448601790181,
+      "learning_rate": 7.754787231976052e-07,
+      "loss": 0.3243,
+      "step": 44635
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.031269602851849,
+      "learning_rate": 7.746022138766823e-07,
+      "loss": 0.3346,
+      "step": 44640
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.53741830317873,
+      "learning_rate": 7.737261585934003e-07,
+      "loss": 0.3324,
+      "step": 44645
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.628902859046572,
+      "learning_rate": 7.728505574418938e-07,
+      "loss": 0.3618,
+      "step": 44650
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.442259678807115,
+      "learning_rate": 7.719754105162514e-07,
+      "loss": 0.3546,
+      "step": 44655
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.405580851793677,
+      "learning_rate": 7.711007179105113e-07,
+      "loss": 0.367,
+      "step": 44660
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.821365305669165,
+      "learning_rate": 7.702264797186615e-07,
+      "loss": 0.3957,
+      "step": 44665
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 6.29747534519174,
+      "learning_rate": 7.693526960346448e-07,
+      "loss": 0.3648,
+      "step": 44670
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.983836888656855,
+      "learning_rate": 7.68479366952351e-07,
+      "loss": 0.3618,
+      "step": 44675
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.943605713343998,
+      "learning_rate": 7.676064925656257e-07,
+      "loss": 0.33,
+      "step": 44680
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.679080666542682,
+      "learning_rate": 7.667340729682615e-07,
+      "loss": 0.3202,
+      "step": 44685
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.156348424319493,
+      "learning_rate": 7.65862108254003e-07,
+      "loss": 0.3289,
+      "step": 44690
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.297079294747502,
+      "learning_rate": 7.649905985165484e-07,
+      "loss": 0.3693,
+      "step": 44695
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.651999008907444,
+      "learning_rate": 7.641195438495435e-07,
+      "loss": 0.3369,
+      "step": 44700
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.797409387390218,
+      "learning_rate": 7.632489443465896e-07,
+      "loss": 0.3401,
+      "step": 44705
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.589892122850783,
+      "learning_rate": 7.623788001012333e-07,
+      "loss": 0.3656,
+      "step": 44710
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.76219234641212,
+      "learning_rate": 7.615091112069772e-07,
+      "loss": 0.3618,
+      "step": 44715
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.428078517908885,
+      "learning_rate": 7.606398777572738e-07,
+      "loss": 0.3447,
+      "step": 44720
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.438532826939408,
+      "learning_rate": 7.597710998455244e-07,
+      "loss": 0.3425,
+      "step": 44725
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.878997266613701,
+      "learning_rate": 7.589027775650854e-07,
+      "loss": 0.3581,
+      "step": 44730
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.867534225517133,
+      "learning_rate": 7.580349110092583e-07,
+      "loss": 0.3405,
+      "step": 44735
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.39431457738843,
+      "learning_rate": 7.57167500271303e-07,
+      "loss": 0.3456,
+      "step": 44740
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.483054982849157,
+      "learning_rate": 7.563005454444245e-07,
+      "loss": 0.3244,
+      "step": 44745
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.332740413440662,
+      "learning_rate": 7.554340466217803e-07,
+      "loss": 0.2978,
+      "step": 44750
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.576924334818634,
+      "learning_rate": 7.54568003896482e-07,
+      "loss": 0.3411,
+      "step": 44755
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.362288551470274,
+      "learning_rate": 7.537024173615865e-07,
+      "loss": 0.3488,
+      "step": 44760
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.541460978341073,
+      "learning_rate": 7.528372871101081e-07,
+      "loss": 0.3459,
+      "step": 44765
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.60757482332367,
+      "learning_rate": 7.519726132350069e-07,
+      "loss": 0.3405,
+      "step": 44770
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.100913946874496,
+      "learning_rate": 7.511083958291953e-07,
+      "loss": 0.3524,
+      "step": 44775
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.476606384585978,
+      "learning_rate": 7.502446349855403e-07,
+      "loss": 0.3519,
+      "step": 44780
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.772859445135238,
+      "learning_rate": 7.493813307968528e-07,
+      "loss": 0.3498,
+      "step": 44785
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.583888825155642,
+      "learning_rate": 7.485184833559023e-07,
+      "loss": 0.3384,
+      "step": 44790
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.248775581687325,
+      "learning_rate": 7.476560927554027e-07,
+      "loss": 0.341,
+      "step": 44795
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.481589917895012,
+      "learning_rate": 7.46794159088024e-07,
+      "loss": 0.3449,
+      "step": 44800
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.139578821681772,
+      "learning_rate": 7.459326824463831e-07,
+      "loss": 0.3481,
+      "step": 44805
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.941603480871794,
+      "learning_rate": 7.450716629230498e-07,
+      "loss": 0.3823,
+      "step": 44810
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.373478099554834,
+      "learning_rate": 7.442111006105457e-07,
+      "loss": 0.3309,
+      "step": 44815
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.671249747303654,
+      "learning_rate": 7.433509956013402e-07,
+      "loss": 0.3588,
+      "step": 44820
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.264948532509474,
+      "learning_rate": 7.424913479878571e-07,
+      "loss": 0.3662,
+      "step": 44825
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.885032567916461,
+      "learning_rate": 7.416321578624685e-07,
+      "loss": 0.3448,
+      "step": 44830
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.939823260071416,
+      "learning_rate": 7.40773425317497e-07,
+      "loss": 0.334,
+      "step": 44835
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.626916438486954,
+      "learning_rate": 7.399151504452201e-07,
+      "loss": 0.3181,
+      "step": 44840
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.4233655638848886,
+      "learning_rate": 7.390573333378598e-07,
+      "loss": 0.3818,
+      "step": 44845
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.402375823574998,
+      "learning_rate": 7.381999740875951e-07,
+      "loss": 0.355,
+      "step": 44850
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.271480811664558,
+      "learning_rate": 7.373430727865521e-07,
+      "loss": 0.3444,
+      "step": 44855
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.992301004960722,
+      "learning_rate": 7.364866295268069e-07,
+      "loss": 0.3463,
+      "step": 44860
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.596595252823135,
+      "learning_rate": 7.356306444003913e-07,
+      "loss": 0.3157,
+      "step": 44865
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.38779584756225,
+      "learning_rate": 7.347751174992818e-07,
+      "loss": 0.3479,
+      "step": 44870
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.808865570577694,
+      "learning_rate": 7.339200489154113e-07,
+      "loss": 0.3121,
+      "step": 44875
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.6246105385618295,
+      "learning_rate": 7.330654387406577e-07,
+      "loss": 0.3295,
+      "step": 44880
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.800382089908357,
+      "learning_rate": 7.322112870668552e-07,
+      "loss": 0.3592,
+      "step": 44885
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.050691840892069,
+      "learning_rate": 7.313575939857843e-07,
+      "loss": 0.3398,
+      "step": 44890
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.5763900688178385,
+      "learning_rate": 7.305043595891793e-07,
+      "loss": 0.307,
+      "step": 44895
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.690380999095491,
+      "learning_rate": 7.296515839687229e-07,
+      "loss": 0.362,
+      "step": 44900
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.106088638126591,
+      "learning_rate": 7.287992672160505e-07,
+      "loss": 0.3437,
+      "step": 44905
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.326280089942625,
+      "learning_rate": 7.279474094227479e-07,
+      "loss": 0.3167,
+      "step": 44910
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.105439104678032,
+      "learning_rate": 7.270960106803504e-07,
+      "loss": 0.3561,
+      "step": 44915
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.626778222128296,
+      "learning_rate": 7.262450710803426e-07,
+      "loss": 0.3346,
+      "step": 44920
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.587701722220041,
+      "learning_rate": 7.253945907141652e-07,
+      "loss": 0.3488,
+      "step": 44925
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.6224501817734716,
+      "learning_rate": 7.245445696732028e-07,
+      "loss": 0.3586,
+      "step": 44930
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.42029361158793,
+      "learning_rate": 7.23695008048797e-07,
+      "loss": 0.338,
+      "step": 44935
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.518060063351799,
+      "learning_rate": 7.228459059322351e-07,
+      "loss": 0.32,
+      "step": 44940
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.28541057087806,
+      "learning_rate": 7.21997263414756e-07,
+      "loss": 0.3068,
+      "step": 44945
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.693612685635041,
+      "learning_rate": 7.21149080587552e-07,
+      "loss": 0.3371,
+      "step": 44950
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.465311875366154,
+      "learning_rate": 7.203013575417628e-07,
+      "loss": 0.3203,
+      "step": 44955
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.811512264611589,
+      "learning_rate": 7.194540943684813e-07,
+      "loss": 0.3669,
+      "step": 44960
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.340615069375246,
+      "learning_rate": 7.186072911587477e-07,
+      "loss": 0.3521,
+      "step": 44965
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.889031936422385,
+      "learning_rate": 7.177609480035568e-07,
+      "loss": 0.3477,
+      "step": 44970
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.927139064168435,
+      "learning_rate": 7.169150649938517e-07,
+      "loss": 0.3436,
+      "step": 44975
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.752788813177939,
+      "learning_rate": 7.160696422205238e-07,
+      "loss": 0.3466,
+      "step": 44980
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.980336606767357,
+      "learning_rate": 7.152246797744195e-07,
+      "loss": 0.3097,
+      "step": 44985
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.27047234839068,
+      "learning_rate": 7.143801777463344e-07,
+      "loss": 0.3529,
+      "step": 44990
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.639829298583398,
+      "learning_rate": 7.135361362270122e-07,
+      "loss": 0.38,
+      "step": 44995
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.845379526215203,
+      "learning_rate": 7.126925553071506e-07,
+      "loss": 0.356,
+      "step": 45000
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.376427871005003,
+      "learning_rate": 7.118494350773946e-07,
+      "loss": 0.3331,
+      "step": 45005
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.7883684348232185,
+      "learning_rate": 7.110067756283429e-07,
+      "loss": 0.3592,
+      "step": 45010
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.481909323704549,
+      "learning_rate": 7.101645770505406e-07,
+      "loss": 0.3049,
+      "step": 45015
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.777910108906637,
+      "learning_rate": 7.093228394344881e-07,
+      "loss": 0.3344,
+      "step": 45020
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.281601427551611,
+      "learning_rate": 7.084815628706327e-07,
+      "loss": 0.3446,
+      "step": 45025
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.5747065573831325,
+      "learning_rate": 7.076407474493724e-07,
+      "loss": 0.3424,
+      "step": 45030
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.031091193354879,
+      "learning_rate": 7.068003932610584e-07,
+      "loss": 0.3299,
+      "step": 45035
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.612869879558003,
+      "learning_rate": 7.05960500395988e-07,
+      "loss": 0.3158,
+      "step": 45040
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.979012364866844,
+      "learning_rate": 7.051210689444143e-07,
+      "loss": 0.3145,
+      "step": 45045
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.7305499080220414,
+      "learning_rate": 7.042820989965355e-07,
+      "loss": 0.3421,
+      "step": 45050
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.739869974815494,
+      "learning_rate": 7.034435906425047e-07,
+      "loss": 0.3426,
+      "step": 45055
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.406403626184369,
+      "learning_rate": 7.026055439724222e-07,
+      "loss": 0.3588,
+      "step": 45060
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.979411572308106,
+      "learning_rate": 7.017679590763388e-07,
+      "loss": 0.3183,
+      "step": 45065
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.61023269706424,
+      "learning_rate": 7.009308360442585e-07,
+      "loss": 0.336,
+      "step": 45070
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.518898134950466,
+      "learning_rate": 7.000941749661322e-07,
+      "loss": 0.3602,
+      "step": 45075
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.958979467630164,
+      "learning_rate": 6.992579759318641e-07,
+      "loss": 0.3445,
+      "step": 45080
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.632274520297883,
+      "learning_rate": 6.984222390313078e-07,
+      "loss": 0.3817,
+      "step": 45085
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.139154209396844,
+      "learning_rate": 6.975869643542649e-07,
+      "loss": 0.3878,
+      "step": 45090
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.87928049637677,
+      "learning_rate": 6.967521519904924e-07,
+      "loss": 0.3593,
+      "step": 45095
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.675614618498371,
+      "learning_rate": 6.959178020296914e-07,
+      "loss": 0.332,
+      "step": 45100
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.034894840758799,
+      "learning_rate": 6.95083914561519e-07,
+      "loss": 0.3697,
+      "step": 45105
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.402565703964291,
+      "learning_rate": 6.942504896755781e-07,
+      "loss": 0.3291,
+      "step": 45110
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.702696547043211,
+      "learning_rate": 6.934175274614258e-07,
+      "loss": 0.3273,
+      "step": 45115
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.606186573893503,
+      "learning_rate": 6.925850280085666e-07,
+      "loss": 0.3238,
+      "step": 45120
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.794160241758433,
+      "learning_rate": 6.917529914064547e-07,
+      "loss": 0.313,
+      "step": 45125
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.9532447284861325,
+      "learning_rate": 6.909214177444995e-07,
+      "loss": 0.3491,
+      "step": 45130
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.643774767180947,
+      "learning_rate": 6.900903071120535e-07,
+      "loss": 0.2958,
+      "step": 45135
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.963371489528325,
+      "learning_rate": 6.892596595984269e-07,
+      "loss": 0.3375,
+      "step": 45140
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.498468147812131,
+      "learning_rate": 6.884294752928738e-07,
+      "loss": 0.3285,
+      "step": 45145
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.9257190081078335,
+      "learning_rate": 6.87599754284602e-07,
+      "loss": 0.3377,
+      "step": 45150
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.498426066136765,
+      "learning_rate": 6.867704966627686e-07,
+      "loss": 0.3332,
+      "step": 45155
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.1564111238574,
+      "learning_rate": 6.859417025164805e-07,
+      "loss": 0.3594,
+      "step": 45160
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.31933813942458,
+      "learning_rate": 6.851133719347974e-07,
+      "loss": 0.3317,
+      "step": 45165
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.825389142389356,
+      "learning_rate": 6.842855050067237e-07,
+      "loss": 0.351,
+      "step": 45170
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.154041907647826,
+      "learning_rate": 6.834581018212195e-07,
+      "loss": 0.3359,
+      "step": 45175
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.944834533182103,
+      "learning_rate": 6.826311624671938e-07,
+      "loss": 0.3291,
+      "step": 45180
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.1692637406001305,
+      "learning_rate": 6.818046870335026e-07,
+      "loss": 0.3538,
+      "step": 45185
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.310571888639479,
+      "learning_rate": 6.809786756089565e-07,
+      "loss": 0.3653,
+      "step": 45190
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.149389655624969,
+      "learning_rate": 6.801531282823115e-07,
+      "loss": 0.3619,
+      "step": 45195
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 8.066637430098174,
+      "learning_rate": 6.793280451422796e-07,
+      "loss": 0.3104,
+      "step": 45200
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.639527242238606,
+      "learning_rate": 6.785034262775175e-07,
+      "loss": 0.3288,
+      "step": 45205
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.760593620632931,
+      "learning_rate": 6.776792717766334e-07,
+      "loss": 0.3352,
+      "step": 45210
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.530244222670664,
+      "learning_rate": 6.76855581728188e-07,
+      "loss": 0.3404,
+      "step": 45215
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.357855873738075,
+      "learning_rate": 6.760323562206894e-07,
+      "loss": 0.3138,
+      "step": 45220
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.766820495803968,
+      "learning_rate": 6.75209595342598e-07,
+      "loss": 0.3868,
+      "step": 45225
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.939430013859291,
+      "learning_rate": 6.743872991823225e-07,
+      "loss": 0.3612,
+      "step": 45230
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.598328199609213,
+      "learning_rate": 6.735654678282211e-07,
+      "loss": 0.3406,
+      "step": 45235
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.021533912883923,
+      "learning_rate": 6.727441013686048e-07,
+      "loss": 0.3353,
+      "step": 45240
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.702531077882601,
+      "learning_rate": 6.719231998917319e-07,
+      "loss": 0.3375,
+      "step": 45245
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.807710449083324,
+      "learning_rate": 6.711027634858131e-07,
+      "loss": 0.3356,
+      "step": 45250
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.393230274046004,
+      "learning_rate": 6.702827922390064e-07,
+      "loss": 0.3259,
+      "step": 45255
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.621976352734638,
+      "learning_rate": 6.694632862394235e-07,
+      "loss": 0.3388,
+      "step": 45260
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.227356045305042,
+      "learning_rate": 6.686442455751213e-07,
+      "loss": 0.3844,
+      "step": 45265
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.613606798064376,
+      "learning_rate": 6.678256703341102e-07,
+      "loss": 0.3353,
+      "step": 45270
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.913645735562954,
+      "learning_rate": 6.670075606043519e-07,
+      "loss": 0.3289,
+      "step": 45275
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.129590843327389,
+      "learning_rate": 6.661899164737529e-07,
+      "loss": 0.3203,
+      "step": 45280
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.261275238586701,
+      "learning_rate": 6.65372738030175e-07,
+      "loss": 0.3607,
+      "step": 45285
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.473467542034233,
+      "learning_rate": 6.645560253614264e-07,
+      "loss": 0.3347,
+      "step": 45290
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.880070134310221,
+      "learning_rate": 6.637397785552658e-07,
+      "loss": 0.3564,
+      "step": 45295
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.4688638540940016,
+      "learning_rate": 6.62923997699404e-07,
+      "loss": 0.3515,
+      "step": 45300
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.005617107071605,
+      "learning_rate": 6.621086828814987e-07,
+      "loss": 0.354,
+      "step": 45305
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.13218954265486,
+      "learning_rate": 6.612938341891617e-07,
+      "loss": 0.3215,
+      "step": 45310
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.595094142958957,
+      "learning_rate": 6.604794517099499e-07,
+      "loss": 0.3391,
+      "step": 45315
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.22501493455035,
+      "learning_rate": 6.59665535531372e-07,
+      "loss": 0.3268,
+      "step": 45320
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.130955010614197,
+      "learning_rate": 6.588520857408881e-07,
+      "loss": 0.3543,
+      "step": 45325
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.501903308732604,
+      "learning_rate": 6.580391024259058e-07,
+      "loss": 0.3237,
+      "step": 45330
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.741030478221836,
+      "learning_rate": 6.572265856737858e-07,
+      "loss": 0.3297,
+      "step": 45335
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.929840417468209,
+      "learning_rate": 6.564145355718337e-07,
+      "loss": 0.3576,
+      "step": 45340
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.996367194908822,
+      "learning_rate": 6.556029522073109e-07,
+      "loss": 0.3457,
+      "step": 45345
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.053237158443029,
+      "learning_rate": 6.547918356674232e-07,
+      "loss": 0.3318,
+      "step": 45350
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.260114867991111,
+      "learning_rate": 6.539811860393292e-07,
+      "loss": 0.316,
+      "step": 45355
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.586001229471724,
+      "learning_rate": 6.531710034101385e-07,
+      "loss": 0.352,
+      "step": 45360
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.597369569231543,
+      "learning_rate": 6.523612878669066e-07,
+      "loss": 0.3236,
+      "step": 45365
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.5669840695910215,
+      "learning_rate": 6.51552039496643e-07,
+      "loss": 0.352,
+      "step": 45370
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.900673091862434,
+      "learning_rate": 6.507432583863039e-07,
+      "loss": 0.342,
+      "step": 45375
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.500785366803441,
+      "learning_rate": 6.499349446227948e-07,
+      "loss": 0.3356,
+      "step": 45380
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.010166337084637,
+      "learning_rate": 6.491270982929759e-07,
+      "loss": 0.3354,
+      "step": 45385
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.7542688293039514,
+      "learning_rate": 6.48319719483651e-07,
+      "loss": 0.3217,
+      "step": 45390
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.746637686073376,
+      "learning_rate": 6.475128082815784e-07,
+      "loss": 0.3543,
+      "step": 45395
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.750444987209024,
+      "learning_rate": 6.467063647734639e-07,
+      "loss": 0.328,
+      "step": 45400
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.671654842897656,
+      "learning_rate": 6.45900389045962e-07,
+      "loss": 0.3354,
+      "step": 45405
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.64219554285188,
+      "learning_rate": 6.450948811856799e-07,
+      "loss": 0.3465,
+      "step": 45410
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.88899211877032,
+      "learning_rate": 6.44289841279172e-07,
+      "loss": 0.3179,
+      "step": 45415
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.236730341052237,
+      "learning_rate": 6.434852694129451e-07,
+      "loss": 0.3394,
+      "step": 45420
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.6235005034713454,
+      "learning_rate": 6.426811656734522e-07,
+      "loss": 0.3201,
+      "step": 45425
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.552357981016617,
+      "learning_rate": 6.418775301470975e-07,
+      "loss": 0.339,
+      "step": 45430
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.778345841882862,
+      "learning_rate": 6.410743629202359e-07,
+      "loss": 0.3385,
+      "step": 45435
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.7549072480216825,
+      "learning_rate": 6.402716640791729e-07,
+      "loss": 0.3196,
+      "step": 45440
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.258553257149767,
+      "learning_rate": 6.394694337101593e-07,
+      "loss": 0.3409,
+      "step": 45445
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.856390728138945,
+      "learning_rate": 6.386676718993994e-07,
+      "loss": 0.3826,
+      "step": 45450
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.347289307679359,
+      "learning_rate": 6.378663787330475e-07,
+      "loss": 0.3561,
+      "step": 45455
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.637035878644347,
+      "learning_rate": 6.370655542972043e-07,
+      "loss": 0.333,
+      "step": 45460
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.761770152061556,
+      "learning_rate": 6.362651986779217e-07,
+      "loss": 0.3456,
+      "step": 45465
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.350322669144332,
+      "learning_rate": 6.354653119612031e-07,
+      "loss": 0.336,
+      "step": 45470
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.55118386582183,
+      "learning_rate": 6.346658942329981e-07,
+      "loss": 0.3171,
+      "step": 45475
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.181971097774808,
+      "learning_rate": 6.33866945579209e-07,
+      "loss": 0.3399,
+      "step": 45480
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.9109321842733795,
+      "learning_rate": 6.330684660856862e-07,
+      "loss": 0.3514,
+      "step": 45485
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.373303427649885,
+      "learning_rate": 6.322704558382281e-07,
+      "loss": 0.316,
+      "step": 45490
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.326360984640099,
+      "learning_rate": 6.314729149225868e-07,
+      "loss": 0.3236,
+      "step": 45495
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.740880807805324,
+      "learning_rate": 6.306758434244597e-07,
+      "loss": 0.3271,
+      "step": 45500
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.865616169432333,
+      "learning_rate": 6.29879241429498e-07,
+      "loss": 0.369,
+      "step": 45505
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.665183473917409,
+      "learning_rate": 6.290831090232979e-07,
+      "loss": 0.3752,
+      "step": 45510
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.695214953843375,
+      "learning_rate": 6.282874462914074e-07,
+      "loss": 0.3228,
+      "step": 45515
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.458721089517722,
+      "learning_rate": 6.274922533193256e-07,
+      "loss": 0.3181,
+      "step": 45520
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.859867224898973,
+      "learning_rate": 6.266975301924977e-07,
+      "loss": 0.3176,
+      "step": 45525
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.802227763539425,
+      "learning_rate": 6.259032769963208e-07,
+      "loss": 0.3374,
+      "step": 45530
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.450749754037471,
+      "learning_rate": 6.251094938161423e-07,
+      "loss": 0.3373,
+      "step": 45535
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.3726303450315696,
+      "learning_rate": 6.24316180737255e-07,
+      "loss": 0.3368,
+      "step": 45540
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.149645236851189,
+      "learning_rate": 6.235233378449074e-07,
+      "loss": 0.3112,
+      "step": 45545
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.2309138674718625,
+      "learning_rate": 6.227309652242902e-07,
+      "loss": 0.3453,
+      "step": 45550
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.7130040876794315,
+      "learning_rate": 6.219390629605504e-07,
+      "loss": 0.3426,
+      "step": 45555
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 6.8412221986098345,
+      "learning_rate": 6.211476311387793e-07,
+      "loss": 0.3321,
+      "step": 45560
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.156391661882892,
+      "learning_rate": 6.203566698440216e-07,
+      "loss": 0.3167,
+      "step": 45565
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.849943062979944,
+      "learning_rate": 6.195661791612684e-07,
+      "loss": 0.3254,
+      "step": 45570
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.223070098014928,
+      "learning_rate": 6.187761591754604e-07,
+      "loss": 0.3541,
+      "step": 45575
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.569147490751449,
+      "learning_rate": 6.179866099714915e-07,
+      "loss": 0.3659,
+      "step": 45580
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.605059064041639,
+      "learning_rate": 6.171975316341993e-07,
+      "loss": 0.2976,
+      "step": 45585
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.320355527719186,
+      "learning_rate": 6.164089242483762e-07,
+      "loss": 0.3378,
+      "step": 45590
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 6.076168671196925,
+      "learning_rate": 6.156207878987591e-07,
+      "loss": 0.3459,
+      "step": 45595
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.171272005921741,
+      "learning_rate": 6.148331226700394e-07,
+      "loss": 0.3014,
+      "step": 45600
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.822150537689046,
+      "learning_rate": 6.140459286468537e-07,
+      "loss": 0.3476,
+      "step": 45605
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.424041346320069,
+      "learning_rate": 6.132592059137887e-07,
+      "loss": 0.3543,
+      "step": 45610
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.40380960558513,
+      "learning_rate": 6.124729545553837e-07,
+      "loss": 0.3399,
+      "step": 45615
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.354273054194549,
+      "learning_rate": 6.116871746561215e-07,
+      "loss": 0.345,
+      "step": 45620
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.716864858849761,
+      "learning_rate": 6.109018663004401e-07,
+      "loss": 0.3325,
+      "step": 45625
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.808751125848432,
+      "learning_rate": 6.101170295727243e-07,
+      "loss": 0.3543,
+      "step": 45630
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.666354198133943,
+      "learning_rate": 6.093326645573072e-07,
+      "loss": 0.3073,
+      "step": 45635
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.472781175296722,
+      "learning_rate": 6.085487713384736e-07,
+      "loss": 0.3425,
+      "step": 45640
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.091290667815048,
+      "learning_rate": 6.077653500004543e-07,
+      "loss": 0.3274,
+      "step": 45645
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.384955573643156,
+      "learning_rate": 6.069824006274338e-07,
+      "loss": 0.3414,
+      "step": 45650
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.716092923737294,
+      "learning_rate": 6.06199923303542e-07,
+      "loss": 0.3383,
+      "step": 45655
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.819062499765147,
+      "learning_rate": 6.054179181128589e-07,
+      "loss": 0.348,
+      "step": 45660
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.232173449885258,
+      "learning_rate": 6.046363851394161e-07,
+      "loss": 0.3337,
+      "step": 45665
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.095511459240964,
+      "learning_rate": 6.038553244671913e-07,
+      "loss": 0.3311,
+      "step": 45670
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.717930865497621,
+      "learning_rate": 6.030747361801143e-07,
+      "loss": 0.3674,
+      "step": 45675
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.905351903685142,
+      "learning_rate": 6.022946203620611e-07,
+      "loss": 0.3819,
+      "step": 45680
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.688057256212101,
+      "learning_rate": 6.015149770968603e-07,
+      "loss": 0.3344,
+      "step": 45685
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.493660416571188,
+      "learning_rate": 6.007358064682872e-07,
+      "loss": 0.3496,
+      "step": 45690
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.526337792244437,
+      "learning_rate": 5.999571085600664e-07,
+      "loss": 0.3367,
+      "step": 45695
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.418054059865418,
+      "learning_rate": 5.991788834558737e-07,
+      "loss": 0.3227,
+      "step": 45700
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.405627532145009,
+      "learning_rate": 5.984011312393317e-07,
+      "loss": 0.3698,
+      "step": 45705
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.778899022311874,
+      "learning_rate": 5.976238519940148e-07,
+      "loss": 0.357,
+      "step": 45710
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.8896373824742,
+      "learning_rate": 5.968470458034431e-07,
+      "loss": 0.349,
+      "step": 45715
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.529989689394601,
+      "learning_rate": 5.96070712751089e-07,
+      "loss": 0.3348,
+      "step": 45720
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.7401925352563685,
+      "learning_rate": 5.952948529203739e-07,
+      "loss": 0.3746,
+      "step": 45725
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.399194053655546,
+      "learning_rate": 5.945194663946651e-07,
+      "loss": 0.3388,
+      "step": 45730
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.894700584365297,
+      "learning_rate": 5.937445532572839e-07,
+      "loss": 0.3295,
+      "step": 45735
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.381968416611394,
+      "learning_rate": 5.929701135914961e-07,
+      "loss": 0.3339,
+      "step": 45740
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.822640331371169,
+      "learning_rate": 5.921961474805187e-07,
+      "loss": 0.3622,
+      "step": 45745
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.263638447804075,
+      "learning_rate": 5.914226550075192e-07,
+      "loss": 0.2818,
+      "step": 45750
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.371105809226048,
+      "learning_rate": 5.906496362556113e-07,
+      "loss": 0.3414,
+      "step": 45755
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.874934446850226,
+      "learning_rate": 5.898770913078606e-07,
+      "loss": 0.3285,
+      "step": 45760
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.50674168631367,
+      "learning_rate": 5.891050202472787e-07,
+      "loss": 0.3216,
+      "step": 45765
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.499476756010894,
+      "learning_rate": 5.883334231568305e-07,
+      "loss": 0.3397,
+      "step": 45770
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.310003041718294,
+      "learning_rate": 5.875623001194259e-07,
+      "loss": 0.3647,
+      "step": 45775
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.961249246450765,
+      "learning_rate": 5.86791651217925e-07,
+      "loss": 0.3353,
+      "step": 45780
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.32262916158411,
+      "learning_rate": 5.860214765351391e-07,
+      "loss": 0.2882,
+      "step": 45785
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.421705242837742,
+      "learning_rate": 5.852517761538246e-07,
+      "loss": 0.3322,
+      "step": 45790
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.605838844262771,
+      "learning_rate": 5.84482550156692e-07,
+      "loss": 0.3156,
+      "step": 45795
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.452032822707266,
+      "learning_rate": 5.837137986263953e-07,
+      "loss": 0.3134,
+      "step": 45800
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.021876932768459,
+      "learning_rate": 5.82945521645542e-07,
+      "loss": 0.2999,
+      "step": 45805
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.067206429559478,
+      "learning_rate": 5.82177719296686e-07,
+      "loss": 0.3067,
+      "step": 45810
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.987951345683465,
+      "learning_rate": 5.814103916623309e-07,
+      "loss": 0.3455,
+      "step": 45815
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.57652436217481,
+      "learning_rate": 5.806435388249315e-07,
+      "loss": 0.3402,
+      "step": 45820
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.812825272638316,
+      "learning_rate": 5.798771608668879e-07,
+      "loss": 0.3352,
+      "step": 45825
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.9354679188340835,
+      "learning_rate": 5.791112578705493e-07,
+      "loss": 0.3111,
+      "step": 45830
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.645811423727055,
+      "learning_rate": 5.783458299182182e-07,
+      "loss": 0.3346,
+      "step": 45835
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.400523525358847,
+      "learning_rate": 5.77580877092141e-07,
+      "loss": 0.362,
+      "step": 45840
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.250070510949164,
+      "learning_rate": 5.768163994745174e-07,
+      "loss": 0.3336,
+      "step": 45845
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.50392073712273,
+      "learning_rate": 5.760523971474913e-07,
+      "loss": 0.3264,
+      "step": 45850
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.587159037433273,
+      "learning_rate": 5.752888701931608e-07,
+      "loss": 0.3125,
+      "step": 45855
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.767630603823105,
+      "learning_rate": 5.745258186935687e-07,
+      "loss": 0.3303,
+      "step": 45860
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.183068224338198,
+      "learning_rate": 5.737632427307077e-07,
+      "loss": 0.3465,
+      "step": 45865
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.81252838819064,
+      "learning_rate": 5.730011423865217e-07,
+      "loss": 0.3482,
+      "step": 45870
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.254997130996933,
+      "learning_rate": 5.722395177428997e-07,
+      "loss": 0.3291,
+      "step": 45875
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.960604300636512,
+      "learning_rate": 5.714783688816833e-07,
+      "loss": 0.3475,
+      "step": 45880
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.052054771109202,
+      "learning_rate": 5.707176958846617e-07,
+      "loss": 0.3386,
+      "step": 45885
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.6240626442632085,
+      "learning_rate": 5.699574988335699e-07,
+      "loss": 0.3317,
+      "step": 45890
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.774035892221966,
+      "learning_rate": 5.69197777810096e-07,
+      "loss": 0.3466,
+      "step": 45895
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.057198059861786,
+      "learning_rate": 5.684385328958769e-07,
+      "loss": 0.3498,
+      "step": 45900
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.172782188468089,
+      "learning_rate": 5.676797641724941e-07,
+      "loss": 0.3058,
+      "step": 45905
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.372762377500973,
+      "learning_rate": 5.669214717214833e-07,
+      "loss": 0.3071,
+      "step": 45910
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.518003780721541,
+      "learning_rate": 5.661636556243238e-07,
+      "loss": 0.3179,
+      "step": 45915
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.696054335029156,
+      "learning_rate": 5.654063159624484e-07,
+      "loss": 0.3298,
+      "step": 45920
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.643356633079536,
+      "learning_rate": 5.646494528172347e-07,
+      "loss": 0.3382,
+      "step": 45925
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.561951257751429,
+      "learning_rate": 5.638930662700132e-07,
+      "loss": 0.3288,
+      "step": 45930
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.067931793655128,
+      "learning_rate": 5.631371564020588e-07,
+      "loss": 0.341,
+      "step": 45935
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.575033423836401,
+      "learning_rate": 5.623817232945994e-07,
+      "loss": 0.3251,
+      "step": 45940
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.669388142208612,
+      "learning_rate": 5.616267670288084e-07,
+      "loss": 0.2978,
+      "step": 45945
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.50147579082452,
+      "learning_rate": 5.60872287685808e-07,
+      "loss": 0.3579,
+      "step": 45950
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.6900589118001035,
+      "learning_rate": 5.60118285346673e-07,
+      "loss": 0.3205,
+      "step": 45955
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.023171376558399,
+      "learning_rate": 5.593647600924218e-07,
+      "loss": 0.3466,
+      "step": 45960
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.471142160791876,
+      "learning_rate": 5.586117120040263e-07,
+      "loss": 0.3089,
+      "step": 45965
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 7.117263786411965,
+      "learning_rate": 5.578591411624034e-07,
+      "loss": 0.3277,
+      "step": 45970
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.581250528514581,
+      "learning_rate": 5.571070476484197e-07,
+      "loss": 0.334,
+      "step": 45975
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.350734881273643,
+      "learning_rate": 5.563554315428926e-07,
+      "loss": 0.3265,
+      "step": 45980
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.2934887443968615,
+      "learning_rate": 5.556042929265848e-07,
+      "loss": 0.3162,
+      "step": 45985
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.192878975313347,
+      "learning_rate": 5.548536318802106e-07,
+      "loss": 0.2995,
+      "step": 45990
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.691314209883258,
+      "learning_rate": 5.54103448484432e-07,
+      "loss": 0.3475,
+      "step": 45995
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.148991913722001,
+      "learning_rate": 5.533537428198582e-07,
+      "loss": 0.3118,
+      "step": 46000
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.674358334706993,
+      "learning_rate": 5.526045149670506e-07,
+      "loss": 0.3397,
+      "step": 46005
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.8153034598221245,
+      "learning_rate": 5.518557650065149e-07,
+      "loss": 0.3446,
+      "step": 46010
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.329807125740926,
+      "learning_rate": 5.511074930187088e-07,
+      "loss": 0.351,
+      "step": 46015
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.416949608034508,
+      "learning_rate": 5.503596990840359e-07,
+      "loss": 0.3168,
+      "step": 46020
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.514112255790046,
+      "learning_rate": 5.496123832828526e-07,
+      "loss": 0.3263,
+      "step": 46025
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.7644315304352265,
+      "learning_rate": 5.4886554569546e-07,
+      "loss": 0.3599,
+      "step": 46030
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.898275421218794,
+      "learning_rate": 5.481191864021074e-07,
+      "loss": 0.3232,
+      "step": 46035
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.717841109761777,
+      "learning_rate": 5.473733054829971e-07,
+      "loss": 0.3429,
+      "step": 46040
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.8288551452757296,
+      "learning_rate": 5.466279030182747e-07,
+      "loss": 0.2774,
+      "step": 46045
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.869427798765506,
+      "learning_rate": 5.458829790880399e-07,
+      "loss": 0.3592,
+      "step": 46050
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 7.413220163693086,
+      "learning_rate": 5.451385337723364e-07,
+      "loss": 0.3292,
+      "step": 46055
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.690746837620549,
+      "learning_rate": 5.443945671511569e-07,
+      "loss": 0.3145,
+      "step": 46060
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.545457596969626,
+      "learning_rate": 5.436510793044464e-07,
+      "loss": 0.3319,
+      "step": 46065
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.308901417366394,
+      "learning_rate": 5.429080703120937e-07,
+      "loss": 0.3173,
+      "step": 46070
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.065855482863909,
+      "learning_rate": 5.421655402539405e-07,
+      "loss": 0.3087,
+      "step": 46075
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.242778617409167,
+      "learning_rate": 5.414234892097731e-07,
+      "loss": 0.3292,
+      "step": 46080
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.498659647638434,
+      "learning_rate": 5.406819172593292e-07,
+      "loss": 0.3348,
+      "step": 46085
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.694945929852806,
+      "learning_rate": 5.399408244822946e-07,
+      "loss": 0.3315,
+      "step": 46090
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.756773699081069,
+      "learning_rate": 5.392002109583006e-07,
+      "loss": 0.33,
+      "step": 46095
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.820117598788975,
+      "learning_rate": 5.384600767669324e-07,
+      "loss": 0.3182,
+      "step": 46100
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.385376533448734,
+      "learning_rate": 5.377204219877186e-07,
+      "loss": 0.3074,
+      "step": 46105
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.957435969071599,
+      "learning_rate": 5.369812467001395e-07,
+      "loss": 0.3437,
+      "step": 46110
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.4331366581863065,
+      "learning_rate": 5.362425509836222e-07,
+      "loss": 0.3089,
+      "step": 46115
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.16092374237919,
+      "learning_rate": 5.35504334917542e-07,
+      "loss": 0.3285,
+      "step": 46120
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 7.583005746145502,
+      "learning_rate": 5.347665985812255e-07,
+      "loss": 0.3172,
+      "step": 46125
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.862785861064724,
+      "learning_rate": 5.340293420539433e-07,
+      "loss": 0.317,
+      "step": 46130
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.497611114307605,
+      "learning_rate": 5.332925654149191e-07,
+      "loss": 0.3123,
+      "step": 46135
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.290820388737501,
+      "learning_rate": 5.325562687433222e-07,
+      "loss": 0.333,
+      "step": 46140
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.978808729520784,
+      "learning_rate": 5.318204521182691e-07,
+      "loss": 0.3637,
+      "step": 46145
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.884145986032276,
+      "learning_rate": 5.310851156188296e-07,
+      "loss": 0.316,
+      "step": 46150
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.803269631029532,
+      "learning_rate": 5.303502593240156e-07,
+      "loss": 0.3391,
+      "step": 46155
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.250440770145616,
+      "learning_rate": 5.296158833127935e-07,
+      "loss": 0.3359,
+      "step": 46160
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.597864642390691,
+      "learning_rate": 5.288819876640733e-07,
+      "loss": 0.3356,
+      "step": 46165
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.169458501789328,
+      "learning_rate": 5.28148572456717e-07,
+      "loss": 0.3477,
+      "step": 46170
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.554852646821307,
+      "learning_rate": 5.274156377695311e-07,
+      "loss": 0.3162,
+      "step": 46175
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.7438969147168395,
+      "learning_rate": 5.266831836812742e-07,
+      "loss": 0.3324,
+      "step": 46180
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.004805065972454,
+      "learning_rate": 5.259512102706521e-07,
+      "loss": 0.3372,
+      "step": 46185
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.724986294140625,
+      "learning_rate": 5.252197176163176e-07,
+      "loss": 0.3584,
+      "step": 46190
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.030369466747653,
+      "learning_rate": 5.244887057968739e-07,
+      "loss": 0.3206,
+      "step": 46195
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.764662918969909,
+      "learning_rate": 5.237581748908705e-07,
+      "loss": 0.3224,
+      "step": 46200
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.597006049270838,
+      "learning_rate": 5.230281249768055e-07,
+      "loss": 0.3502,
+      "step": 46205
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.372302759834244,
+      "learning_rate": 5.222985561331279e-07,
+      "loss": 0.3243,
+      "step": 46210
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.488255978863521,
+      "learning_rate": 5.215694684382305e-07,
+      "loss": 0.3318,
+      "step": 46215
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.2101604116508184,
+      "learning_rate": 5.208408619704602e-07,
+      "loss": 0.3551,
+      "step": 46220
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.484927201361623,
+      "learning_rate": 5.201127368081072e-07,
+      "loss": 0.3254,
+      "step": 46225
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.812075795884978,
+      "learning_rate": 5.193850930294109e-07,
+      "loss": 0.3521,
+      "step": 46230
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.227904412365689,
+      "learning_rate": 5.186579307125617e-07,
+      "loss": 0.3233,
+      "step": 46235
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.48733536905027,
+      "learning_rate": 5.179312499356942e-07,
+      "loss": 0.3517,
+      "step": 46240
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.9012095004225715,
+      "learning_rate": 5.172050507768961e-07,
+      "loss": 0.3212,
+      "step": 46245
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.213389781683783,
+      "learning_rate": 5.164793333141982e-07,
+      "loss": 0.3478,
+      "step": 46250
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.549484073230738,
+      "learning_rate": 5.157540976255848e-07,
+      "loss": 0.3423,
+      "step": 46255
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.7062223505257474,
+      "learning_rate": 5.150293437889825e-07,
+      "loss": 0.32,
+      "step": 46260
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.482824436298924,
+      "learning_rate": 5.143050718822723e-07,
+      "loss": 0.3258,
+      "step": 46265
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.460343629837303,
+      "learning_rate": 5.135812819832775e-07,
+      "loss": 0.308,
+      "step": 46270
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.612973732820175,
+      "learning_rate": 5.128579741697742e-07,
+      "loss": 0.3001,
+      "step": 46275
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.801668871586329,
+      "learning_rate": 5.121351485194864e-07,
+      "loss": 0.3431,
+      "step": 46280
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.711319975233357,
+      "learning_rate": 5.114128051100825e-07,
+      "loss": 0.3386,
+      "step": 46285
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.656998448073274,
+      "learning_rate": 5.10690944019182e-07,
+      "loss": 0.3205,
+      "step": 46290
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.47779574776888,
+      "learning_rate": 5.099695653243531e-07,
+      "loss": 0.3251,
+      "step": 46295
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.191342791615926,
+      "learning_rate": 5.092486691031095e-07,
+      "loss": 0.3344,
+      "step": 46300
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.7625385521267605,
+      "learning_rate": 5.085282554329163e-07,
+      "loss": 0.3457,
+      "step": 46305
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.3258968327760146,
+      "learning_rate": 5.078083243911846e-07,
+      "loss": 0.3167,
+      "step": 46310
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 6.728115614650199,
+      "learning_rate": 5.070888760552733e-07,
+      "loss": 0.3193,
+      "step": 46315
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.710623912333051,
+      "learning_rate": 5.063699105024916e-07,
+      "loss": 0.3463,
+      "step": 46320
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.868176111442243,
+      "learning_rate": 5.056514278100943e-07,
+      "loss": 0.3337,
+      "step": 46325
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.897804826606414,
+      "learning_rate": 5.04933428055287e-07,
+      "loss": 0.3413,
+      "step": 46330
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.48526100218459,
+      "learning_rate": 5.042159113152201e-07,
+      "loss": 0.3398,
+      "step": 46335
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.179763745121832,
+      "learning_rate": 5.034988776669958e-07,
+      "loss": 0.3351,
+      "step": 46340
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.451738688720434,
+      "learning_rate": 5.02782327187662e-07,
+      "loss": 0.3259,
+      "step": 46345
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.330320228509962,
+      "learning_rate": 5.020662599542137e-07,
+      "loss": 0.3317,
+      "step": 46350
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.0568547250309175,
+      "learning_rate": 5.013506760435966e-07,
+      "loss": 0.3104,
+      "step": 46355
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.541983116924022,
+      "learning_rate": 5.006355755327041e-07,
+      "loss": 0.3397,
+      "step": 46360
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.547305181395205,
+      "learning_rate": 4.999209584983755e-07,
+      "loss": 0.335,
+      "step": 46365
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.776630827040175,
+      "learning_rate": 4.992068250174014e-07,
+      "loss": 0.3363,
+      "step": 46370
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.068470514595292,
+      "learning_rate": 4.984931751665167e-07,
+      "loss": 0.3185,
+      "step": 46375
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.344035870284013,
+      "learning_rate": 4.977800090224078e-07,
+      "loss": 0.3338,
+      "step": 46380
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.781788711941363,
+      "learning_rate": 4.970673266617055e-07,
+      "loss": 0.3219,
+      "step": 46385
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.230395607717361,
+      "learning_rate": 4.96355128160993e-07,
+      "loss": 0.3348,
+      "step": 46390
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.450801750708208,
+      "learning_rate": 4.95643413596798e-07,
+      "loss": 0.3466,
+      "step": 46395
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.648472660225281,
+      "learning_rate": 4.949321830455966e-07,
+      "loss": 0.3337,
+      "step": 46400
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.438694675231268,
+      "learning_rate": 4.942214365838155e-07,
+      "loss": 0.3145,
+      "step": 46405
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.364386738114903,
+      "learning_rate": 4.935111742878257e-07,
+      "loss": 0.3491,
+      "step": 46410
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.074838813178466,
+      "learning_rate": 4.928013962339495e-07,
+      "loss": 0.3263,
+      "step": 46415
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.774070797456436,
+      "learning_rate": 4.920921024984543e-07,
+      "loss": 0.3414,
+      "step": 46420
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.153494379702369,
+      "learning_rate": 4.913832931575585e-07,
+      "loss": 0.2957,
+      "step": 46425
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.746596480095154,
+      "learning_rate": 4.906749682874257e-07,
+      "loss": 0.3139,
+      "step": 46430
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.9289718278731165,
+      "learning_rate": 4.899671279641682e-07,
+      "loss": 0.3287,
+      "step": 46435
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.1760119861251415,
+      "learning_rate": 4.892597722638476e-07,
+      "loss": 0.344,
+      "step": 46440
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.576907565430694,
+      "learning_rate": 4.885529012624713e-07,
+      "loss": 0.322,
+      "step": 46445
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.912295010945285,
+      "learning_rate": 4.878465150359962e-07,
+      "loss": 0.3272,
+      "step": 46450
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 6.02785863344295,
+      "learning_rate": 4.871406136603279e-07,
+      "loss": 0.3484,
+      "step": 46455
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.395925493871996,
+      "learning_rate": 4.864351972113163e-07,
+      "loss": 0.3187,
+      "step": 46460
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.405042994822878,
+      "learning_rate": 4.857302657647633e-07,
+      "loss": 0.3206,
+      "step": 46465
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.565139131974331,
+      "learning_rate": 4.850258193964152e-07,
+      "loss": 0.3281,
+      "step": 46470
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.404712308416535,
+      "learning_rate": 4.843218581819703e-07,
+      "loss": 0.3524,
+      "step": 46475
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.0384397446786275,
+      "learning_rate": 4.836183821970708e-07,
+      "loss": 0.3484,
+      "step": 46480
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.936290176541444,
+      "learning_rate": 4.829153915173074e-07,
+      "loss": 0.3112,
+      "step": 46485
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.399157552521201,
+      "learning_rate": 4.822128862182218e-07,
+      "loss": 0.3077,
+      "step": 46490
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.886471784660722,
+      "learning_rate": 4.815108663752988e-07,
+      "loss": 0.3339,
+      "step": 46495
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.611838625994117,
+      "learning_rate": 4.808093320639762e-07,
+      "loss": 0.3332,
+      "step": 46500
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.615169741809314,
+      "learning_rate": 4.801082833596344e-07,
+      "loss": 0.327,
+      "step": 46505
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 6.705207984017148,
+      "learning_rate": 4.794077203376063e-07,
+      "loss": 0.3373,
+      "step": 46510
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.54452367481361,
+      "learning_rate": 4.787076430731696e-07,
+      "loss": 0.3451,
+      "step": 46515
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.528245726811377,
+      "learning_rate": 4.780080516415497e-07,
+      "loss": 0.3445,
+      "step": 46520
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.643215275404237,
+      "learning_rate": 4.773089461179226e-07,
+      "loss": 0.3437,
+      "step": 46525
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.265512148486137,
+      "learning_rate": 4.76610326577408e-07,
+      "loss": 0.342,
+      "step": 46530
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.445016881382162,
+      "learning_rate": 4.7591219309507886e-07,
+      "loss": 0.3369,
+      "step": 46535
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.721337830790504,
+      "learning_rate": 4.752145457459495e-07,
+      "loss": 0.3187,
+      "step": 46540
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.354957394363971,
+      "learning_rate": 4.7451738460498664e-07,
+      "loss": 0.3251,
+      "step": 46545
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.767370799936989,
+      "learning_rate": 4.7382070974710426e-07,
+      "loss": 0.3112,
+      "step": 46550
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.415702968447362,
+      "learning_rate": 4.731245212471608e-07,
+      "loss": 0.3086,
+      "step": 46555
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.533344978473302,
+      "learning_rate": 4.724288191799675e-07,
+      "loss": 0.3226,
+      "step": 46560
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.99179213634612,
+      "learning_rate": 4.717336036202785e-07,
+      "loss": 0.3524,
+      "step": 46565
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.707278535069128,
+      "learning_rate": 4.7103887464279907e-07,
+      "loss": 0.3235,
+      "step": 46570
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.751202072124831,
+      "learning_rate": 4.703446323221811e-07,
+      "loss": 0.3125,
+      "step": 46575
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.664623967429161,
+      "learning_rate": 4.6965087673302155e-07,
+      "loss": 0.3242,
+      "step": 46580
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.453480376202116,
+      "learning_rate": 4.689576079498703e-07,
+      "loss": 0.352,
+      "step": 46585
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.418918181892823,
+      "learning_rate": 4.6826482604722057e-07,
+      "loss": 0.373,
+      "step": 46590
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.493629490574223,
+      "learning_rate": 4.6757253109951607e-07,
+      "loss": 0.3284,
+      "step": 46595
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.269964245495313,
+      "learning_rate": 4.668807231811462e-07,
+      "loss": 0.3288,
+      "step": 46600
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.53863055003813,
+      "learning_rate": 4.661894023664482e-07,
+      "loss": 0.3379,
+      "step": 46605
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.255523569854855,
+      "learning_rate": 4.654985687297087e-07,
+      "loss": 0.2973,
+      "step": 46610
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.526356223682446,
+      "learning_rate": 4.648082223451594e-07,
+      "loss": 0.3266,
+      "step": 46615
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.421706598087153,
+      "learning_rate": 4.641183632869828e-07,
+      "loss": 0.3049,
+      "step": 46620
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.573584685497384,
+      "learning_rate": 4.6342899162930557e-07,
+      "loss": 0.3584,
+      "step": 46625
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.687356946628921,
+      "learning_rate": 4.6274010744620567e-07,
+      "loss": 0.3014,
+      "step": 46630
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.136176011071282,
+      "learning_rate": 4.620517108117045e-07,
+      "loss": 0.2939,
+      "step": 46635
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.316344056456715,
+      "learning_rate": 4.6136380179977456e-07,
+      "loss": 0.3301,
+      "step": 46640
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.3163159087472724,
+      "learning_rate": 4.6067638048433616e-07,
+      "loss": 0.3377,
+      "step": 46645
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.459929959955323,
+      "learning_rate": 4.599894469392524e-07,
+      "loss": 0.3122,
+      "step": 46650
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.416213608940536,
+      "learning_rate": 4.593030012383409e-07,
+      "loss": 0.3132,
+      "step": 46655
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.687807797740529,
+      "learning_rate": 4.586170434553611e-07,
+      "loss": 0.3252,
+      "step": 46660
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.107190719012708,
+      "learning_rate": 4.5793157366402165e-07,
+      "loss": 0.3048,
+      "step": 46665
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.956873437855812,
+      "learning_rate": 4.572465919379815e-07,
+      "loss": 0.3359,
+      "step": 46670
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.735126849106469,
+      "learning_rate": 4.565620983508423e-07,
+      "loss": 0.3672,
+      "step": 46675
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.472038017751582,
+      "learning_rate": 4.558780929761586e-07,
+      "loss": 0.3004,
+      "step": 46680
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.6455324004840675,
+      "learning_rate": 4.5519457588742876e-07,
+      "loss": 0.3655,
+      "step": 46685
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.675128394429441,
+      "learning_rate": 4.5451154715809855e-07,
+      "loss": 0.3308,
+      "step": 46690
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.246548711309785,
+      "learning_rate": 4.538290068615636e-07,
+      "loss": 0.3331,
+      "step": 46695
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.2548599837512375,
+      "learning_rate": 4.5314695507116535e-07,
+      "loss": 0.2764,
+      "step": 46700
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.7598452175108985,
+      "learning_rate": 4.5246539186019455e-07,
+      "loss": 0.3209,
+      "step": 46705
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.525337179491603,
+      "learning_rate": 4.517843173018871e-07,
+      "loss": 0.3056,
+      "step": 46710
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.403329916359623,
+      "learning_rate": 4.5110373146942666e-07,
+      "loss": 0.3364,
+      "step": 46715
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.586314300497535,
+      "learning_rate": 4.504236344359458e-07,
+      "loss": 0.3289,
+      "step": 46720
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.279324615183544,
+      "learning_rate": 4.4974402627452565e-07,
+      "loss": 0.3424,
+      "step": 46725
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.547243659490779,
+      "learning_rate": 4.490649070581904e-07,
+      "loss": 0.3025,
+      "step": 46730
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.927879511934787,
+      "learning_rate": 4.4838627685991563e-07,
+      "loss": 0.3238,
+      "step": 46735
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.6418661930447636,
+      "learning_rate": 4.477081357526242e-07,
+      "loss": 0.3417,
+      "step": 46740
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.328746805770058,
+      "learning_rate": 4.4703048380918434e-07,
+      "loss": 0.3315,
+      "step": 46745
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.469255223881732,
+      "learning_rate": 4.463533211024118e-07,
+      "loss": 0.2829,
+      "step": 46750
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.294176882946335,
+      "learning_rate": 4.456766477050728e-07,
+      "loss": 0.3192,
+      "step": 46755
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 7.866310515172893,
+      "learning_rate": 4.450004636898769e-07,
+      "loss": 0.3148,
+      "step": 46760
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.792377233032178,
+      "learning_rate": 4.4432476912948496e-07,
+      "loss": 0.3386,
+      "step": 46765
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.222320133431159,
+      "learning_rate": 4.436495640965022e-07,
+      "loss": 0.36,
+      "step": 46770
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.149002208127003,
+      "learning_rate": 4.429748486634811e-07,
+      "loss": 0.3481,
+      "step": 46775
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.3096026730296835,
+      "learning_rate": 4.4230062290292586e-07,
+      "loss": 0.3084,
+      "step": 46780
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.3440879972361515,
+      "learning_rate": 4.4162688688728194e-07,
+      "loss": 0.3442,
+      "step": 46785
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.76826264700687,
+      "learning_rate": 4.4095364068894755e-07,
+      "loss": 0.3541,
+      "step": 46790
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.85168291242625,
+      "learning_rate": 4.402808843802653e-07,
+      "loss": 0.3164,
+      "step": 46795
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.88661915370977,
+      "learning_rate": 4.3960861803352463e-07,
+      "loss": 0.3345,
+      "step": 46800
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.421894520432229,
+      "learning_rate": 4.389368417209655e-07,
+      "loss": 0.3303,
+      "step": 46805
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.668288951826391,
+      "learning_rate": 4.3826555551477135e-07,
+      "loss": 0.3409,
+      "step": 46810
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.513252889281303,
+      "learning_rate": 4.3759475948707554e-07,
+      "loss": 0.3097,
+      "step": 46815
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.114137039029897,
+      "learning_rate": 4.3692445370995875e-07,
+      "loss": 0.3519,
+      "step": 46820
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.072764221298067,
+      "learning_rate": 4.3625463825544777e-07,
+      "loss": 0.292,
+      "step": 46825
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.560733578451888,
+      "learning_rate": 4.355853131955173e-07,
+      "loss": 0.3347,
+      "step": 46830
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.779777260598413,
+      "learning_rate": 4.3491647860208877e-07,
+      "loss": 0.321,
+      "step": 46835
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.687178481559941,
+      "learning_rate": 4.3424813454703285e-07,
+      "loss": 0.3136,
+      "step": 46840
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.475830281431158,
+      "learning_rate": 4.335802811021644e-07,
+      "loss": 0.2911,
+      "step": 46845
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.479512983952609,
+      "learning_rate": 4.329129183392483e-07,
+      "loss": 0.2995,
+      "step": 46850
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.822087055753844,
+      "learning_rate": 4.3224604632999543e-07,
+      "loss": 0.3124,
+      "step": 46855
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.655909129757384,
+      "learning_rate": 4.3157966514606295e-07,
+      "loss": 0.3021,
+      "step": 46860
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.529545803506113,
+      "learning_rate": 4.309137748590581e-07,
+      "loss": 0.2858,
+      "step": 46865
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.256877250547265,
+      "learning_rate": 4.3024837554053234e-07,
+      "loss": 0.3019,
+      "step": 46870
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.345462366679429,
+      "learning_rate": 4.2958346726198754e-07,
+      "loss": 0.3902,
+      "step": 46875
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.630696105087748,
+      "learning_rate": 4.289190500948698e-07,
+      "loss": 0.3299,
+      "step": 46880
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.8876201973539155,
+      "learning_rate": 4.282551241105731e-07,
+      "loss": 0.3342,
+      "step": 46885
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.424498315357881,
+      "learning_rate": 4.2759168938044105e-07,
+      "loss": 0.336,
+      "step": 46890
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.799727148662146,
+      "learning_rate": 4.269287459757604e-07,
+      "loss": 0.3308,
+      "step": 46895
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.187968502330678,
+      "learning_rate": 4.262662939677692e-07,
+      "loss": 0.3175,
+      "step": 46900
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.715993452326925,
+      "learning_rate": 4.256043334276494e-07,
+      "loss": 0.3065,
+      "step": 46905
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.771388550950795,
+      "learning_rate": 4.2494286442653245e-07,
+      "loss": 0.3536,
+      "step": 46910
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.868514320748043,
+      "learning_rate": 4.24281887035497e-07,
+      "loss": 0.3239,
+      "step": 46915
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.562426506289905,
+      "learning_rate": 4.236214013255657e-07,
+      "loss": 0.3238,
+      "step": 46920
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.947891848815056,
+      "learning_rate": 4.22961407367713e-07,
+      "loss": 0.3429,
+      "step": 46925
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.446925438033762,
+      "learning_rate": 4.2230190523285586e-07,
+      "loss": 0.2965,
+      "step": 46930
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.79046869026732,
+      "learning_rate": 4.2164289499186327e-07,
+      "loss": 0.3009,
+      "step": 46935
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.84656612305597,
+      "learning_rate": 4.209843767155475e-07,
+      "loss": 0.3524,
+      "step": 46940
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.612128800859239,
+      "learning_rate": 4.2032635047466795e-07,
+      "loss": 0.3391,
+      "step": 46945
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.211833582382217,
+      "learning_rate": 4.1966881633993485e-07,
+      "loss": 0.3058,
+      "step": 46950
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.996857196530111,
+      "learning_rate": 4.1901177438200103e-07,
+      "loss": 0.2945,
+      "step": 46955
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.858078115209991,
+      "learning_rate": 4.1835522467147005e-07,
+      "loss": 0.3363,
+      "step": 46960
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.490565710898016,
+      "learning_rate": 4.176991672788899e-07,
+      "loss": 0.2949,
+      "step": 46965
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.489166325576985,
+      "learning_rate": 4.1704360227475814e-07,
+      "loss": 0.3116,
+      "step": 46970
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.51153850421733,
+      "learning_rate": 4.163885297295178e-07,
+      "loss": 0.3105,
+      "step": 46975
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.807365501389042,
+      "learning_rate": 4.157339497135582e-07,
+      "loss": 0.3112,
+      "step": 46980
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.516339470155919,
+      "learning_rate": 4.15079862297218e-07,
+      "loss": 0.3238,
+      "step": 46985
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.889724706848004,
+      "learning_rate": 4.1442626755078095e-07,
+      "loss": 0.3237,
+      "step": 46990
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.885939649506536,
+      "learning_rate": 4.137731655444799e-07,
+      "loss": 0.3177,
+      "step": 46995
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.767865505475133,
+      "learning_rate": 4.1312055634849137e-07,
+      "loss": 0.3126,
+      "step": 47000
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.925559666567569,
+      "learning_rate": 4.1246844003294317e-07,
+      "loss": 0.3316,
+      "step": 47005
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.789895457295862,
+      "learning_rate": 4.1181681666790817e-07,
+      "loss": 0.3124,
+      "step": 47010
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.48421426232905,
+      "learning_rate": 4.1116568632340424e-07,
+      "loss": 0.3126,
+      "step": 47015
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.780515250240848,
+      "learning_rate": 4.105150490694004e-07,
+      "loss": 0.3327,
+      "step": 47020
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.373610834287886,
+      "learning_rate": 4.098649049758097e-07,
+      "loss": 0.3343,
+      "step": 47025
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.302592901513993,
+      "learning_rate": 4.0921525411249164e-07,
+      "loss": 0.3208,
+      "step": 47030
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.92804899036466,
+      "learning_rate": 4.0856609654925606e-07,
+      "loss": 0.3551,
+      "step": 47035
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.689449107764158,
+      "learning_rate": 4.0791743235585603e-07,
+      "loss": 0.334,
+      "step": 47040
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.276759902726785,
+      "learning_rate": 4.072692616019952e-07,
+      "loss": 0.3155,
+      "step": 47045
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.3844193098775355,
+      "learning_rate": 4.066215843573207e-07,
+      "loss": 0.3324,
+      "step": 47050
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.815219341175224,
+      "learning_rate": 4.0597440069143014e-07,
+      "loss": 0.3221,
+      "step": 47055
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.692955230893898,
+      "learning_rate": 4.053277106738651e-07,
+      "loss": 0.3194,
+      "step": 47060
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.6323274112565445,
+      "learning_rate": 4.0468151437411453e-07,
+      "loss": 0.3141,
+      "step": 47065
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.631325692097944,
+      "learning_rate": 4.040358118616167e-07,
+      "loss": 0.3533,
+      "step": 47070
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.927335165269176,
+      "learning_rate": 4.033906032057533e-07,
+      "loss": 0.3314,
+      "step": 47075
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.877498789312842,
+      "learning_rate": 4.027458884758573e-07,
+      "loss": 0.3216,
+      "step": 47080
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.246272771666314,
+      "learning_rate": 4.0210166774120374e-07,
+      "loss": 0.3425,
+      "step": 47085
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.5247644234160145,
+      "learning_rate": 4.01457941071019e-07,
+      "loss": 0.3153,
+      "step": 47090
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.463647383380514,
+      "learning_rate": 4.008147085344721e-07,
+      "loss": 0.3311,
+      "step": 47095
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.559348588349339,
+      "learning_rate": 4.0017197020068287e-07,
+      "loss": 0.3375,
+      "step": 47100
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.0138635726505365,
+      "learning_rate": 3.995297261387171e-07,
+      "loss": 0.3173,
+      "step": 47105
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.303357727529372,
+      "learning_rate": 3.988879764175857e-07,
+      "loss": 0.2982,
+      "step": 47110
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.298462690324856,
+      "learning_rate": 3.982467211062463e-07,
+      "loss": 0.3373,
+      "step": 47115
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.304089836229813,
+      "learning_rate": 3.9760596027360655e-07,
+      "loss": 0.3076,
+      "step": 47120
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.839555906421958,
+      "learning_rate": 3.9696569398851803e-07,
+      "loss": 0.3307,
+      "step": 47125
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.749661601195171,
+      "learning_rate": 3.9632592231978075e-07,
+      "loss": 0.3214,
+      "step": 47130
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.55621875569353,
+      "learning_rate": 3.9568664533614023e-07,
+      "loss": 0.3002,
+      "step": 47135
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.0782440393019375,
+      "learning_rate": 3.950478631062904e-07,
+      "loss": 0.3302,
+      "step": 47140
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.815075060577531,
+      "learning_rate": 3.944095756988714e-07,
+      "loss": 0.3214,
+      "step": 47145
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.07546656532758,
+      "learning_rate": 3.9377178318246833e-07,
+      "loss": 0.3033,
+      "step": 47150
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.256030271952451,
+      "learning_rate": 3.9313448562561694e-07,
+      "loss": 0.3452,
+      "step": 47155
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.7105927018970934,
+      "learning_rate": 3.924976830967958e-07,
+      "loss": 0.3139,
+      "step": 47160
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.395540297868007,
+      "learning_rate": 3.9186137566443406e-07,
+      "loss": 0.3214,
+      "step": 47165
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.68165705261665,
+      "learning_rate": 3.9122556339690475e-07,
+      "loss": 0.3125,
+      "step": 47170
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.225631421305167,
+      "learning_rate": 3.905902463625277e-07,
+      "loss": 0.2875,
+      "step": 47175
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.282772827428368,
+      "learning_rate": 3.899554246295717e-07,
+      "loss": 0.2889,
+      "step": 47180
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.838280264949798,
+      "learning_rate": 3.893210982662515e-07,
+      "loss": 0.3209,
+      "step": 47185
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.877907268717985,
+      "learning_rate": 3.8868726734072716e-07,
+      "loss": 0.3296,
+      "step": 47190
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.854364067016355,
+      "learning_rate": 3.880539319211085e-07,
+      "loss": 0.3373,
+      "step": 47195
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 6.232484394987799,
+      "learning_rate": 3.8742109207544787e-07,
+      "loss": 0.3294,
+      "step": 47200
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.679387401102267,
+      "learning_rate": 3.8678874787174804e-07,
+      "loss": 0.3355,
+      "step": 47205
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.816690183043029,
+      "learning_rate": 3.861568993779569e-07,
+      "loss": 0.2978,
+      "step": 47210
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.60766726117633,
+      "learning_rate": 3.855255466619695e-07,
+      "loss": 0.2869,
+      "step": 47215
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.56949189783215,
+      "learning_rate": 3.848946897916267e-07,
+      "loss": 0.3089,
+      "step": 47220
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.957594418906804,
+      "learning_rate": 3.8426432883471863e-07,
+      "loss": 0.314,
+      "step": 47225
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.282474940892623,
+      "learning_rate": 3.8363446385897896e-07,
+      "loss": 0.3333,
+      "step": 47230
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.122244530794561,
+      "learning_rate": 3.8300509493208906e-07,
+      "loss": 0.3353,
+      "step": 47235
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 6.846221853812618,
+      "learning_rate": 3.823762221216787e-07,
+      "loss": 0.329,
+      "step": 47240
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.04568969048718,
+      "learning_rate": 3.8174784549532163e-07,
+      "loss": 0.36,
+      "step": 47245
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.340629017456971,
+      "learning_rate": 3.8111996512054104e-07,
+      "loss": 0.3588,
+      "step": 47250
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.644180897846171,
+      "learning_rate": 3.8049258106480525e-07,
+      "loss": 0.2957,
+      "step": 47255
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.14247660758316,
+      "learning_rate": 3.7986569339552804e-07,
+      "loss": 0.3081,
+      "step": 47260
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.974849820472255,
+      "learning_rate": 3.7923930218007334e-07,
+      "loss": 0.324,
+      "step": 47265
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.597716890240932,
+      "learning_rate": 3.786134074857478e-07,
+      "loss": 0.3488,
+      "step": 47270
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.484337347488252,
+      "learning_rate": 3.779880093798072e-07,
+      "loss": 0.3258,
+      "step": 47275
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 6.491127473804733,
+      "learning_rate": 3.773631079294543e-07,
+      "loss": 0.3093,
+      "step": 47280
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.897306774978777,
+      "learning_rate": 3.767387032018355e-07,
+      "loss": 0.3383,
+      "step": 47285
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.787247236921363,
+      "learning_rate": 3.761147952640487e-07,
+      "loss": 0.3003,
+      "step": 47290
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.772062282452134,
+      "learning_rate": 3.7549138418313257e-07,
+      "loss": 0.3312,
+      "step": 47295
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.361851150668638,
+      "learning_rate": 3.748684700260774e-07,
+      "loss": 0.3252,
+      "step": 47300
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.3037318819974235,
+      "learning_rate": 3.742460528598169e-07,
+      "loss": 0.3248,
+      "step": 47305
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.569751016260492,
+      "learning_rate": 3.7362413275123356e-07,
+      "loss": 0.3017,
+      "step": 47310
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.312452729981813,
+      "learning_rate": 3.7300270976715514e-07,
+      "loss": 0.3413,
+      "step": 47315
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.298661718228763,
+      "learning_rate": 3.723817839743554e-07,
+      "loss": 0.3193,
+      "step": 47320
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.563306621629549,
+      "learning_rate": 3.717613554395566e-07,
+      "loss": 0.3081,
+      "step": 47325
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.366297381680645,
+      "learning_rate": 3.7114142422942534e-07,
+      "loss": 0.3139,
+      "step": 47330
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.856945936723807,
+      "learning_rate": 3.705219904105778e-07,
+      "loss": 0.372,
+      "step": 47335
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.350410815432054,
+      "learning_rate": 3.6990305404957404e-07,
+      "loss": 0.2971,
+      "step": 47340
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.962664741492998,
+      "learning_rate": 3.6928461521292036e-07,
+      "loss": 0.316,
+      "step": 47345
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.734462988417854,
+      "learning_rate": 3.686666739670719e-07,
+      "loss": 0.3232,
+      "step": 47350
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.996910036124418,
+      "learning_rate": 3.680492303784289e-07,
+      "loss": 0.3275,
+      "step": 47355
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.170838149373807,
+      "learning_rate": 3.674322845133388e-07,
+      "loss": 0.316,
+      "step": 47360
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.105056301647215,
+      "learning_rate": 3.6681583643809416e-07,
+      "loss": 0.304,
+      "step": 47365
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.798121810378818,
+      "learning_rate": 3.661998862189359e-07,
+      "loss": 0.328,
+      "step": 47370
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.925917264774166,
+      "learning_rate": 3.65584433922051e-07,
+      "loss": 0.3384,
+      "step": 47375
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.889806270974378,
+      "learning_rate": 3.649694796135711e-07,
+      "loss": 0.3082,
+      "step": 47380
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.766925180638513,
+      "learning_rate": 3.643550233595772e-07,
+      "loss": 0.3309,
+      "step": 47385
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.921202646989476,
+      "learning_rate": 3.637410652260942e-07,
+      "loss": 0.3213,
+      "step": 47390
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.470605708476463,
+      "learning_rate": 3.631276052790955e-07,
+      "loss": 0.3101,
+      "step": 47395
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.007220511121349,
+      "learning_rate": 3.6251464358450005e-07,
+      "loss": 0.3194,
+      "step": 47400
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.285716298893889,
+      "learning_rate": 3.619021802081724e-07,
+      "loss": 0.3095,
+      "step": 47405
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.376894087725877,
+      "learning_rate": 3.6129021521592533e-07,
+      "loss": 0.3388,
+      "step": 47410
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.372636904633146,
+      "learning_rate": 3.606787486735158e-07,
+      "loss": 0.3347,
+      "step": 47415
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.155466949842631,
+      "learning_rate": 3.600677806466507e-07,
+      "loss": 0.3194,
+      "step": 47420
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.440585903857065,
+      "learning_rate": 3.594573112009803e-07,
+      "loss": 0.3292,
+      "step": 47425
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.925003167153565,
+      "learning_rate": 3.5884734040210147e-07,
+      "loss": 0.34,
+      "step": 47430
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.586192778987076,
+      "learning_rate": 3.5823786831555906e-07,
+      "loss": 0.337,
+      "step": 47435
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.935545151503379,
+      "learning_rate": 3.5762889500684294e-07,
+      "loss": 0.3562,
+      "step": 47440
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.423750760500127,
+      "learning_rate": 3.570204205413913e-07,
+      "loss": 0.284,
+      "step": 47445
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.483843025042958,
+      "learning_rate": 3.564124449845857e-07,
+      "loss": 0.3167,
+      "step": 47450
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.115499035227642,
+      "learning_rate": 3.558049684017573e-07,
+      "loss": 0.3542,
+      "step": 47455
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.597403206810282,
+      "learning_rate": 3.551979908581804e-07,
+      "loss": 0.3213,
+      "step": 47460
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.082674894256887,
+      "learning_rate": 3.5459151241907853e-07,
+      "loss": 0.3356,
+      "step": 47465
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.905002825773987,
+      "learning_rate": 3.539855331496211e-07,
+      "loss": 0.3163,
+      "step": 47470
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.285027227430392,
+      "learning_rate": 3.533800531149223e-07,
+      "loss": 0.3021,
+      "step": 47475
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.63205866538077,
+      "learning_rate": 3.527750723800444e-07,
+      "loss": 0.3138,
+      "step": 47480
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.489272684043733,
+      "learning_rate": 3.5217059100999487e-07,
+      "loss": 0.3194,
+      "step": 47485
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.9383592562869207,
+      "learning_rate": 3.5156660906972674e-07,
+      "loss": 0.2854,
+      "step": 47490
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.558018527068866,
+      "learning_rate": 3.50963126624142e-07,
+      "loss": 0.3072,
+      "step": 47495
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.4583540022292105,
+      "learning_rate": 3.5036014373808716e-07,
+      "loss": 0.3087,
+      "step": 47500
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.95993238449542,
+      "learning_rate": 3.497576604763553e-07,
+      "loss": 0.2955,
+      "step": 47505
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.438456015228428,
+      "learning_rate": 3.4915567690368634e-07,
+      "loss": 0.3387,
+      "step": 47510
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.490596650882147,
+      "learning_rate": 3.485541930847647e-07,
+      "loss": 0.3084,
+      "step": 47515
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.661525193352011,
+      "learning_rate": 3.479532090842241e-07,
+      "loss": 0.3273,
+      "step": 47520
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.502288950846946,
+      "learning_rate": 3.4735272496664197e-07,
+      "loss": 0.3447,
+      "step": 47525
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.2317737559166755,
+      "learning_rate": 3.4675274079654374e-07,
+      "loss": 0.3172,
+      "step": 47530
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.948338884048343,
+      "learning_rate": 3.461532566383985e-07,
+      "loss": 0.3162,
+      "step": 47535
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.63747731015127,
+      "learning_rate": 3.4555427255662633e-07,
+      "loss": 0.2946,
+      "step": 47540
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.812212196188589,
+      "learning_rate": 3.44955788615588e-07,
+      "loss": 0.3121,
+      "step": 47545
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 6.128273368307266,
+      "learning_rate": 3.4435780487959593e-07,
+      "loss": 0.3157,
+      "step": 47550
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.638519535583288,
+      "learning_rate": 3.4376032141290315e-07,
+      "loss": 0.3339,
+      "step": 47555
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.321372248864944,
+      "learning_rate": 3.4316333827971336e-07,
+      "loss": 0.3697,
+      "step": 47560
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.509442331848998,
+      "learning_rate": 3.4256685554417637e-07,
+      "loss": 0.3068,
+      "step": 47565
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.670117243748899,
+      "learning_rate": 3.4197087327038523e-07,
+      "loss": 0.3229,
+      "step": 47570
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.590114746808976,
+      "learning_rate": 3.4137539152238053e-07,
+      "loss": 0.3275,
+      "step": 47575
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.2817912196370935,
+      "learning_rate": 3.4078041036415097e-07,
+      "loss": 0.3199,
+      "step": 47580
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.471639898787827,
+      "learning_rate": 3.401859298596277e-07,
+      "loss": 0.3002,
+      "step": 47585
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.69528646125724,
+      "learning_rate": 3.3959195007269297e-07,
+      "loss": 0.338,
+      "step": 47590
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.715543890771803,
+      "learning_rate": 3.3899847106717063e-07,
+      "loss": 0.3418,
+      "step": 47595
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.217419483804426,
+      "learning_rate": 3.3840549290683313e-07,
+      "loss": 0.3275,
+      "step": 47600
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.790687169048269,
+      "learning_rate": 3.378130156553988e-07,
+      "loss": 0.3228,
+      "step": 47605
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.399420151552524,
+      "learning_rate": 3.372210393765313e-07,
+      "loss": 0.3275,
+      "step": 47610
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 6.143734271429322,
+      "learning_rate": 3.3662956413384187e-07,
+      "loss": 0.3386,
+      "step": 47615
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.7673449533142325,
+      "learning_rate": 3.3603858999088635e-07,
+      "loss": 0.3118,
+      "step": 47620
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.854963445528044,
+      "learning_rate": 3.354481170111684e-07,
+      "loss": 0.3279,
+      "step": 47625
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.078103501442932,
+      "learning_rate": 3.3485814525813675e-07,
+      "loss": 0.3434,
+      "step": 47630
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.034977945477632,
+      "learning_rate": 3.3426867479518563e-07,
+      "loss": 0.2907,
+      "step": 47635
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.383761195192665,
+      "learning_rate": 3.3367970568565665e-07,
+      "loss": 0.3415,
+      "step": 47640
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.386557697058935,
+      "learning_rate": 3.3309123799283804e-07,
+      "loss": 0.3412,
+      "step": 47645
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.333341839070483,
+      "learning_rate": 3.32503271779962e-07,
+      "loss": 0.3098,
+      "step": 47650
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.34362991618549,
+      "learning_rate": 3.319158071102091e-07,
+      "loss": 0.3106,
+      "step": 47655
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.254660178129415,
+      "learning_rate": 3.313288440467044e-07,
+      "loss": 0.3172,
+      "step": 47660
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.0917619833565775,
+      "learning_rate": 3.307423826525202e-07,
+      "loss": 0.3005,
+      "step": 47665
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.117232752981816,
+      "learning_rate": 3.301564229906729e-07,
+      "loss": 0.3165,
+      "step": 47670
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.57774333667238,
+      "learning_rate": 3.2957096512412856e-07,
+      "loss": 0.337,
+      "step": 47675
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.408126168209125,
+      "learning_rate": 3.2898600911579646e-07,
+      "loss": 0.3365,
+      "step": 47680
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.889320483711481,
+      "learning_rate": 3.2840155502853133e-07,
+      "loss": 0.3189,
+      "step": 47685
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.362285259304799,
+      "learning_rate": 3.278176029251373e-07,
+      "loss": 0.354,
+      "step": 47690
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.8069990841662324,
+      "learning_rate": 3.272341528683609e-07,
+      "loss": 0.3232,
+      "step": 47695
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.26464172603452,
+      "learning_rate": 3.266512049208981e-07,
+      "loss": 0.32,
+      "step": 47700
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.565820319698187,
+      "learning_rate": 3.260687591453876e-07,
+      "loss": 0.3212,
+      "step": 47705
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.741254672867791,
+      "learning_rate": 3.2548681560441775e-07,
+      "loss": 0.3151,
+      "step": 47710
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.563889335836407,
+      "learning_rate": 3.2490537436051905e-07,
+      "loss": 0.2942,
+      "step": 47715
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.133163083253335,
+      "learning_rate": 3.2432443547617045e-07,
+      "loss": 0.3355,
+      "step": 47720
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.16777636294225,
+      "learning_rate": 3.2374399901379696e-07,
+      "loss": 0.348,
+      "step": 47725
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.777794135019218,
+      "learning_rate": 3.2316406503576816e-07,
+      "loss": 0.3163,
+      "step": 47730
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.558942667410501,
+      "learning_rate": 3.2258463360440083e-07,
+      "loss": 0.3191,
+      "step": 47735
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.955876311305226,
+      "learning_rate": 3.2200570478195857e-07,
+      "loss": 0.3727,
+      "step": 47740
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.512120144335555,
+      "learning_rate": 3.214272786306483e-07,
+      "loss": 0.3147,
+      "step": 47745
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.178539157244023,
+      "learning_rate": 3.208493552126257e-07,
+      "loss": 0.3429,
+      "step": 47750
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.697660185506693,
+      "learning_rate": 3.202719345899896e-07,
+      "loss": 0.3188,
+      "step": 47755
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.448063930221362,
+      "learning_rate": 3.1969501682478807e-07,
+      "loss": 0.3018,
+      "step": 47760
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.3512123122337805,
+      "learning_rate": 3.1911860197901267e-07,
+      "loss": 0.3118,
+      "step": 47765
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.9023805265563696,
+      "learning_rate": 3.185426901146016e-07,
+      "loss": 0.2779,
+      "step": 47770
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.751318911524001,
+      "learning_rate": 3.179672812934398e-07,
+      "loss": 0.3207,
+      "step": 47775
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.11648296728079,
+      "learning_rate": 3.1739237557735613e-07,
+      "loss": 0.3055,
+      "step": 47780
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.574882169294049,
+      "learning_rate": 3.168179730281279e-07,
+      "loss": 0.3356,
+      "step": 47785
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.027181941766636,
+      "learning_rate": 3.162440737074768e-07,
+      "loss": 0.3081,
+      "step": 47790
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.112720326862771,
+      "learning_rate": 3.156706776770718e-07,
+      "loss": 0.3308,
+      "step": 47795
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.473631153161822,
+      "learning_rate": 3.1509778499852597e-07,
+      "loss": 0.2996,
+      "step": 47800
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.23761028541806,
+      "learning_rate": 3.1452539573339826e-07,
+      "loss": 0.3251,
+      "step": 47805
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.861753624417564,
+      "learning_rate": 3.1395350994319617e-07,
+      "loss": 0.3309,
+      "step": 47810
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.10625851829245,
+      "learning_rate": 3.1338212768936947e-07,
+      "loss": 0.3801,
+      "step": 47815
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.485121295117052,
+      "learning_rate": 3.128112490333179e-07,
+      "loss": 0.315,
+      "step": 47820
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.166181052258781,
+      "learning_rate": 3.12240874036383e-07,
+      "loss": 0.3275,
+      "step": 47825
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.497452351302918,
+      "learning_rate": 3.1167100275985464e-07,
+      "loss": 0.3266,
+      "step": 47830
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.630048357828678,
+      "learning_rate": 3.111016352649687e-07,
+      "loss": 0.3196,
+      "step": 47835
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.718753036645414,
+      "learning_rate": 3.1053277161290585e-07,
+      "loss": 0.3036,
+      "step": 47840
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.262779770629185,
+      "learning_rate": 3.0996441186479265e-07,
+      "loss": 0.3289,
+      "step": 47845
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.573369072323231,
+      "learning_rate": 3.0939655608170306e-07,
+      "loss": 0.3034,
+      "step": 47850
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.434752011193726,
+      "learning_rate": 3.0882920432465324e-07,
+      "loss": 0.3272,
+      "step": 47855
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.4238194974092595,
+      "learning_rate": 3.0826235665461004e-07,
+      "loss": 0.3355,
+      "step": 47860
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.945574492558196,
+      "learning_rate": 3.0769601313248244e-07,
+      "loss": 0.2991,
+      "step": 47865
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.430777388168083,
+      "learning_rate": 3.0713017381912736e-07,
+      "loss": 0.3396,
+      "step": 47870
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.468089592102152,
+      "learning_rate": 3.0656483877534557e-07,
+      "loss": 0.3102,
+      "step": 47875
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.902235614015058,
+      "learning_rate": 3.0600000806188633e-07,
+      "loss": 0.3292,
+      "step": 47880
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.368090791078648,
+      "learning_rate": 3.0543568173944273e-07,
+      "loss": 0.296,
+      "step": 47885
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.083264614720332,
+      "learning_rate": 3.0487185986865285e-07,
+      "loss": 0.3502,
+      "step": 47890
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.317452748816307,
+      "learning_rate": 3.043085425101033e-07,
+      "loss": 0.3234,
+      "step": 47895
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.408397455549362,
+      "learning_rate": 3.0374572972432405e-07,
+      "loss": 0.314,
+      "step": 47900
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.5821145538223345,
+      "learning_rate": 3.0318342157179326e-07,
+      "loss": 0.3383,
+      "step": 47905
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.343002499016802,
+      "learning_rate": 3.0262161811293155e-07,
+      "loss": 0.3213,
+      "step": 47910
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.287329186077056,
+      "learning_rate": 3.020603194081084e-07,
+      "loss": 0.2996,
+      "step": 47915
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.71053580808724,
+      "learning_rate": 3.0149952551763726e-07,
+      "loss": 0.311,
+      "step": 47920
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.076330986299103,
+      "learning_rate": 3.0093923650177827e-07,
+      "loss": 0.3253,
+      "step": 47925
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.483208349833166,
+      "learning_rate": 3.0037945242073706e-07,
+      "loss": 0.3263,
+      "step": 47930
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.786764538160286,
+      "learning_rate": 2.998201733346645e-07,
+      "loss": 0.3244,
+      "step": 47935
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.341878249099872,
+      "learning_rate": 2.9926139930365807e-07,
+      "loss": 0.3228,
+      "step": 47940
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.6848355866544065,
+      "learning_rate": 2.987031303877602e-07,
+      "loss": 0.3238,
+      "step": 47945
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.875822688048872,
+      "learning_rate": 2.9814536664695914e-07,
+      "loss": 0.327,
+      "step": 47950
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.128496690051945,
+      "learning_rate": 2.9758810814118967e-07,
+      "loss": 0.3279,
+      "step": 47955
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.502126110869699,
+      "learning_rate": 2.9703135493033053e-07,
+      "loss": 0.3209,
+      "step": 47960
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.372510229707003,
+      "learning_rate": 2.964751070742089e-07,
+      "loss": 0.3188,
+      "step": 47965
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.99897515610518,
+      "learning_rate": 2.9591936463259585e-07,
+      "loss": 0.3095,
+      "step": 47970
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.836799461654691,
+      "learning_rate": 2.953641276652064e-07,
+      "loss": 0.3462,
+      "step": 47975
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.0562436096974,
+      "learning_rate": 2.94809396231705e-07,
+      "loss": 0.3292,
+      "step": 47980
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.5157550259236485,
+      "learning_rate": 2.942551703916996e-07,
+      "loss": 0.3549,
+      "step": 47985
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.155132591518107,
+      "learning_rate": 2.937014502047442e-07,
+      "loss": 0.353,
+      "step": 47990
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.4462293514944955,
+      "learning_rate": 2.931482357303389e-07,
+      "loss": 0.3202,
+      "step": 47995
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.008498180222657,
+      "learning_rate": 2.9259552702792793e-07,
+      "loss": 0.3432,
+      "step": 48000
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 6.013341013228123,
+      "learning_rate": 2.9204332415690264e-07,
+      "loss": 0.3249,
+      "step": 48005
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.294916659419839,
+      "learning_rate": 2.9149162717660107e-07,
+      "loss": 0.2977,
+      "step": 48010
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.596923682020863,
+      "learning_rate": 2.9094043614630305e-07,
+      "loss": 0.3387,
+      "step": 48015
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.340613792857137,
+      "learning_rate": 2.9038975112523847e-07,
+      "loss": 0.2886,
+      "step": 48020
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.588770766517988,
+      "learning_rate": 2.898395721725805e-07,
+      "loss": 0.297,
+      "step": 48025
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.636047363816947,
+      "learning_rate": 2.89289899347448e-07,
+      "loss": 0.3187,
+      "step": 48030
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.456184039149601,
+      "learning_rate": 2.8874073270890536e-07,
+      "loss": 0.2915,
+      "step": 48035
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.119085168184362,
+      "learning_rate": 2.8819207231596424e-07,
+      "loss": 0.34,
+      "step": 48040
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.550098220068273,
+      "learning_rate": 2.8764391822757867e-07,
+      "loss": 0.2936,
+      "step": 48045
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.840685867505534,
+      "learning_rate": 2.87096270502652e-07,
+      "loss": 0.335,
+      "step": 48050
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 8.596024096044914,
+      "learning_rate": 2.8654912920003106e-07,
+      "loss": 0.3078,
+      "step": 48055
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.39991037935613,
+      "learning_rate": 2.8600249437850725e-07,
+      "loss": 0.3083,
+      "step": 48060
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.66445646412765,
+      "learning_rate": 2.8545636609682017e-07,
+      "loss": 0.3247,
+      "step": 48065
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.939081279564444,
+      "learning_rate": 2.849107444136534e-07,
+      "loss": 0.3177,
+      "step": 48070
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.371676845073246,
+      "learning_rate": 2.8436562938763676e-07,
+      "loss": 0.2955,
+      "step": 48075
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.7915892877165485,
+      "learning_rate": 2.8382102107734513e-07,
+      "loss": 0.3347,
+      "step": 48080
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.900237168206849,
+      "learning_rate": 2.8327691954129823e-07,
+      "loss": 0.3351,
+      "step": 48085
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.6266733418582895,
+      "learning_rate": 2.827333248379632e-07,
+      "loss": 0.2983,
+      "step": 48090
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.536261795730333,
+      "learning_rate": 2.821902370257507e-07,
+      "loss": 0.3433,
+      "step": 48095
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.028518494385905,
+      "learning_rate": 2.8164765616301827e-07,
+      "loss": 0.3138,
+      "step": 48100
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.13318173852176,
+      "learning_rate": 2.8110558230806993e-07,
+      "loss": 0.3143,
+      "step": 48105
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.754033215845489,
+      "learning_rate": 2.805640155191519e-07,
+      "loss": 0.3252,
+      "step": 48110
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.318566146912634,
+      "learning_rate": 2.800229558544598e-07,
+      "loss": 0.3267,
+      "step": 48115
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.6740527822773155,
+      "learning_rate": 2.7948240337213164e-07,
+      "loss": 0.3072,
+      "step": 48120
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.607471005702266,
+      "learning_rate": 2.789423581302525e-07,
+      "loss": 0.3046,
+      "step": 48125
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.386526630340396,
+      "learning_rate": 2.7840282018685215e-07,
+      "loss": 0.3458,
+      "step": 48130
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.719610208826684,
+      "learning_rate": 2.77863789599907e-07,
+      "loss": 0.3177,
+      "step": 48135
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.751449920202733,
+      "learning_rate": 2.773252664273385e-07,
+      "loss": 0.2937,
+      "step": 48140
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.935405482501159,
+      "learning_rate": 2.767872507270125e-07,
+      "loss": 0.3242,
+      "step": 48145
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.553882914254541,
+      "learning_rate": 2.7624974255674175e-07,
+      "loss": 0.3157,
+      "step": 48150
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.8498608330619675,
+      "learning_rate": 2.757127419742828e-07,
+      "loss": 0.316,
+      "step": 48155
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.544397331845315,
+      "learning_rate": 2.751762490373405e-07,
+      "loss": 0.319,
+      "step": 48160
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.354977104200328,
+      "learning_rate": 2.7464026380356277e-07,
+      "loss": 0.3192,
+      "step": 48165
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.785778652903822,
+      "learning_rate": 2.741047863305424e-07,
+      "loss": 0.3041,
+      "step": 48170
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.947356521333472,
+      "learning_rate": 2.735698166758205e-07,
+      "loss": 0.3401,
+      "step": 48175
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.265983571552172,
+      "learning_rate": 2.7303535489688004e-07,
+      "loss": 0.3404,
+      "step": 48180
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.923016936776703,
+      "learning_rate": 2.725014010511534e-07,
+      "loss": 0.35,
+      "step": 48185
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.29443219370602,
+      "learning_rate": 2.719679551960147e-07,
+      "loss": 0.2947,
+      "step": 48190
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.403594296346075,
+      "learning_rate": 2.714350173887853e-07,
+      "loss": 0.3307,
+      "step": 48195
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.832278401792976,
+      "learning_rate": 2.7090258768673274e-07,
+      "loss": 0.3229,
+      "step": 48200
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.291655696679214,
+      "learning_rate": 2.703706661470679e-07,
+      "loss": 0.2846,
+      "step": 48205
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.725478258645117,
+      "learning_rate": 2.6983925282694887e-07,
+      "loss": 0.337,
+      "step": 48210
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.636798919742124,
+      "learning_rate": 2.693083477834768e-07,
+      "loss": 0.3493,
+      "step": 48215
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.384085649462769,
+      "learning_rate": 2.6877795107370206e-07,
+      "loss": 0.2948,
+      "step": 48220
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.095443922903502,
+      "learning_rate": 2.682480627546169e-07,
+      "loss": 0.3046,
+      "step": 48225
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.340265699931569,
+      "learning_rate": 2.677186828831596e-07,
+      "loss": 0.3098,
+      "step": 48230
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.573530998144871,
+      "learning_rate": 2.671898115162158e-07,
+      "loss": 0.3356,
+      "step": 48235
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.553303339441197,
+      "learning_rate": 2.666614487106134e-07,
+      "loss": 0.2995,
+      "step": 48240
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.380895526287801,
+      "learning_rate": 2.661335945231286e-07,
+      "loss": 0.3202,
+      "step": 48245
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.645901369789934,
+      "learning_rate": 2.6560624901048103e-07,
+      "loss": 0.3077,
+      "step": 48250
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.9588769543741975,
+      "learning_rate": 2.650794122293365e-07,
+      "loss": 0.3044,
+      "step": 48255
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.7119343066023,
+      "learning_rate": 2.645530842363059e-07,
+      "loss": 0.2998,
+      "step": 48260
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.200997395871063,
+      "learning_rate": 2.640272650879455e-07,
+      "loss": 0.3053,
+      "step": 48265
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.358109505683691,
+      "learning_rate": 2.635019548407575e-07,
+      "loss": 0.2984,
+      "step": 48270
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.821216192556364,
+      "learning_rate": 2.6297715355118714e-07,
+      "loss": 0.3151,
+      "step": 48275
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.443574438783689,
+      "learning_rate": 2.624528612756283e-07,
+      "loss": 0.3325,
+      "step": 48280
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.514500606658843,
+      "learning_rate": 2.6192907807041756e-07,
+      "loss": 0.3345,
+      "step": 48285
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.763407786139578,
+      "learning_rate": 2.6140580399183824e-07,
+      "loss": 0.308,
+      "step": 48290
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.506694333749776,
+      "learning_rate": 2.6088303909611866e-07,
+      "loss": 0.2847,
+      "step": 48295
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.439390995243577,
+      "learning_rate": 2.603607834394312e-07,
+      "loss": 0.2956,
+      "step": 48300
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.328194427436979,
+      "learning_rate": 2.5983903707789583e-07,
+      "loss": 0.3656,
+      "step": 48305
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.788156999472909,
+      "learning_rate": 2.593178000675761e-07,
+      "loss": 0.315,
+      "step": 48310
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.622047510705419,
+      "learning_rate": 2.5879707246448003e-07,
+      "loss": 0.3451,
+      "step": 48315
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.675275570473539,
+      "learning_rate": 2.5827685432456394e-07,
+      "loss": 0.3276,
+      "step": 48320
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.65125097104725,
+      "learning_rate": 2.577571457037259e-07,
+      "loss": 0.3243,
+      "step": 48325
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.543791125320193,
+      "learning_rate": 2.572379466578123e-07,
+      "loss": 0.3808,
+      "step": 48330
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.341357498998987,
+      "learning_rate": 2.5671925724261193e-07,
+      "loss": 0.3199,
+      "step": 48335
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.640376502483177,
+      "learning_rate": 2.562010775138618e-07,
+      "loss": 0.2978,
+      "step": 48340
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.520983776948888,
+      "learning_rate": 2.5568340752724185e-07,
+      "loss": 0.3284,
+      "step": 48345
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.565842464269002,
+      "learning_rate": 2.5516624733837756e-07,
+      "loss": 0.3174,
+      "step": 48350
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.199199793015165,
+      "learning_rate": 2.546495970028412e-07,
+      "loss": 0.3123,
+      "step": 48355
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.0215539304949495,
+      "learning_rate": 2.5413345657614774e-07,
+      "loss": 0.3042,
+      "step": 48360
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.293481420370327,
+      "learning_rate": 2.536178261137601e-07,
+      "loss": 0.3513,
+      "step": 48365
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.107031913559453,
+      "learning_rate": 2.5310270567108385e-07,
+      "loss": 0.3289,
+      "step": 48370
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.21557193488729,
+      "learning_rate": 2.5258809530347204e-07,
+      "loss": 0.2998,
+      "step": 48375
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.784881851311318,
+      "learning_rate": 2.52073995066221e-07,
+      "loss": 0.3874,
+      "step": 48380
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.177736110934562,
+      "learning_rate": 2.5156040501457256e-07,
+      "loss": 0.3371,
+      "step": 48385
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.341872033030984,
+      "learning_rate": 2.5104732520371656e-07,
+      "loss": 0.337,
+      "step": 48390
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.691009906416703,
+      "learning_rate": 2.5053475568878384e-07,
+      "loss": 0.3476,
+      "step": 48395
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.9428860993574215,
+      "learning_rate": 2.50022696524852e-07,
+      "loss": 0.3407,
+      "step": 48400
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.584417431322471,
+      "learning_rate": 2.495111477669454e-07,
+      "loss": 0.3304,
+      "step": 48405
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.607442312863722,
+      "learning_rate": 2.4900010947003117e-07,
+      "loss": 0.3314,
+      "step": 48410
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.505455187839345,
+      "learning_rate": 2.4848958168902317e-07,
+      "loss": 0.3172,
+      "step": 48415
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.054524398874461,
+      "learning_rate": 2.479795644787786e-07,
+      "loss": 0.3252,
+      "step": 48420
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.75513974897599,
+      "learning_rate": 2.474700578941036e-07,
+      "loss": 0.297,
+      "step": 48425
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.251761658206857,
+      "learning_rate": 2.46961061989745e-07,
+      "loss": 0.3399,
+      "step": 48430
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.022015028529235,
+      "learning_rate": 2.4645257682039626e-07,
+      "loss": 0.3622,
+      "step": 48435
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.2767440878294325,
+      "learning_rate": 2.459446024406975e-07,
+      "loss": 0.329,
+      "step": 48440
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.817468601166096,
+      "learning_rate": 2.4543713890523237e-07,
+      "loss": 0.3575,
+      "step": 48445
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.741023448168125,
+      "learning_rate": 2.449301862685305e-07,
+      "loss": 0.3118,
+      "step": 48450
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.611835086952191,
+      "learning_rate": 2.4442374458506613e-07,
+      "loss": 0.3131,
+      "step": 48455
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.326487496709411,
+      "learning_rate": 2.439178139092574e-07,
+      "loss": 0.2935,
+      "step": 48460
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.878973820550783,
+      "learning_rate": 2.434123942954703e-07,
+      "loss": 0.3207,
+      "step": 48465
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.719651152630841,
+      "learning_rate": 2.42907485798014e-07,
+      "loss": 0.319,
+      "step": 48470
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.083885644675477,
+      "learning_rate": 2.424030884711426e-07,
+      "loss": 0.3317,
+      "step": 48475
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 6.191228762060517,
+      "learning_rate": 2.418992023690564e-07,
+      "loss": 0.3175,
+      "step": 48480
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.888713359558324,
+      "learning_rate": 2.4139582754589996e-07,
+      "loss": 0.3227,
+      "step": 48485
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.9695097719879,
+      "learning_rate": 2.4089296405576334e-07,
+      "loss": 0.3097,
+      "step": 48490
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.831692727842725,
+      "learning_rate": 2.403906119526811e-07,
+      "loss": 0.324,
+      "step": 48495
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.769783229227645,
+      "learning_rate": 2.3988877129063336e-07,
+      "loss": 0.3347,
+      "step": 48500
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.395280734211998,
+      "learning_rate": 2.393874421235448e-07,
+      "loss": 0.3117,
+      "step": 48505
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.613298053689998,
+      "learning_rate": 2.388866245052868e-07,
+      "loss": 0.3316,
+      "step": 48510
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.291076970437595,
+      "learning_rate": 2.3838631848967286e-07,
+      "loss": 0.3143,
+      "step": 48515
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.117887596077893,
+      "learning_rate": 2.3788652413046341e-07,
+      "loss": 0.3072,
+      "step": 48520
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.382365807835714,
+      "learning_rate": 2.3738724148136383e-07,
+      "loss": 0.3198,
+      "step": 48525
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.866468843543766,
+      "learning_rate": 2.3688847059602393e-07,
+      "loss": 0.3539,
+      "step": 48530
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.515002010174925,
+      "learning_rate": 2.3639021152803977e-07,
+      "loss": 0.3472,
+      "step": 48535
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.201456980862455,
+      "learning_rate": 2.358924643309507e-07,
+      "loss": 0.3177,
+      "step": 48540
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.648208394085241,
+      "learning_rate": 2.353952290582412e-07,
+      "loss": 0.3321,
+      "step": 48545
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.358863393218217,
+      "learning_rate": 2.3489850576334294e-07,
+      "loss": 0.2982,
+      "step": 48550
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.173404261867256,
+      "learning_rate": 2.3440229449962936e-07,
+      "loss": 0.3139,
+      "step": 48555
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.362149370807334,
+      "learning_rate": 2.339065953204217e-07,
+      "loss": 0.3312,
+      "step": 48560
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.400072583425098,
+      "learning_rate": 2.3341140827898512e-07,
+      "loss": 0.348,
+      "step": 48565
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.604628847225456,
+      "learning_rate": 2.3291673342852872e-07,
+      "loss": 0.3405,
+      "step": 48570
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.34353342388737,
+      "learning_rate": 2.3242257082220886e-07,
+      "loss": 0.3149,
+      "step": 48575
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.279829571817658,
+      "learning_rate": 2.3192892051312365e-07,
+      "loss": 0.3226,
+      "step": 48580
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.48094737447629,
+      "learning_rate": 2.3143578255432008e-07,
+      "loss": 0.3033,
+      "step": 48585
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.294683799143649,
+      "learning_rate": 2.3094315699878633e-07,
+      "loss": 0.3142,
+      "step": 48590
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.722491596409391,
+      "learning_rate": 2.3045104389945838e-07,
+      "loss": 0.2974,
+      "step": 48595
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.256774612443694,
+      "learning_rate": 2.2995944330921504e-07,
+      "loss": 0.3026,
+      "step": 48600
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.231845486959496,
+      "learning_rate": 2.294683552808813e-07,
+      "loss": 0.3165,
+      "step": 48605
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.10781775538336,
+      "learning_rate": 2.2897777986722713e-07,
+      "loss": 0.3125,
+      "step": 48610
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.513689272308508,
+      "learning_rate": 2.284877171209654e-07,
+      "loss": 0.3132,
+      "step": 48615
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.935012379172134,
+      "learning_rate": 2.2799816709475842e-07,
+      "loss": 0.3079,
+      "step": 48620
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.069317755187218,
+      "learning_rate": 2.2750912984120798e-07,
+      "loss": 0.3369,
+      "step": 48625
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.479800405668751,
+      "learning_rate": 2.2702060541286374e-07,
+      "loss": 0.3139,
+      "step": 48630
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.91373517562012,
+      "learning_rate": 2.2653259386222092e-07,
+      "loss": 0.317,
+      "step": 48635
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.191209904819061,
+      "learning_rate": 2.260450952417176e-07,
+      "loss": 0.3472,
+      "step": 48640
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.738003488603212,
+      "learning_rate": 2.255581096037379e-07,
+      "loss": 0.3352,
+      "step": 48645
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.462488744125328,
+      "learning_rate": 2.2507163700061007e-07,
+      "loss": 0.3288,
+      "step": 48650
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.460285896466044,
+      "learning_rate": 2.2458567748460834e-07,
+      "loss": 0.3027,
+      "step": 48655
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.835187356687008,
+      "learning_rate": 2.2410023110795154e-07,
+      "loss": 0.3217,
+      "step": 48660
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.4936032920551465,
+      "learning_rate": 2.2361529792280235e-07,
+      "loss": 0.3171,
+      "step": 48665
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.279379016726062,
+      "learning_rate": 2.2313087798126908e-07,
+      "loss": 0.2975,
+      "step": 48670
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.173206576804991,
+      "learning_rate": 2.2264697133540514e-07,
+      "loss": 0.3115,
+      "step": 48675
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.264923126794077,
+      "learning_rate": 2.2216357803720834e-07,
+      "loss": 0.3162,
+      "step": 48680
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.154763208221577,
+      "learning_rate": 2.2168069813862102e-07,
+      "loss": 0.3072,
+      "step": 48685
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 7.52416814448695,
+      "learning_rate": 2.211983316915306e-07,
+      "loss": 0.3169,
+      "step": 48690
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.8239161198660465,
+      "learning_rate": 2.2071647874777058e-07,
+      "loss": 0.3462,
+      "step": 48695
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.743538371289594,
+      "learning_rate": 2.2023513935911734e-07,
+      "loss": 0.3607,
+      "step": 48700
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.349279132536488,
+      "learning_rate": 2.197543135772928e-07,
+      "loss": 0.3328,
+      "step": 48705
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.75414314900954,
+      "learning_rate": 2.192740014539646e-07,
+      "loss": 0.3145,
+      "step": 48710
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.228703320654876,
+      "learning_rate": 2.1879420304074362e-07,
+      "loss": 0.3201,
+      "step": 48715
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.063793745033728,
+      "learning_rate": 2.1831491838918694e-07,
+      "loss": 0.2697,
+      "step": 48720
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.536008369701413,
+      "learning_rate": 2.1783614755079453e-07,
+      "loss": 0.3163,
+      "step": 48725
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.980949776874142,
+      "learning_rate": 2.1735789057701407e-07,
+      "loss": 0.3169,
+      "step": 48730
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.498869917530692,
+      "learning_rate": 2.168801475192356e-07,
+      "loss": 0.3127,
+      "step": 48735
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.935501703167351,
+      "learning_rate": 2.1640291842879468e-07,
+      "loss": 0.328,
+      "step": 48740
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.179056348410177,
+      "learning_rate": 2.1592620335697144e-07,
+      "loss": 0.3088,
+      "step": 48745
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.170036214389213,
+      "learning_rate": 2.1545000235499158e-07,
+      "loss": 0.3049,
+      "step": 48750
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.957812299908879,
+      "learning_rate": 2.1497431547402524e-07,
+      "loss": 0.2965,
+      "step": 48755
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.052972801970208,
+      "learning_rate": 2.14499142765186e-07,
+      "loss": 0.2887,
+      "step": 48760
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.790161194528763,
+      "learning_rate": 2.140244842795347e-07,
+      "loss": 0.3333,
+      "step": 48765
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.821468909260666,
+      "learning_rate": 2.135503400680744e-07,
+      "loss": 0.2937,
+      "step": 48770
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.477710390220459,
+      "learning_rate": 2.130767101817538e-07,
+      "loss": 0.304,
+      "step": 48775
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.8026257383268662,
+      "learning_rate": 2.1260359467146774e-07,
+      "loss": 0.2992,
+      "step": 48780
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.476622510171095,
+      "learning_rate": 2.121309935880528e-07,
+      "loss": 0.3191,
+      "step": 48785
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.753207872192619,
+      "learning_rate": 2.116589069822944e-07,
+      "loss": 0.3132,
+      "step": 48790
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.326528265772871,
+      "learning_rate": 2.1118733490491816e-07,
+      "loss": 0.3121,
+      "step": 48795
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.452515476611273,
+      "learning_rate": 2.1071627740659738e-07,
+      "loss": 0.2989,
+      "step": 48800
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.090399182091981,
+      "learning_rate": 2.1024573453794994e-07,
+      "loss": 0.3416,
+      "step": 48805
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.486045533116705,
+      "learning_rate": 2.097757063495365e-07,
+      "loss": 0.3273,
+      "step": 48810
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.786701399162015,
+      "learning_rate": 2.0930619289186438e-07,
+      "loss": 0.3166,
+      "step": 48815
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.578007534203883,
+      "learning_rate": 2.0883719421538496e-07,
+      "loss": 0.3507,
+      "step": 48820
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.6778620259830825,
+      "learning_rate": 2.08368710370494e-07,
+      "loss": 0.3054,
+      "step": 48825
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.672715960745678,
+      "learning_rate": 2.0790074140753181e-07,
+      "loss": 0.3208,
+      "step": 48830
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.316246102513153,
+      "learning_rate": 2.074332873767848e-07,
+      "loss": 0.2947,
+      "step": 48835
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.057774998230994,
+      "learning_rate": 2.0696634832848117e-07,
+      "loss": 0.3299,
+      "step": 48840
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.789371789010931,
+      "learning_rate": 2.0649992431279743e-07,
+      "loss": 0.3352,
+      "step": 48845
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.0122057332275025,
+      "learning_rate": 2.0603401537985235e-07,
+      "loss": 0.3315,
+      "step": 48850
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.324040746124454,
+      "learning_rate": 2.055686215797098e-07,
+      "loss": 0.3209,
+      "step": 48855
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.923613594591983,
+      "learning_rate": 2.0510374296237755e-07,
+      "loss": 0.3089,
+      "step": 48860
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.495983971185289,
+      "learning_rate": 2.0463937957781065e-07,
+      "loss": 0.3005,
+      "step": 48865
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.800563641183687,
+      "learning_rate": 2.0417553147590475e-07,
+      "loss": 0.3153,
+      "step": 48870
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.527806384209996,
+      "learning_rate": 2.0371219870650438e-07,
+      "loss": 0.3016,
+      "step": 48875
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.4945441170043114,
+      "learning_rate": 2.0324938131939588e-07,
+      "loss": 0.3009,
+      "step": 48880
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.329729661208378,
+      "learning_rate": 2.0278707936431053e-07,
+      "loss": 0.3187,
+      "step": 48885
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.839627579434529,
+      "learning_rate": 2.0232529289092583e-07,
+      "loss": 0.3242,
+      "step": 48890
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.548891180838907,
+      "learning_rate": 2.0186402194886156e-07,
+      "loss": 0.3208,
+      "step": 48895
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.540142339527456,
+      "learning_rate": 2.0140326658768473e-07,
+      "loss": 0.3337,
+      "step": 48900
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.807465281063605,
+      "learning_rate": 2.0094302685690402e-07,
+      "loss": 0.3054,
+      "step": 48905
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.38012020671858,
+      "learning_rate": 2.0048330280597495e-07,
+      "loss": 0.3184,
+      "step": 48910
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.4830538938930244,
+      "learning_rate": 2.0002409448429738e-07,
+      "loss": 0.3248,
+      "step": 48915
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.261476325984831,
+      "learning_rate": 1.995654019412141e-07,
+      "loss": 0.3565,
+      "step": 48920
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.579978223742383,
+      "learning_rate": 1.9910722522601455e-07,
+      "loss": 0.3226,
+      "step": 48925
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.892672366414126,
+      "learning_rate": 1.9864956438793215e-07,
+      "loss": 0.3424,
+      "step": 48930
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.572617588825795,
+      "learning_rate": 1.981924194761431e-07,
+      "loss": 0.3241,
+      "step": 48935
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.561838938066252,
+      "learning_rate": 1.9773579053977143e-07,
+      "loss": 0.3191,
+      "step": 48940
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.667196042380077,
+      "learning_rate": 1.9727967762788236e-07,
+      "loss": 0.2989,
+      "step": 48945
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.180949793823962,
+      "learning_rate": 1.9682408078948835e-07,
+      "loss": 0.3102,
+      "step": 48950
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.6479624764642145,
+      "learning_rate": 1.9636900007354464e-07,
+      "loss": 0.3413,
+      "step": 48955
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.571997812211011,
+      "learning_rate": 1.9591443552895273e-07,
+      "loss": 0.3337,
+      "step": 48960
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.376117436048435,
+      "learning_rate": 1.954603872045563e-07,
+      "loss": 0.2789,
+      "step": 48965
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.540570814512873,
+      "learning_rate": 1.9500685514914463e-07,
+      "loss": 0.3183,
+      "step": 48970
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.778227282715019,
+      "learning_rate": 1.9455383941145322e-07,
+      "loss": 0.3109,
+      "step": 48975
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.472267768248948,
+      "learning_rate": 1.9410134004015925e-07,
+      "loss": 0.3048,
+      "step": 48980
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.253835751740025,
+      "learning_rate": 1.936493570838871e-07,
+      "loss": 0.3057,
+      "step": 48985
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.312471739582353,
+      "learning_rate": 1.9319789059120298e-07,
+      "loss": 0.3459,
+      "step": 48990
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.629782722527147,
+      "learning_rate": 1.9274694061061972e-07,
+      "loss": 0.3007,
+      "step": 48995
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.44227520736774,
+      "learning_rate": 1.9229650719059412e-07,
+      "loss": 0.2945,
+      "step": 49000
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.620035652646511,
+      "learning_rate": 1.918465903795258e-07,
+      "loss": 0.331,
+      "step": 49005
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.5860921186999954,
+      "learning_rate": 1.913971902257622e-07,
+      "loss": 0.3342,
+      "step": 49010
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.6381365769760965,
+      "learning_rate": 1.909483067775919e-07,
+      "loss": 0.2948,
+      "step": 49015
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.669053813092479,
+      "learning_rate": 1.9049994008325022e-07,
+      "loss": 0.3256,
+      "step": 49020
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.011530251180718,
+      "learning_rate": 1.900520901909164e-07,
+      "loss": 0.2778,
+      "step": 49025
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.766175203791341,
+      "learning_rate": 1.8960475714871252e-07,
+      "loss": 0.3287,
+      "step": 49030
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.432442576871143,
+      "learning_rate": 1.8915794100470842e-07,
+      "loss": 0.3115,
+      "step": 49035
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.799932467394133,
+      "learning_rate": 1.8871164180691458e-07,
+      "loss": 0.3475,
+      "step": 49040
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.0604929545214405,
+      "learning_rate": 1.8826585960328936e-07,
+      "loss": 0.3452,
+      "step": 49045
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.67532811778708,
+      "learning_rate": 1.8782059444173327e-07,
+      "loss": 0.3077,
+      "step": 49050
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.018812750614052,
+      "learning_rate": 1.8737584637009143e-07,
+      "loss": 0.3176,
+      "step": 49055
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.718907937700562,
+      "learning_rate": 1.86931615436155e-07,
+      "loss": 0.3308,
+      "step": 49060
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.215585333702029,
+      "learning_rate": 1.8648790168765807e-07,
+      "loss": 0.3195,
+      "step": 49065
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.6316512024975225,
+      "learning_rate": 1.8604470517228024e-07,
+      "loss": 0.3455,
+      "step": 49070
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 41.47359500203935,
+      "learning_rate": 1.8560202593764343e-07,
+      "loss": 0.3519,
+      "step": 49075
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.400187130489294,
+      "learning_rate": 1.8515986403131736e-07,
+      "loss": 0.3255,
+      "step": 49080
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.302854837180305,
+      "learning_rate": 1.8471821950081347e-07,
+      "loss": 0.2863,
+      "step": 49085
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.912319645572431,
+      "learning_rate": 1.8427709239358772e-07,
+      "loss": 0.3328,
+      "step": 49090
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.7087066202501315,
+      "learning_rate": 1.8383648275704214e-07,
+      "loss": 0.3128,
+      "step": 49095
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.029776857488416,
+      "learning_rate": 1.8339639063852165e-07,
+      "loss": 0.2887,
+      "step": 49100
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.40376162963985,
+      "learning_rate": 1.829568160853168e-07,
+      "loss": 0.3137,
+      "step": 49105
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.992769460417173,
+      "learning_rate": 1.8251775914466086e-07,
+      "loss": 0.3242,
+      "step": 49110
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.448704637868842,
+      "learning_rate": 1.8207921986373223e-07,
+      "loss": 0.3034,
+      "step": 49115
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.127651118463407,
+      "learning_rate": 1.8164119828965542e-07,
+      "loss": 0.3707,
+      "step": 49120
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.571870026758502,
+      "learning_rate": 1.812036944694967e-07,
+      "loss": 0.3391,
+      "step": 49125
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.023194414765683,
+      "learning_rate": 1.8076670845026844e-07,
+      "loss": 0.2649,
+      "step": 49130
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.794355404408927,
+      "learning_rate": 1.8033024027892588e-07,
+      "loss": 0.3236,
+      "step": 49135
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.447515597431007,
+      "learning_rate": 1.7989429000236923e-07,
+      "loss": 0.3395,
+      "step": 49140
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.053976987911622,
+      "learning_rate": 1.7945885766744497e-07,
+      "loss": 0.3379,
+      "step": 49145
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.2742981381202245,
+      "learning_rate": 1.7902394332094008e-07,
+      "loss": 0.3287,
+      "step": 49150
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.835210739452109,
+      "learning_rate": 1.7858954700958996e-07,
+      "loss": 0.3457,
+      "step": 49155
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.215503775241933,
+      "learning_rate": 1.7815566878007064e-07,
+      "loss": 0.3217,
+      "step": 49160
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.584841483240674,
+      "learning_rate": 1.7772230867900587e-07,
+      "loss": 0.3237,
+      "step": 49165
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.654602953852498,
+      "learning_rate": 1.7728946675296122e-07,
+      "loss": 0.3368,
+      "step": 49170
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.831925840461089,
+      "learning_rate": 1.7685714304844726e-07,
+      "loss": 0.2665,
+      "step": 49175
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.466258043059292,
+      "learning_rate": 1.7642533761192016e-07,
+      "loss": 0.3556,
+      "step": 49180
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.825888510969101,
+      "learning_rate": 1.759940504897778e-07,
+      "loss": 0.3107,
+      "step": 49185
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.050732475009098,
+      "learning_rate": 1.7556328172836533e-07,
+      "loss": 0.2989,
+      "step": 49190
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.385419000452627,
+      "learning_rate": 1.7513303137396909e-07,
+      "loss": 0.3139,
+      "step": 49195
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.219793906782917,
+      "learning_rate": 1.7470329947282372e-07,
+      "loss": 0.3046,
+      "step": 49200
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.721609037698076,
+      "learning_rate": 1.742740860711034e-07,
+      "loss": 0.2973,
+      "step": 49205
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.771189694919438,
+      "learning_rate": 1.7384539121493016e-07,
+      "loss": 0.3297,
+      "step": 49210
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.635735068267216,
+      "learning_rate": 1.7341721495036933e-07,
+      "loss": 0.33,
+      "step": 49215
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.438209290682146,
+      "learning_rate": 1.7298955732343082e-07,
+      "loss": 0.2988,
+      "step": 49220
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.483691036687797,
+      "learning_rate": 1.7256241838006672e-07,
+      "loss": 0.334,
+      "step": 49225
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.242663895481422,
+      "learning_rate": 1.7213579816617588e-07,
+      "loss": 0.3316,
+      "step": 49230
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.365577246093222,
+      "learning_rate": 1.717096967276005e-07,
+      "loss": 0.3115,
+      "step": 49235
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.4489874026179255,
+      "learning_rate": 1.7128411411012792e-07,
+      "loss": 0.3319,
+      "step": 49240
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.4687936706397196,
+      "learning_rate": 1.7085905035948703e-07,
+      "loss": 0.324,
+      "step": 49245
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.92941646054833,
+      "learning_rate": 1.704345055213541e-07,
+      "loss": 0.2879,
+      "step": 49250
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.5929394896555715,
+      "learning_rate": 1.7001047964134875e-07,
+      "loss": 0.3413,
+      "step": 49255
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.680853375632767,
+      "learning_rate": 1.6958697276503288e-07,
+      "loss": 0.2759,
+      "step": 49260
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.380558261041858,
+      "learning_rate": 1.6916398493791507e-07,
+      "loss": 0.308,
+      "step": 49265
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.27707596732731,
+      "learning_rate": 1.6874151620544732e-07,
+      "loss": 0.3338,
+      "step": 49270
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.861723308646287,
+      "learning_rate": 1.6831956661302607e-07,
+      "loss": 0.2765,
+      "step": 49275
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.851889775601613,
+      "learning_rate": 1.6789813620599116e-07,
+      "loss": 0.3043,
+      "step": 49280
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.758076018914188,
+      "learning_rate": 1.674772250296264e-07,
+      "loss": 0.3296,
+      "step": 49285
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.495918493541164,
+      "learning_rate": 1.6705683312916166e-07,
+      "loss": 0.309,
+      "step": 49290
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.611089598335909,
+      "learning_rate": 1.6663696054976975e-07,
+      "loss": 0.3084,
+      "step": 49295
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.339235409887899,
+      "learning_rate": 1.6621760733656734e-07,
+      "loss": 0.3135,
+      "step": 49300
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.342331312679037,
+      "learning_rate": 1.6579877353461616e-07,
+      "loss": 0.3142,
+      "step": 49305
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.24058743942037,
+      "learning_rate": 1.6538045918892243e-07,
+      "loss": 0.3111,
+      "step": 49310
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.938797914048927,
+      "learning_rate": 1.649626643444352e-07,
+      "loss": 0.3406,
+      "step": 49315
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.456487517168583,
+      "learning_rate": 1.6454538904604744e-07,
+      "loss": 0.3171,
+      "step": 49320
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.485593420063372,
+      "learning_rate": 1.6412863333859885e-07,
+      "loss": 0.3154,
+      "step": 49325
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.279469998994342,
+      "learning_rate": 1.6371239726687082e-07,
+      "loss": 0.3019,
+      "step": 49330
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.258712483394506,
+      "learning_rate": 1.632966808755898e-07,
+      "loss": 0.2986,
+      "step": 49335
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.3646259616186995,
+      "learning_rate": 1.6288148420942672e-07,
+      "loss": 0.313,
+      "step": 49340
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.689260127423614,
+      "learning_rate": 1.6246680731299536e-07,
+      "loss": 0.3125,
+      "step": 49345
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.383301993976913,
+      "learning_rate": 1.620526502308556e-07,
+      "loss": 0.3004,
+      "step": 49350
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.503178423014603,
+      "learning_rate": 1.616390130075096e-07,
+      "loss": 0.3835,
+      "step": 49355
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.374325781219972,
+      "learning_rate": 1.6122589568740576e-07,
+      "loss": 0.3118,
+      "step": 49360
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.4613099149958195,
+      "learning_rate": 1.608132983149341e-07,
+      "loss": 0.3093,
+      "step": 49365
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.004622784835653,
+      "learning_rate": 1.6040122093443022e-07,
+      "loss": 0.3362,
+      "step": 49370
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.956152039375732,
+      "learning_rate": 1.5998966359017488e-07,
+      "loss": 0.338,
+      "step": 49375
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.690217694208695,
+      "learning_rate": 1.5957862632638987e-07,
+      "loss": 0.3198,
+      "step": 49380
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.690957375560183,
+      "learning_rate": 1.591681091872438e-07,
+      "loss": 0.2817,
+      "step": 49385
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.003496220778562,
+      "learning_rate": 1.5875811221684912e-07,
+      "loss": 0.3261,
+      "step": 49390
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.545500669800517,
+      "learning_rate": 1.583486354592606e-07,
+      "loss": 0.2923,
+      "step": 49395
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.203199731145849,
+      "learning_rate": 1.5793967895848027e-07,
+      "loss": 0.3127,
+      "step": 49400
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.576779035060099,
+      "learning_rate": 1.5753124275844967e-07,
+      "loss": 0.2906,
+      "step": 49405
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.00853311597256,
+      "learning_rate": 1.5712332690305977e-07,
+      "loss": 0.3344,
+      "step": 49410
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.09984843274896,
+      "learning_rate": 1.567159314361405e-07,
+      "loss": 0.3177,
+      "step": 49415
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.731019089443841,
+      "learning_rate": 1.563090564014702e-07,
+      "loss": 0.33,
+      "step": 49420
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.699799571044952,
+      "learning_rate": 1.559027018427689e-07,
+      "loss": 0.3077,
+      "step": 49425
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.915617021912783,
+      "learning_rate": 1.5549686780369999e-07,
+      "loss": 0.3241,
+      "step": 49430
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.445358027809136,
+      "learning_rate": 1.5509155432787414e-07,
+      "loss": 0.3111,
+      "step": 49435
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.8622196114555045,
+      "learning_rate": 1.5468676145884208e-07,
+      "loss": 0.3391,
+      "step": 49440
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.794898668313322,
+      "learning_rate": 1.5428248924010182e-07,
+      "loss": 0.3349,
+      "step": 49445
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.148367879965321,
+      "learning_rate": 1.538787377150941e-07,
+      "loss": 0.3473,
+      "step": 49450
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.018720295060291,
+      "learning_rate": 1.5347550692720313e-07,
+      "loss": 0.3127,
+      "step": 49455
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.126032955709537,
+      "learning_rate": 1.5307279691975873e-07,
+      "loss": 0.3122,
+      "step": 49460
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.645274597456493,
+      "learning_rate": 1.5267060773603294e-07,
+      "loss": 0.3082,
+      "step": 49465
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.539644192415214,
+      "learning_rate": 1.5226893941924392e-07,
+      "loss": 0.3386,
+      "step": 49470
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.8818796766097154,
+      "learning_rate": 1.5186779201255108e-07,
+      "loss": 0.2992,
+      "step": 49475
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.664261169459152,
+      "learning_rate": 1.514671655590605e-07,
+      "loss": 0.3001,
+      "step": 49480
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.481285845733247,
+      "learning_rate": 1.5106706010182214e-07,
+      "loss": 0.3058,
+      "step": 49485
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.316575255719252,
+      "learning_rate": 1.5066747568382723e-07,
+      "loss": 0.3043,
+      "step": 49490
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.342556509589545,
+      "learning_rate": 1.502684123480147e-07,
+      "loss": 0.3385,
+      "step": 49495
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.764862801595888,
+      "learning_rate": 1.4986987013726363e-07,
+      "loss": 0.319,
+      "step": 49500
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.182945761708489,
+      "learning_rate": 1.4947184909440138e-07,
+      "loss": 0.341,
+      "step": 49505
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.218815957627943,
+      "learning_rate": 1.4907434926219545e-07,
+      "loss": 0.3083,
+      "step": 49510
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.512042894996303,
+      "learning_rate": 1.4867737068335941e-07,
+      "loss": 0.3451,
+      "step": 49515
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.04880001576011,
+      "learning_rate": 1.4828091340055085e-07,
+      "loss": 0.3417,
+      "step": 49520
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.874911424061834,
+      "learning_rate": 1.4788497745636955e-07,
+      "loss": 0.3032,
+      "step": 49525
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.53295590430745,
+      "learning_rate": 1.4748956289336203e-07,
+      "loss": 0.2939,
+      "step": 49530
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.210848335554472,
+      "learning_rate": 1.4709466975401708e-07,
+      "loss": 0.2776,
+      "step": 49535
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.4851897229701,
+      "learning_rate": 1.4670029808076634e-07,
+      "loss": 0.3267,
+      "step": 49540
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.5599165578640894,
+      "learning_rate": 1.463064479159887e-07,
+      "loss": 0.3307,
+      "step": 49545
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.367003965898049,
+      "learning_rate": 1.459131193020036e-07,
+      "loss": 0.3101,
+      "step": 49550
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.9377164708803516,
+      "learning_rate": 1.455203122810772e-07,
+      "loss": 0.2938,
+      "step": 49555
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.61850378796758,
+      "learning_rate": 1.451280268954175e-07,
+      "loss": 0.3013,
+      "step": 49560
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.324312135203147,
+      "learning_rate": 1.4473626318717737e-07,
+      "loss": 0.3012,
+      "step": 49565
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.807004375756143,
+      "learning_rate": 1.4434502119845319e-07,
+      "loss": 0.2695,
+      "step": 49570
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.920628767107733,
+      "learning_rate": 1.4395430097128627e-07,
+      "loss": 0.2959,
+      "step": 49575
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.364254147624005,
+      "learning_rate": 1.4356410254766196e-07,
+      "loss": 0.3229,
+      "step": 49580
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.088700187800069,
+      "learning_rate": 1.4317442596950726e-07,
+      "loss": 0.3166,
+      "step": 49585
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.216262749120216,
+      "learning_rate": 1.4278527127869535e-07,
+      "loss": 0.3044,
+      "step": 49590
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.014721378114283,
+      "learning_rate": 1.4239663851704277e-07,
+      "loss": 0.2915,
+      "step": 49595
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.690237564984018,
+      "learning_rate": 1.4200852772630946e-07,
+      "loss": 0.3207,
+      "step": 49600
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.391227977336331,
+      "learning_rate": 1.4162093894819984e-07,
+      "loss": 0.289,
+      "step": 49605
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.051753165624551,
+      "learning_rate": 1.4123387222436168e-07,
+      "loss": 0.2867,
+      "step": 49610
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.724103774928652,
+      "learning_rate": 1.4084732759638786e-07,
+      "loss": 0.2793,
+      "step": 49615
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.668499657819339,
+      "learning_rate": 1.4046130510581403e-07,
+      "loss": 0.3097,
+      "step": 49620
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.669216058746535,
+      "learning_rate": 1.4007580479411863e-07,
+      "loss": 0.3034,
+      "step": 49625
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.937021637540202,
+      "learning_rate": 1.396908267027275e-07,
+      "loss": 0.3261,
+      "step": 49630
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.467460570511647,
+      "learning_rate": 1.3930637087300635e-07,
+      "loss": 0.3193,
+      "step": 49635
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.883384265978267,
+      "learning_rate": 1.3892243734626831e-07,
+      "loss": 0.3153,
+      "step": 49640
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.8550942496126055,
+      "learning_rate": 1.3853902616376703e-07,
+      "loss": 0.3356,
+      "step": 49645
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.3256939303010205,
+      "learning_rate": 1.3815613736670397e-07,
+      "loss": 0.3203,
+      "step": 49650
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.881945932100836,
+      "learning_rate": 1.3777377099621958e-07,
+      "loss": 0.3214,
+      "step": 49655
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.33756910057617,
+      "learning_rate": 1.3739192709340321e-07,
+      "loss": 0.3182,
+      "step": 49660
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.679228348160198,
+      "learning_rate": 1.370106056992837e-07,
+      "loss": 0.3017,
+      "step": 49665
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.820551716831073,
+      "learning_rate": 1.3662980685483662e-07,
+      "loss": 0.2874,
+      "step": 49670
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.277810315998395,
+      "learning_rate": 1.3624953060098145e-07,
+      "loss": 0.3168,
+      "step": 49675
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.091580304351212,
+      "learning_rate": 1.3586977697857882e-07,
+      "loss": 0.3188,
+      "step": 49680
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.907577781622324,
+      "learning_rate": 1.3549054602843558e-07,
+      "loss": 0.3495,
+      "step": 49685
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.410263128612622,
+      "learning_rate": 1.3511183779130243e-07,
+      "loss": 0.3316,
+      "step": 49690
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.517350551085498,
+      "learning_rate": 1.347336523078724e-07,
+      "loss": 0.2913,
+      "step": 49695
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.555476506318306,
+      "learning_rate": 1.3435598961878404e-07,
+      "loss": 0.3228,
+      "step": 49700
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.866765603990417,
+      "learning_rate": 1.3397884976461716e-07,
+      "loss": 0.3171,
+      "step": 49705
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.8702380224929875,
+      "learning_rate": 1.3360223278589935e-07,
+      "loss": 0.3126,
+      "step": 49710
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.811413132990187,
+      "learning_rate": 1.3322613872309874e-07,
+      "loss": 0.2999,
+      "step": 49715
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.271145464293548,
+      "learning_rate": 1.328505676166275e-07,
+      "loss": 0.3083,
+      "step": 49720
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.5236074869459,
+      "learning_rate": 1.3247551950684334e-07,
+      "loss": 0.334,
+      "step": 49725
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.34227982709005,
+      "learning_rate": 1.321009944340468e-07,
+      "loss": 0.3082,
+      "step": 49730
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.538327848443762,
+      "learning_rate": 1.3172699243848174e-07,
+      "loss": 0.2933,
+      "step": 49735
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.385324363870202,
+      "learning_rate": 1.313535135603372e-07,
+      "loss": 0.3179,
+      "step": 49740
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.608675996383441,
+      "learning_rate": 1.309805578397444e-07,
+      "loss": 0.2889,
+      "step": 49745
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.448251701218472,
+      "learning_rate": 1.3060812531677902e-07,
+      "loss": 0.3121,
+      "step": 49750
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.425920239944105,
+      "learning_rate": 1.302362160314613e-07,
+      "loss": 0.3083,
+      "step": 49755
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.530434310691104,
+      "learning_rate": 1.2986483002375316e-07,
+      "loss": 0.2997,
+      "step": 49760
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.525187526324167,
+      "learning_rate": 1.294939673335638e-07,
+      "loss": 0.301,
+      "step": 49765
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.210430002668408,
+      "learning_rate": 1.2912362800074185e-07,
+      "loss": 0.32,
+      "step": 49770
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.765892537710484,
+      "learning_rate": 1.287538120650833e-07,
+      "loss": 0.3037,
+      "step": 49775
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.444989167541033,
+      "learning_rate": 1.283845195663258e-07,
+      "loss": 0.2965,
+      "step": 49780
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.08504032953018,
+      "learning_rate": 1.2801575054415262e-07,
+      "loss": 0.2916,
+      "step": 49785
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.452497176596786,
+      "learning_rate": 1.2764750503818756e-07,
+      "loss": 0.3068,
+      "step": 49790
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.493273299309027,
+      "learning_rate": 1.2727978308800237e-07,
+      "loss": 0.3258,
+      "step": 49795
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.738191199702944,
+      "learning_rate": 1.269125847331093e-07,
+      "loss": 0.3204,
+      "step": 49800
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.32332726726489,
+      "learning_rate": 1.2654591001296513e-07,
+      "loss": 0.3149,
+      "step": 49805
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.2361901713688175,
+      "learning_rate": 1.2617975896697165e-07,
+      "loss": 0.302,
+      "step": 49810
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.346275468583683,
+      "learning_rate": 1.2581413163447243e-07,
+      "loss": 0.3074,
+      "step": 49815
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.000187465023883,
+      "learning_rate": 1.2544902805475657e-07,
+      "loss": 0.307,
+      "step": 49820
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.387719405463538,
+      "learning_rate": 1.2508444826705545e-07,
+      "loss": 0.3053,
+      "step": 49825
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.6913978646148995,
+      "learning_rate": 1.24720392310545e-07,
+      "loss": 0.308,
+      "step": 49830
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.840259371819478,
+      "learning_rate": 1.2435686022434446e-07,
+      "loss": 0.358,
+      "step": 49835
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.471284889555268,
+      "learning_rate": 1.2399385204751702e-07,
+      "loss": 0.3018,
+      "step": 49840
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.426073064427697,
+      "learning_rate": 1.236313678190698e-07,
+      "loss": 0.3217,
+      "step": 49845
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.578297894204973,
+      "learning_rate": 1.232694075779539e-07,
+      "loss": 0.3089,
+      "step": 49850
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.269928237123571,
+      "learning_rate": 1.2290797136306154e-07,
+      "loss": 0.3162,
+      "step": 49855
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.69279024555451,
+      "learning_rate": 1.2254705921323329e-07,
+      "loss": 0.3412,
+      "step": 49860
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.293566881388998,
+      "learning_rate": 1.2218667116724814e-07,
+      "loss": 0.323,
+      "step": 49865
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.383699298284824,
+      "learning_rate": 1.2182680726383344e-07,
+      "loss": 0.2978,
+      "step": 49870
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.379892479437391,
+      "learning_rate": 1.2146746754165662e-07,
+      "loss": 0.3105,
+      "step": 49875
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.370298698302884,
+      "learning_rate": 1.2110865203933175e-07,
+      "loss": 0.3196,
+      "step": 49880
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.849092078290703,
+      "learning_rate": 1.2075036079541413e-07,
+      "loss": 0.3048,
+      "step": 49885
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.735955032794627,
+      "learning_rate": 1.2039259384840407e-07,
+      "loss": 0.3441,
+      "step": 49890
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.994709702676883,
+      "learning_rate": 1.2003535123674526e-07,
+      "loss": 0.346,
+      "step": 49895
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.038490161528967,
+      "learning_rate": 1.196786329988242e-07,
+      "loss": 0.3123,
+      "step": 49900
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 7.386741008813954,
+      "learning_rate": 1.1932243917297304e-07,
+      "loss": 0.311,
+      "step": 49905
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.751250148862671,
+      "learning_rate": 1.1896676979746613e-07,
+      "loss": 0.3255,
+      "step": 49910
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.343921331576727,
+      "learning_rate": 1.1861162491052069e-07,
+      "loss": 0.3684,
+      "step": 49915
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.897884985248466,
+      "learning_rate": 1.1825700455030009e-07,
+      "loss": 0.3058,
+      "step": 49920
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.617782399359281,
+      "learning_rate": 1.1790290875490884e-07,
+      "loss": 0.3025,
+      "step": 49925
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.331148135171444,
+      "learning_rate": 1.175493375623965e-07,
+      "loss": 0.3246,
+      "step": 49930
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.879795047588261,
+      "learning_rate": 1.1719629101075547e-07,
+      "loss": 0.3164,
+      "step": 49935
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.736021943573125,
+      "learning_rate": 1.1684376913792206e-07,
+      "loss": 0.3051,
+      "step": 49940
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.547630193154529,
+      "learning_rate": 1.1649177198177764e-07,
+      "loss": 0.2954,
+      "step": 49945
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.544386783902927,
+      "learning_rate": 1.1614029958014416e-07,
+      "loss": 0.351,
+      "step": 49950
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.8959323739397345,
+      "learning_rate": 1.1578935197078978e-07,
+      "loss": 0.3315,
+      "step": 49955
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.45870327068497,
+      "learning_rate": 1.1543892919142541e-07,
+      "loss": 0.2912,
+      "step": 49960
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.886514263568746,
+      "learning_rate": 1.1508903127970539e-07,
+      "loss": 0.3026,
+      "step": 49965
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.763516862907151,
+      "learning_rate": 1.1473965827322742e-07,
+      "loss": 0.313,
+      "step": 49970
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.5561197614542195,
+      "learning_rate": 1.1439081020953368e-07,
+      "loss": 0.3288,
+      "step": 49975
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.456281693326938,
+      "learning_rate": 1.1404248712610922e-07,
+      "loss": 0.2934,
+      "step": 49980
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.52520700641327,
+      "learning_rate": 1.136946890603824e-07,
+      "loss": 0.3026,
+      "step": 49985
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.241907598125358,
+      "learning_rate": 1.1334741604972666e-07,
+      "loss": 0.3145,
+      "step": 49990
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.390998513873911,
+      "learning_rate": 1.1300066813145771e-07,
+      "loss": 0.3377,
+      "step": 49995
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.276889926816193,
+      "learning_rate": 1.1265444534283465e-07,
+      "loss": 0.36,
+      "step": 50000
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.7710719196745943,
+      "learning_rate": 1.1230874772106104e-07,
+      "loss": 0.3169,
+      "step": 50005
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.728281178786274,
+      "learning_rate": 1.1196357530328272e-07,
+      "loss": 0.3403,
+      "step": 50010
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.400948003221979,
+      "learning_rate": 1.116189281265917e-07,
+      "loss": 0.3041,
+      "step": 50015
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.822692472062315,
+      "learning_rate": 1.1127480622802056e-07,
+      "loss": 0.2699,
+      "step": 50020
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.452412220408982,
+      "learning_rate": 1.109312096445475e-07,
+      "loss": 0.2988,
+      "step": 50025
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.304613634028073,
+      "learning_rate": 1.1058813841309246e-07,
+      "loss": 0.3306,
+      "step": 50030
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.549176845585562,
+      "learning_rate": 1.1024559257052036e-07,
+      "loss": 0.3103,
+      "step": 50035
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.5313067796554245,
+      "learning_rate": 1.0990357215364067e-07,
+      "loss": 0.3472,
+      "step": 50040
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.262994160246877,
+      "learning_rate": 1.0956207719920231e-07,
+      "loss": 0.3213,
+      "step": 50045
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.934739579450185,
+      "learning_rate": 1.0922110774390315e-07,
+      "loss": 0.3274,
+      "step": 50050
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.058559541882657,
+      "learning_rate": 1.0888066382438056e-07,
+      "loss": 0.2915,
+      "step": 50055
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.914887366797448,
+      "learning_rate": 1.0854074547721583e-07,
+      "loss": 0.3336,
+      "step": 50060
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.565490339689186,
+      "learning_rate": 1.082013527389364e-07,
+      "loss": 0.3224,
+      "step": 50065
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.869785079152946,
+      "learning_rate": 1.0786248564601031e-07,
+      "loss": 0.3075,
+      "step": 50070
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.418472574847041,
+      "learning_rate": 1.0752414423485125e-07,
+      "loss": 0.2944,
+      "step": 50075
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.437129635372558,
+      "learning_rate": 1.071863285418151e-07,
+      "loss": 0.3191,
+      "step": 50080
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.648109498374428,
+      "learning_rate": 1.0684903860320117e-07,
+      "loss": 0.3547,
+      "step": 50085
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.490040844108931,
+      "learning_rate": 1.065122744552538e-07,
+      "loss": 0.3147,
+      "step": 50090
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.036602894296083,
+      "learning_rate": 1.061760361341585e-07,
+      "loss": 0.3318,
+      "step": 50095
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.197471938487059,
+      "learning_rate": 1.0584032367604636e-07,
+      "loss": 0.3113,
+      "step": 50100
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.6151931997693305,
+      "learning_rate": 1.0550513711699129e-07,
+      "loss": 0.2871,
+      "step": 50105
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.347851844019679,
+      "learning_rate": 1.0517047649300949e-07,
+      "loss": 0.327,
+      "step": 50110
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.412524011331123,
+      "learning_rate": 1.048363418400633e-07,
+      "loss": 0.3219,
+      "step": 50115
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.6523083585095,
+      "learning_rate": 1.0450273319405569e-07,
+      "loss": 0.3497,
+      "step": 50120
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.194009118699771,
+      "learning_rate": 1.0416965059083517e-07,
+      "loss": 0.3483,
+      "step": 50125
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.825721169687486,
+      "learning_rate": 1.0383709406619258e-07,
+      "loss": 0.3421,
+      "step": 50130
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.34509778362892,
+      "learning_rate": 1.0350506365586266e-07,
+      "loss": 0.317,
+      "step": 50135
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.429398368110269,
+      "learning_rate": 1.0317355939552354e-07,
+      "loss": 0.31,
+      "step": 50140
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.374883202619751,
+      "learning_rate": 1.0284258132079671e-07,
+      "loss": 0.3399,
+      "step": 50145
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.480022612470613,
+      "learning_rate": 1.0251212946724765e-07,
+      "loss": 0.3126,
+      "step": 50150
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.525072716123713,
+      "learning_rate": 1.0218220387038402e-07,
+      "loss": 0.3126,
+      "step": 50155
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.86095362396342,
+      "learning_rate": 1.0185280456565916e-07,
+      "loss": 0.3288,
+      "step": 50160
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.324954894412038,
+      "learning_rate": 1.0152393158846696e-07,
+      "loss": 0.314,
+      "step": 50165
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.124842077445845,
+      "learning_rate": 1.0119558497414694e-07,
+      "loss": 0.3308,
+      "step": 50170
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.028774692554803,
+      "learning_rate": 1.0086776475798144e-07,
+      "loss": 0.3266,
+      "step": 50175
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.359988857098868,
+      "learning_rate": 1.0054047097519615e-07,
+      "loss": 0.335,
+      "step": 50180
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.753341481215494,
+      "learning_rate": 1.0021370366096017e-07,
+      "loss": 0.3254,
+      "step": 50185
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.104290093826002,
+      "learning_rate": 9.988746285038597e-08,
+      "loss": 0.3108,
+      "step": 50190
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.243322106856568,
+      "learning_rate": 9.956174857853052e-08,
+      "loss": 0.3075,
+      "step": 50195
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.263327233984905,
+      "learning_rate": 9.92365608803919e-08,
+      "loss": 0.2867,
+      "step": 50200
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.490589622086641,
+      "learning_rate": 9.891189979091332e-08,
+      "loss": 0.2977,
+      "step": 50205
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.216487071346811,
+      "learning_rate": 9.858776534498126e-08,
+      "loss": 0.2882,
+      "step": 50210
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.462840569198014,
+      "learning_rate": 9.82641575774257e-08,
+      "loss": 0.3089,
+      "step": 50215
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.284583532159085,
+      "learning_rate": 9.794107652301931e-08,
+      "loss": 0.3371,
+      "step": 50220
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.6185822898545315,
+      "learning_rate": 9.761852221647883e-08,
+      "loss": 0.2963,
+      "step": 50225
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.596465076660738,
+      "learning_rate": 9.729649469246372e-08,
+      "loss": 0.3246,
+      "step": 50230
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.963895889346747,
+      "learning_rate": 9.697499398557796e-08,
+      "loss": 0.3311,
+      "step": 50235
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.850941408885426,
+      "learning_rate": 9.665402013036728e-08,
+      "loss": 0.3294,
+      "step": 50240
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.692459297777503,
+      "learning_rate": 9.633357316132296e-08,
+      "loss": 0.2874,
+      "step": 50245
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.072748928491446,
+      "learning_rate": 9.601365311287746e-08,
+      "loss": 0.3128,
+      "step": 50250
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.471451359124489,
+      "learning_rate": 9.569426001940718e-08,
+      "loss": 0.3171,
+      "step": 50255
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.685336420093143,
+      "learning_rate": 9.53753939152341e-08,
+      "loss": 0.2928,
+      "step": 50260
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.6468380348736265,
+      "learning_rate": 9.505705483461969e-08,
+      "loss": 0.3119,
+      "step": 50265
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.98904304011332,
+      "learning_rate": 9.473924281177272e-08,
+      "loss": 0.3268,
+      "step": 50270
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.858420597418929,
+      "learning_rate": 9.442195788084196e-08,
+      "loss": 0.3224,
+      "step": 50275
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.872932590235281,
+      "learning_rate": 9.410520007592239e-08,
+      "loss": 0.2924,
+      "step": 50280
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.801675970462846,
+      "learning_rate": 9.378896943105009e-08,
+      "loss": 0.3019,
+      "step": 50285
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 3.978054949396103,
+      "learning_rate": 9.347326598020568e-08,
+      "loss": 0.305,
+      "step": 50290
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.741435531719537,
+      "learning_rate": 9.315808975731367e-08,
+      "loss": 0.3038,
+      "step": 50295
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.233291559055745,
+      "learning_rate": 9.284344079623974e-08,
+      "loss": 0.3248,
+      "step": 50300
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.910759586828501,
+      "learning_rate": 9.252931913079521e-08,
+      "loss": 0.3128,
+      "step": 50305
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.431837773424744,
+      "learning_rate": 9.221572479473361e-08,
+      "loss": 0.3459,
+      "step": 50310
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.644901691822145,
+      "learning_rate": 9.19026578217519e-08,
+      "loss": 0.351,
+      "step": 50315
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.842504398392006,
+      "learning_rate": 9.159011824549146e-08,
+      "loss": 0.2872,
+      "step": 50320
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.165281179086053,
+      "learning_rate": 9.127810609953436e-08,
+      "loss": 0.3334,
+      "step": 50325
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.7562824696128265,
+      "learning_rate": 9.096662141740931e-08,
+      "loss": 0.2934,
+      "step": 50330
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.744936752511307,
+      "learning_rate": 9.065566423258621e-08,
+      "loss": 0.3094,
+      "step": 50335
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.186146136092161,
+      "learning_rate": 9.034523457847777e-08,
+      "loss": 0.2915,
+      "step": 50340
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.611058917440578,
+      "learning_rate": 9.003533248844232e-08,
+      "loss": 0.3197,
+      "step": 50345
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.868501460717823,
+      "learning_rate": 8.972595799577933e-08,
+      "loss": 0.2739,
+      "step": 50350
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.467519703279405,
+      "learning_rate": 8.94171111337333e-08,
+      "loss": 0.3662,
+      "step": 50355
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.486339241583467,
+      "learning_rate": 8.91087919354905e-08,
+      "loss": 0.3105,
+      "step": 50360
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.3476448630490845,
+      "learning_rate": 8.88010004341816e-08,
+      "loss": 0.3027,
+      "step": 50365
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.754309533708539,
+      "learning_rate": 8.849373666287964e-08,
+      "loss": 0.295,
+      "step": 50370
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.482721899119557,
+      "learning_rate": 8.818700065460206e-08,
+      "loss": 0.3103,
+      "step": 50375
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.038167578700791,
+      "learning_rate": 8.788079244230863e-08,
+      "loss": 0.2929,
+      "step": 50380
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.824291950693941,
+      "learning_rate": 8.757511205890301e-08,
+      "loss": 0.32,
+      "step": 50385
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.742804541170898,
+      "learning_rate": 8.726995953723172e-08,
+      "loss": 0.3246,
+      "step": 50390
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.057409040654424,
+      "learning_rate": 8.696533491008463e-08,
+      "loss": 0.3376,
+      "step": 50395
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.186778939491333,
+      "learning_rate": 8.666123821019501e-08,
+      "loss": 0.2972,
+      "step": 50400
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.212756169312982,
+      "learning_rate": 8.635766947024004e-08,
+      "loss": 0.3137,
+      "step": 50405
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.09780841698599,
+      "learning_rate": 8.605462872283865e-08,
+      "loss": 0.3228,
+      "step": 50410
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.363497119152206,
+      "learning_rate": 8.575211600055478e-08,
+      "loss": 0.318,
+      "step": 50415
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.611170072335453,
+      "learning_rate": 8.545013133589463e-08,
+      "loss": 0.2828,
+      "step": 50420
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.655326337914719,
+      "learning_rate": 8.514867476130673e-08,
+      "loss": 0.3046,
+      "step": 50425
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.815812570756807,
+      "learning_rate": 8.484774630918457e-08,
+      "loss": 0.2977,
+      "step": 50430
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.265079692222116,
+      "learning_rate": 8.454734601186454e-08,
+      "loss": 0.3096,
+      "step": 50435
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.265372329352981,
+      "learning_rate": 8.42474739016258e-08,
+      "loss": 0.327,
+      "step": 50440
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.675661363396215,
+      "learning_rate": 8.394813001069035e-08,
+      "loss": 0.2908,
+      "step": 50445
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.689339208331142,
+      "learning_rate": 8.364931437122525e-08,
+      "loss": 0.2852,
+      "step": 50450
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.095372852553672,
+      "learning_rate": 8.33510270153387e-08,
+      "loss": 0.2904,
+      "step": 50455
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.838247684494885,
+      "learning_rate": 8.305326797508229e-08,
+      "loss": 0.315,
+      "step": 50460
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.446557729521616,
+      "learning_rate": 8.275603728245319e-08,
+      "loss": 0.3177,
+      "step": 50465
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.594023950904019,
+      "learning_rate": 8.245933496938918e-08,
+      "loss": 0.3099,
+      "step": 50470
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.988865002310838,
+      "learning_rate": 8.216316106777256e-08,
+      "loss": 0.3001,
+      "step": 50475
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.488323348782298,
+      "learning_rate": 8.186751560942785e-08,
+      "loss": 0.2989,
+      "step": 50480
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.099294440369799,
+      "learning_rate": 8.157239862612409e-08,
+      "loss": 0.282,
+      "step": 50485
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.45459536026761,
+      "learning_rate": 8.127781014957259e-08,
+      "loss": 0.2877,
+      "step": 50490
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.452539301615366,
+      "learning_rate": 8.098375021142857e-08,
+      "loss": 0.2955,
+      "step": 50495
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.680812573590136,
+      "learning_rate": 8.06902188432901e-08,
+      "loss": 0.3018,
+      "step": 50500
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.443811170868664,
+      "learning_rate": 8.039721607669859e-08,
+      "loss": 0.322,
+      "step": 50505
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.783655534184915,
+      "learning_rate": 8.010474194313721e-08,
+      "loss": 0.3034,
+      "step": 50510
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.367756587249246,
+      "learning_rate": 7.981279647403528e-08,
+      "loss": 0.3133,
+      "step": 50515
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.537199180987554,
+      "learning_rate": 7.952137970076212e-08,
+      "loss": 0.2931,
+      "step": 50520
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.9418821906300536,
+      "learning_rate": 7.92304916546327e-08,
+      "loss": 0.3302,
+      "step": 50525
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.4318586025931825,
+      "learning_rate": 7.894013236690368e-08,
+      "loss": 0.3137,
+      "step": 50530
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.775949796296001,
+      "learning_rate": 7.86503018687762e-08,
+      "loss": 0.3038,
+      "step": 50535
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.858496311439641,
+      "learning_rate": 7.836100019139314e-08,
+      "loss": 0.356,
+      "step": 50540
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.593863582966466,
+      "learning_rate": 7.80722273658413e-08,
+      "loss": 0.2841,
+      "step": 50545
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.52734634243024,
+      "learning_rate": 7.778398342315085e-08,
+      "loss": 0.3098,
+      "step": 50550
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.32822498058099,
+      "learning_rate": 7.74962683942948e-08,
+      "loss": 0.3201,
+      "step": 50555
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.69765347920918,
+      "learning_rate": 7.720908231018953e-08,
+      "loss": 0.3095,
+      "step": 50560
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.737113373927384,
+      "learning_rate": 7.69224252016948e-08,
+      "loss": 0.2923,
+      "step": 50565
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.888793443075394,
+      "learning_rate": 7.663629709961152e-08,
+      "loss": 0.3189,
+      "step": 50570
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.555395636448465,
+      "learning_rate": 7.635069803468675e-08,
+      "loss": 0.3505,
+      "step": 50575
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.568835597456061,
+      "learning_rate": 7.606562803760987e-08,
+      "loss": 0.304,
+      "step": 50580
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.858743827885877,
+      "learning_rate": 7.578108713901189e-08,
+      "loss": 0.3302,
+      "step": 50585
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.78588864979234,
+      "learning_rate": 7.549707536946837e-08,
+      "loss": 0.3167,
+      "step": 50590
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.6799114837666345,
+      "learning_rate": 7.52135927594977e-08,
+      "loss": 0.3198,
+      "step": 50595
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.362470865277408,
+      "learning_rate": 7.493063933956101e-08,
+      "loss": 0.3333,
+      "step": 50600
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.20192899568831,
+      "learning_rate": 7.464821514006293e-08,
+      "loss": 0.3239,
+      "step": 50605
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.395174539192422,
+      "learning_rate": 7.436632019135193e-08,
+      "loss": 0.2918,
+      "step": 50610
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.5539554577142445,
+      "learning_rate": 7.408495452371766e-08,
+      "loss": 0.3208,
+      "step": 50615
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.737968736331961,
+      "learning_rate": 7.380411816739541e-08,
+      "loss": 0.3223,
+      "step": 50620
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.413379561627084,
+      "learning_rate": 7.352381115256157e-08,
+      "loss": 0.3122,
+      "step": 50625
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.892905667302753,
+      "learning_rate": 7.324403350933595e-08,
+      "loss": 0.3419,
+      "step": 50630
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.62858968304428,
+      "learning_rate": 7.296478526778283e-08,
+      "loss": 0.3023,
+      "step": 50635
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.427630774315769,
+      "learning_rate": 7.268606645790766e-08,
+      "loss": 0.3073,
+      "step": 50640
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.596735394376822,
+      "learning_rate": 7.240787710966146e-08,
+      "loss": 0.3014,
+      "step": 50645
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.994393004728246,
+      "learning_rate": 7.213021725293534e-08,
+      "loss": 0.2905,
+      "step": 50650
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.852710346439697,
+      "learning_rate": 7.185308691756543e-08,
+      "loss": 0.3172,
+      "step": 50655
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.806108788659589,
+      "learning_rate": 7.157648613333124e-08,
+      "loss": 0.3141,
+      "step": 50660
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.6204702214533775,
+      "learning_rate": 7.130041492995454e-08,
+      "loss": 0.3417,
+      "step": 50665
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.475126761571318,
+      "learning_rate": 7.102487333709995e-08,
+      "loss": 0.3054,
+      "step": 50670
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.006984284747317,
+      "learning_rate": 7.0749861384376e-08,
+      "loss": 0.2868,
+      "step": 50675
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.614477656095139,
+      "learning_rate": 7.047537910133406e-08,
+      "loss": 0.3288,
+      "step": 50680
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.424057852884947,
+      "learning_rate": 7.020142651746831e-08,
+      "loss": 0.2923,
+      "step": 50685
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.626664531497191,
+      "learning_rate": 6.992800366221574e-08,
+      "loss": 0.3219,
+      "step": 50690
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.436549046950527,
+      "learning_rate": 6.965511056495789e-08,
+      "loss": 0.2995,
+      "step": 50695
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.327832563040486,
+      "learning_rate": 6.938274725501736e-08,
+      "loss": 0.3295,
+      "step": 50700
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.868305097715036,
+      "learning_rate": 6.911091376166135e-08,
+      "loss": 0.3089,
+      "step": 50705
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.752385747264929,
+      "learning_rate": 6.883961011409923e-08,
+      "loss": 0.3074,
+      "step": 50710
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.443208251519891,
+      "learning_rate": 6.856883634148326e-08,
+      "loss": 0.3151,
+      "step": 50715
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.88315770304472,
+      "learning_rate": 6.829859247291071e-08,
+      "loss": 0.3095,
+      "step": 50720
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.459678871667474,
+      "learning_rate": 6.802887853741891e-08,
+      "loss": 0.2983,
+      "step": 50725
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.499026110495813,
+      "learning_rate": 6.775969456399134e-08,
+      "loss": 0.2603,
+      "step": 50730
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.395285627269967,
+      "learning_rate": 6.749104058155209e-08,
+      "loss": 0.319,
+      "step": 50735
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.158844704155739,
+      "learning_rate": 6.722291661896863e-08,
+      "loss": 0.3155,
+      "step": 50740
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.065675143292178,
+      "learning_rate": 6.695532270505345e-08,
+      "loss": 0.3409,
+      "step": 50745
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.556871803932034,
+      "learning_rate": 6.668825886855967e-08,
+      "loss": 0.2996,
+      "step": 50750
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.228147917553167,
+      "learning_rate": 6.642172513818545e-08,
+      "loss": 0.301,
+      "step": 50755
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.318879529234859,
+      "learning_rate": 6.615572154256956e-08,
+      "loss": 0.3336,
+      "step": 50760
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.846962166350977,
+      "learning_rate": 6.589024811029632e-08,
+      "loss": 0.3293,
+      "step": 50765
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.43983729156032,
+      "learning_rate": 6.562530486989238e-08,
+      "loss": 0.3256,
+      "step": 50770
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.480583732531065,
+      "learning_rate": 6.536089184982552e-08,
+      "loss": 0.2767,
+      "step": 50775
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.536696216981906,
+      "learning_rate": 6.509700907851024e-08,
+      "loss": 0.3306,
+      "step": 50780
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.449547707099758,
+      "learning_rate": 6.483365658429997e-08,
+      "loss": 0.3008,
+      "step": 50785
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.851405238208118,
+      "learning_rate": 6.457083439549372e-08,
+      "loss": 0.3008,
+      "step": 50790
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.771014055575448,
+      "learning_rate": 6.430854254033391e-08,
+      "loss": 0.3318,
+      "step": 50795
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.547439424919254,
+      "learning_rate": 6.404678104700301e-08,
+      "loss": 0.3117,
+      "step": 50800
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.995651136815408,
+      "learning_rate": 6.378554994363018e-08,
+      "loss": 0.314,
+      "step": 50805
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.35402770455956,
+      "learning_rate": 6.352484925828462e-08,
+      "loss": 0.299,
+      "step": 50810
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.387502089185089,
+      "learning_rate": 6.326467901898114e-08,
+      "loss": 0.2782,
+      "step": 50815
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.901541909443361,
+      "learning_rate": 6.300503925367518e-08,
+      "loss": 0.2988,
+      "step": 50820
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.331618446550664,
+      "learning_rate": 6.274592999026552e-08,
+      "loss": 0.2851,
+      "step": 50825
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.4672706474430415,
+      "learning_rate": 6.248735125659655e-08,
+      "loss": 0.3123,
+      "step": 50830
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.282887454706712,
+      "learning_rate": 6.222930308045161e-08,
+      "loss": 0.3238,
+      "step": 50835
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.571514056594294,
+      "learning_rate": 6.197178548956073e-08,
+      "loss": 0.3184,
+      "step": 50840
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.2451860988571575,
+      "learning_rate": 6.171479851159456e-08,
+      "loss": 0.3283,
+      "step": 50845
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.180400392962699,
+      "learning_rate": 6.145834217416769e-08,
+      "loss": 0.3083,
+      "step": 50850
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.105132746372893,
+      "learning_rate": 6.120241650483694e-08,
+      "loss": 0.3176,
+      "step": 50855
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 6.134096959045546,
+      "learning_rate": 6.094702153110365e-08,
+      "loss": 0.3211,
+      "step": 50860
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.592566732063401,
+      "learning_rate": 6.069215728041034e-08,
+      "loss": 0.3463,
+      "step": 50865
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.702514719183494,
+      "learning_rate": 6.04378237801434e-08,
+      "loss": 0.3038,
+      "step": 50870
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.787882802699646,
+      "learning_rate": 6.018402105763265e-08,
+      "loss": 0.3235,
+      "step": 50875
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.672450653568845,
+      "learning_rate": 5.993074914014962e-08,
+      "loss": 0.2991,
+      "step": 50880
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.100656691620931,
+      "learning_rate": 5.967800805490975e-08,
+      "loss": 0.3047,
+      "step": 50885
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.012854771157292,
+      "learning_rate": 5.9425797829071875e-08,
+      "loss": 0.33,
+      "step": 50890
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.461547457372237,
+      "learning_rate": 5.917411848973542e-08,
+      "loss": 0.3391,
+      "step": 50895
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.726259847332957,
+      "learning_rate": 5.8922970063946535e-08,
+      "loss": 0.3128,
+      "step": 50900
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.952165097565031,
+      "learning_rate": 5.8672352578690306e-08,
+      "loss": 0.3241,
+      "step": 50905
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.04859829766221,
+      "learning_rate": 5.842226606089796e-08,
+      "loss": 0.2879,
+      "step": 50910
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.279618153517121,
+      "learning_rate": 5.817271053744189e-08,
+      "loss": 0.3278,
+      "step": 50915
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.714338243746879,
+      "learning_rate": 5.7923686035137874e-08,
+      "loss": 0.3472,
+      "step": 50920
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.0741075367910735,
+      "learning_rate": 5.767519258074505e-08,
+      "loss": 0.2991,
+      "step": 50925
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.503275509095795,
+      "learning_rate": 5.74272302009643e-08,
+      "loss": 0.3175,
+      "step": 50930
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.277906641815037,
+      "learning_rate": 5.717979892244152e-08,
+      "loss": 0.2854,
+      "step": 50935
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.622521615117127,
+      "learning_rate": 5.693289877176378e-08,
+      "loss": 0.286,
+      "step": 50940
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.644644371729939,
+      "learning_rate": 5.668652977546096e-08,
+      "loss": 0.3159,
+      "step": 50945
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.451748444550624,
+      "learning_rate": 5.6440691960007455e-08,
+      "loss": 0.3119,
+      "step": 50950
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.506514099024686,
+      "learning_rate": 5.619538535181879e-08,
+      "loss": 0.3424,
+      "step": 50955
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.58903115694264,
+      "learning_rate": 5.5950609977254986e-08,
+      "loss": 0.3177,
+      "step": 50960
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.2189616600085635,
+      "learning_rate": 5.5706365862617794e-08,
+      "loss": 0.3221,
+      "step": 50965
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.6120194439308335,
+      "learning_rate": 5.5462653034152326e-08,
+      "loss": 0.2973,
+      "step": 50970
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.596783433291919,
+      "learning_rate": 5.521947151804707e-08,
+      "loss": 0.3219,
+      "step": 50975
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.899675789218501,
+      "learning_rate": 5.497682134043225e-08,
+      "loss": 0.3215,
+      "step": 50980
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.9943283942817644,
+      "learning_rate": 5.4734702527382e-08,
+      "loss": 0.2995,
+      "step": 50985
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.9511983233665795,
+      "learning_rate": 5.4493115104913286e-08,
+      "loss": 0.3044,
+      "step": 50990
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.271563576894699,
+      "learning_rate": 5.4252059098985346e-08,
+      "loss": 0.3179,
+      "step": 50995
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.651601483403973,
+      "learning_rate": 5.401153453550134e-08,
+      "loss": 0.3342,
+      "step": 51000
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.765661565149231,
+      "learning_rate": 5.377154144030561e-08,
+      "loss": 0.3276,
+      "step": 51005
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.2832444858579,
+      "learning_rate": 5.353207983918751e-08,
+      "loss": 0.3297,
+      "step": 51010
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.706003775371774,
+      "learning_rate": 5.3293149757877585e-08,
+      "loss": 0.3668,
+      "step": 51015
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.440220912483668,
+      "learning_rate": 5.3054751222050836e-08,
+      "loss": 0.3351,
+      "step": 51020
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.076771025973336,
+      "learning_rate": 5.281688425732345e-08,
+      "loss": 0.2955,
+      "step": 51025
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.538951980008243,
+      "learning_rate": 5.2579548889254985e-08,
+      "loss": 0.3271,
+      "step": 51030
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.41404410727894,
+      "learning_rate": 5.2342745143348364e-08,
+      "loss": 0.3055,
+      "step": 51035
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.5271931424653165,
+      "learning_rate": 5.2106473045049896e-08,
+      "loss": 0.2741,
+      "step": 51040
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.326926162576558,
+      "learning_rate": 5.187073261974762e-08,
+      "loss": 0.2969,
+      "step": 51045
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.427165943700475,
+      "learning_rate": 5.1635523892772935e-08,
+      "loss": 0.2884,
+      "step": 51050
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.8610150988511895,
+      "learning_rate": 5.1400846889400055e-08,
+      "loss": 0.3083,
+      "step": 51055
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.53356862857095,
+      "learning_rate": 5.116670163484605e-08,
+      "loss": 0.315,
+      "step": 51060
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.142706146080753,
+      "learning_rate": 5.093308815427023e-08,
+      "loss": 0.2879,
+      "step": 51065
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.666059296165824,
+      "learning_rate": 5.0700006472776395e-08,
+      "loss": 0.3335,
+      "step": 51070
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.73198153320675,
+      "learning_rate": 5.0467456615410085e-08,
+      "loss": 0.3152,
+      "step": 51075
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.183226817065994,
+      "learning_rate": 5.023543860715907e-08,
+      "loss": 0.3174,
+      "step": 51080
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.299660855333153,
+      "learning_rate": 5.000395247295564e-08,
+      "loss": 0.3039,
+      "step": 51085
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.366329296066329,
+      "learning_rate": 4.977299823767268e-08,
+      "loss": 0.2989,
+      "step": 51090
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.3837064005194355,
+      "learning_rate": 4.954257592612921e-08,
+      "loss": 0.3149,
+      "step": 51095
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.894362997635067,
+      "learning_rate": 4.931268556308322e-08,
+      "loss": 0.3248,
+      "step": 51100
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.174686637715214,
+      "learning_rate": 4.9083327173238274e-08,
+      "loss": 0.3341,
+      "step": 51105
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.73115555649793,
+      "learning_rate": 4.885450078124021e-08,
+      "loss": 0.2888,
+      "step": 51110
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.2739387759155,
+      "learning_rate": 4.862620641167659e-08,
+      "loss": 0.2974,
+      "step": 51115
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.497355303918984,
+      "learning_rate": 4.8398444089079454e-08,
+      "loss": 0.3225,
+      "step": 51120
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 6.301918952129178,
+      "learning_rate": 4.817121383792256e-08,
+      "loss": 0.2912,
+      "step": 51125
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.889121262789907,
+      "learning_rate": 4.7944515682623036e-08,
+      "loss": 0.3586,
+      "step": 51130
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.9330507710958935,
+      "learning_rate": 4.7718349647540296e-08,
+      "loss": 0.2921,
+      "step": 51135
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.044985545361549,
+      "learning_rate": 4.749271575697656e-08,
+      "loss": 0.2951,
+      "step": 51140
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.502781456033189,
+      "learning_rate": 4.7267614035177986e-08,
+      "loss": 0.3031,
+      "step": 51145
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.126933613816304,
+      "learning_rate": 4.704304450633246e-08,
+      "loss": 0.2781,
+      "step": 51150
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.140035963321687,
+      "learning_rate": 4.6819007194570664e-08,
+      "loss": 0.3282,
+      "step": 51155
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.728470333110667,
+      "learning_rate": 4.659550212396668e-08,
+      "loss": 0.3068,
+      "step": 51160
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.469656500938408,
+      "learning_rate": 4.6372529318536864e-08,
+      "loss": 0.2826,
+      "step": 51165
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.610129460771441,
+      "learning_rate": 4.615008880224092e-08,
+      "loss": 0.2893,
+      "step": 51170
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.800904336974467,
+      "learning_rate": 4.5928180598980855e-08,
+      "loss": 0.2943,
+      "step": 51175
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.472736543334713,
+      "learning_rate": 4.570680473260203e-08,
+      "loss": 0.2851,
+      "step": 51180
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.478488408638247,
+      "learning_rate": 4.548596122689153e-08,
+      "loss": 0.3035,
+      "step": 51185
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.712890660944273,
+      "learning_rate": 4.526565010558093e-08,
+      "loss": 0.3037,
+      "step": 51190
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.858576691153519,
+      "learning_rate": 4.504587139234351e-08,
+      "loss": 0.3491,
+      "step": 51195
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.931933719338714,
+      "learning_rate": 4.4826625110794276e-08,
+      "loss": 0.2804,
+      "step": 51200
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.455080269802044,
+      "learning_rate": 4.4607911284493267e-08,
+      "loss": 0.3195,
+      "step": 51205
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.832757994390502,
+      "learning_rate": 4.438972993694168e-08,
+      "loss": 0.3347,
+      "step": 51210
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.385494312504423,
+      "learning_rate": 4.4172081091585215e-08,
+      "loss": 0.2808,
+      "step": 51215
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.961706642024568,
+      "learning_rate": 4.3954964771809606e-08,
+      "loss": 0.312,
+      "step": 51220
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.303491552788849,
+      "learning_rate": 4.373838100094563e-08,
+      "loss": 0.3034,
+      "step": 51225
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.684195748815649,
+      "learning_rate": 4.352232980226689e-08,
+      "loss": 0.2999,
+      "step": 51230
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.311559898568494,
+      "learning_rate": 4.3306811198988164e-08,
+      "loss": 0.3033,
+      "step": 51235
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.405922005429559,
+      "learning_rate": 4.3091825214268135e-08,
+      "loss": 0.2905,
+      "step": 51240
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.652308238280765,
+      "learning_rate": 4.2877371871207774e-08,
+      "loss": 0.3346,
+      "step": 51245
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.63095466104666,
+      "learning_rate": 4.266345119285143e-08,
+      "loss": 0.2865,
+      "step": 51250
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.0170678122206604,
+      "learning_rate": 4.245006320218625e-08,
+      "loss": 0.3189,
+      "step": 51255
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.813475266790062,
+      "learning_rate": 4.223720792214003e-08,
+      "loss": 0.3066,
+      "step": 51260
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.700602080491322,
+      "learning_rate": 4.202488537558669e-08,
+      "loss": 0.2889,
+      "step": 51265
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.156138280176386,
+      "learning_rate": 4.181309558534075e-08,
+      "loss": 0.2992,
+      "step": 51270
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.059320303157331,
+      "learning_rate": 4.160183857415956e-08,
+      "loss": 0.3108,
+      "step": 51275
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.848833840171049,
+      "learning_rate": 4.139111436474441e-08,
+      "loss": 0.3032,
+      "step": 51280
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.945736089338037,
+      "learning_rate": 4.1180922979737194e-08,
+      "loss": 0.2917,
+      "step": 51285
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.4686169764965005,
+      "learning_rate": 4.097126444172539e-08,
+      "loss": 0.2792,
+      "step": 51290
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.608989767600169,
+      "learning_rate": 4.0762138773236534e-08,
+      "loss": 0.3302,
+      "step": 51295
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.151291916298146,
+      "learning_rate": 4.055354599674321e-08,
+      "loss": 0.2901,
+      "step": 51300
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.319251022104954,
+      "learning_rate": 4.034548613465861e-08,
+      "loss": 0.3149,
+      "step": 51305
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.911591303924704,
+      "learning_rate": 4.013795920934038e-08,
+      "loss": 0.3434,
+      "step": 51310
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.348200707148874,
+      "learning_rate": 3.993096524308793e-08,
+      "loss": 0.2969,
+      "step": 51315
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.4209763762854575,
+      "learning_rate": 3.972450425814345e-08,
+      "loss": 0.3114,
+      "step": 51320
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.306674523177662,
+      "learning_rate": 3.951857627669309e-08,
+      "loss": 0.2987,
+      "step": 51325
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.723021871881002,
+      "learning_rate": 3.931318132086359e-08,
+      "loss": 0.3104,
+      "step": 51330
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.567668525573129,
+      "learning_rate": 3.910831941272564e-08,
+      "loss": 0.2921,
+      "step": 51335
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.229991115510662,
+      "learning_rate": 3.890399057429328e-08,
+      "loss": 0.3124,
+      "step": 51340
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.938609083720758,
+      "learning_rate": 3.870019482752174e-08,
+      "loss": 0.3118,
+      "step": 51345
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.000319005002347,
+      "learning_rate": 3.8496932194310723e-08,
+      "loss": 0.3314,
+      "step": 51350
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.225371342694177,
+      "learning_rate": 3.8294202696500525e-08,
+      "loss": 0.309,
+      "step": 51355
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.265058831026209,
+      "learning_rate": 3.809200635587651e-08,
+      "loss": 0.3353,
+      "step": 51360
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.097003599934477,
+      "learning_rate": 3.7890343194164625e-08,
+      "loss": 0.302,
+      "step": 51365
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.3101364174846015,
+      "learning_rate": 3.768921323303476e-08,
+      "loss": 0.3466,
+      "step": 51370
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.4002258948487425,
+      "learning_rate": 3.748861649409907e-08,
+      "loss": 0.3138,
+      "step": 51375
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.635858072390705,
+      "learning_rate": 3.7288552998913095e-08,
+      "loss": 0.3192,
+      "step": 51380
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.601753315150774,
+      "learning_rate": 3.708902276897408e-08,
+      "loss": 0.2795,
+      "step": 51385
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.001443794317098,
+      "learning_rate": 3.6890025825723206e-08,
+      "loss": 0.3158,
+      "step": 51390
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.085064681723496,
+      "learning_rate": 3.669156219054226e-08,
+      "loss": 0.3092,
+      "step": 51395
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.575427050517259,
+      "learning_rate": 3.649363188475752e-08,
+      "loss": 0.2978,
+      "step": 51400
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.644921343058895,
+      "learning_rate": 3.629623492963807e-08,
+      "loss": 0.3113,
+      "step": 51405
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.142196906322257,
+      "learning_rate": 3.6099371346394184e-08,
+      "loss": 0.2854,
+      "step": 51410
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.597506810627093,
+      "learning_rate": 3.59030411561806e-08,
+      "loss": 0.3206,
+      "step": 51415
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.846672191601049,
+      "learning_rate": 3.570724438009376e-08,
+      "loss": 0.3142,
+      "step": 51420
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.158642645229595,
+      "learning_rate": 3.551198103917241e-08,
+      "loss": 0.2925,
+      "step": 51425
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.710377301342723,
+      "learning_rate": 3.5317251154398656e-08,
+      "loss": 0.3104,
+      "step": 51430
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.510829582646343,
+      "learning_rate": 3.512305474669686e-08,
+      "loss": 0.2849,
+      "step": 51435
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.73073715515635,
+      "learning_rate": 3.4929391836934776e-08,
+      "loss": 0.3341,
+      "step": 51440
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.3911957216853,
+      "learning_rate": 3.4736262445921876e-08,
+      "loss": 0.2867,
+      "step": 51445
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.404336851311999,
+      "learning_rate": 3.4543666594410996e-08,
+      "loss": 0.2891,
+      "step": 51450
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.560208321176111,
+      "learning_rate": 3.4351604303097255e-08,
+      "loss": 0.3255,
+      "step": 51455
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.625273373884491,
+      "learning_rate": 3.4160075592619134e-08,
+      "loss": 0.3257,
+      "step": 51460
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.270953574225311,
+      "learning_rate": 3.39690804835563e-08,
+      "loss": 0.3018,
+      "step": 51465
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.729310342874606,
+      "learning_rate": 3.377861899643287e-08,
+      "loss": 0.3353,
+      "step": 51470
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.256171996866667,
+      "learning_rate": 3.358869115171415e-08,
+      "loss": 0.3147,
+      "step": 51475
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.7195182519162735,
+      "learning_rate": 3.339929696980937e-08,
+      "loss": 0.3106,
+      "step": 51480
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.75503189027886,
+      "learning_rate": 3.3210436471069477e-08,
+      "loss": 0.2842,
+      "step": 51485
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.901900190672175,
+      "learning_rate": 3.3022109675787675e-08,
+      "loss": 0.3004,
+      "step": 51490
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.243141756074635,
+      "learning_rate": 3.283431660420111e-08,
+      "loss": 0.3036,
+      "step": 51495
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.267345204612427,
+      "learning_rate": 3.2647057276489755e-08,
+      "loss": 0.3327,
+      "step": 51500
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.208475444683537,
+      "learning_rate": 3.2460331712774186e-08,
+      "loss": 0.3096,
+      "step": 51505
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.9232891298562445,
+      "learning_rate": 3.227413993311945e-08,
+      "loss": 0.2926,
+      "step": 51510
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.590923787583403,
+      "learning_rate": 3.208848195753289e-08,
+      "loss": 0.335,
+      "step": 51515
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.423562550792244,
+      "learning_rate": 3.1903357805963545e-08,
+      "loss": 0.2954,
+      "step": 51520
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.804914673741037,
+      "learning_rate": 3.171876749830494e-08,
+      "loss": 0.3487,
+      "step": 51525
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.087525771742065,
+      "learning_rate": 3.153471105439121e-08,
+      "loss": 0.3366,
+      "step": 51530
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.711356948745704,
+      "learning_rate": 3.135118849399987e-08,
+      "loss": 0.3195,
+      "step": 51535
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.575049828526733,
+      "learning_rate": 3.116819983685182e-08,
+      "loss": 0.2956,
+      "step": 51540
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.173594648449249,
+      "learning_rate": 3.098574510261021e-08,
+      "loss": 0.3174,
+      "step": 51545
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.5551981854017525,
+      "learning_rate": 3.080382431087992e-08,
+      "loss": 0.3095,
+      "step": 51550
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.905985507843454,
+      "learning_rate": 3.0622437481209745e-08,
+      "loss": 0.3246,
+      "step": 51555
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.348849932831417,
+      "learning_rate": 3.044158463308966e-08,
+      "loss": 0.2852,
+      "step": 51560
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.68815883620634,
+      "learning_rate": 3.026126578595412e-08,
+      "loss": 0.3406,
+      "step": 51565
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.206814663854788,
+      "learning_rate": 3.008148095917873e-08,
+      "loss": 0.3073,
+      "step": 51570
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.6143632469739515,
+      "learning_rate": 2.990223017208138e-08,
+      "loss": 0.285,
+      "step": 51575
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.68741877363965,
+      "learning_rate": 2.972351344392499e-08,
+      "loss": 0.324,
+      "step": 51580
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.234203589268513,
+      "learning_rate": 2.9545330793911974e-08,
+      "loss": 0.3339,
+      "step": 51585
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.545525205202262,
+      "learning_rate": 2.93676822411898e-08,
+      "loss": 0.3037,
+      "step": 51590
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.615354537959156,
+      "learning_rate": 2.919056780484708e-08,
+      "loss": 0.3033,
+      "step": 51595
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.030765295019107,
+      "learning_rate": 2.901398750391582e-08,
+      "loss": 0.2997,
+      "step": 51600
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.8715659585737985,
+      "learning_rate": 2.8837941357370282e-08,
+      "loss": 0.3097,
+      "step": 51605
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.201650727094648,
+      "learning_rate": 2.8662429384127e-08,
+      "loss": 0.3409,
+      "step": 51610
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.282882441299974,
+      "learning_rate": 2.8487451603046444e-08,
+      "loss": 0.305,
+      "step": 51615
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.689086307926574,
+      "learning_rate": 2.831300803292969e-08,
+      "loss": 0.3469,
+      "step": 51620
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.862361634197883,
+      "learning_rate": 2.8139098692522292e-08,
+      "loss": 0.3079,
+      "step": 51625
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.8813773822524857,
+      "learning_rate": 2.7965723600510974e-08,
+      "loss": 0.2906,
+      "step": 51630
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.575902476121729,
+      "learning_rate": 2.7792882775525832e-08,
+      "loss": 0.3047,
+      "step": 51635
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.260125105414601,
+      "learning_rate": 2.7620576236139786e-08,
+      "loss": 0.2872,
+      "step": 51640
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.193285116127877,
+      "learning_rate": 2.7448804000867468e-08,
+      "loss": 0.3405,
+      "step": 51645
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.352027310051853,
+      "learning_rate": 2.7277566088166897e-08,
+      "loss": 0.2881,
+      "step": 51650
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.235383909350402,
+      "learning_rate": 2.7106862516437793e-08,
+      "loss": 0.3034,
+      "step": 51655
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.5599531517103316,
+      "learning_rate": 2.6936693304023266e-08,
+      "loss": 0.2789,
+      "step": 51660
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.697232998562075,
+      "learning_rate": 2.676705846920924e-08,
+      "loss": 0.2952,
+      "step": 51665
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.849987052839738,
+      "learning_rate": 2.6597958030223357e-08,
+      "loss": 0.3316,
+      "step": 51670
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.211524333137218,
+      "learning_rate": 2.6429392005236087e-08,
+      "loss": 0.3232,
+      "step": 51675
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.765563415974672,
+      "learning_rate": 2.6261360412360714e-08,
+      "loss": 0.3159,
+      "step": 51680
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.4935754631227915,
+      "learning_rate": 2.6093863269652242e-08,
+      "loss": 0.3423,
+      "step": 51685
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.765543462548494,
+      "learning_rate": 2.5926900595110717e-08,
+      "loss": 0.3273,
+      "step": 51690
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.701805706145789,
+      "learning_rate": 2.5760472406675118e-08,
+      "loss": 0.3113,
+      "step": 51695
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.246181524495169,
+      "learning_rate": 2.559457872223059e-08,
+      "loss": 0.3164,
+      "step": 51700
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.368401265857004,
+      "learning_rate": 2.542921955960176e-08,
+      "loss": 0.2878,
+      "step": 51705
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.916691159377702,
+      "learning_rate": 2.5264394936557746e-08,
+      "loss": 0.3101,
+      "step": 51710
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.364876281440367,
+      "learning_rate": 2.510010487080994e-08,
+      "loss": 0.3126,
+      "step": 51715
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.790348168262585,
+      "learning_rate": 2.4936349380011437e-08,
+      "loss": 0.2963,
+      "step": 51720
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.620303678991867,
+      "learning_rate": 2.477312848175928e-08,
+      "loss": 0.3283,
+      "step": 51725
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.8825015085588825,
+      "learning_rate": 2.4610442193591655e-08,
+      "loss": 0.3052,
+      "step": 51730
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.2828952826924755,
+      "learning_rate": 2.4448290532990137e-08,
+      "loss": 0.3046,
+      "step": 51735
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.332009873377474,
+      "learning_rate": 2.4286673517378566e-08,
+      "loss": 0.346,
+      "step": 51740
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.223505913109157,
+      "learning_rate": 2.4125591164123052e-08,
+      "loss": 0.3069,
+      "step": 51745
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.502265405918352,
+      "learning_rate": 2.3965043490533635e-08,
+      "loss": 0.3479,
+      "step": 51750
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.206542323519605,
+      "learning_rate": 2.3805030513860406e-08,
+      "loss": 0.3287,
+      "step": 51755
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.136697508095126,
+      "learning_rate": 2.3645552251299053e-08,
+      "loss": 0.3265,
+      "step": 51760
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.465478249630526,
+      "learning_rate": 2.348660871998476e-08,
+      "loss": 0.3125,
+      "step": 51765
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.413908591851821,
+      "learning_rate": 2.3328199936997752e-08,
+      "loss": 0.2979,
+      "step": 51770
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.691844200408493,
+      "learning_rate": 2.3170325919359416e-08,
+      "loss": 0.3216,
+      "step": 51775
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.83099009078426,
+      "learning_rate": 2.30129866840334e-08,
+      "loss": 0.3186,
+      "step": 51780
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.525809301467718,
+      "learning_rate": 2.285618224792785e-08,
+      "loss": 0.3263,
+      "step": 51785
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.6011477729389485,
+      "learning_rate": 2.2699912627890952e-08,
+      "loss": 0.313,
+      "step": 51790
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.483392038556261,
+      "learning_rate": 2.2544177840715385e-08,
+      "loss": 0.3074,
+      "step": 51795
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.992611406745169,
+      "learning_rate": 2.2388977903134433e-08,
+      "loss": 0.3218,
+      "step": 51800
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.399920938997252,
+      "learning_rate": 2.2234312831825866e-08,
+      "loss": 0.3174,
+      "step": 51805
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.5807378365562785,
+      "learning_rate": 2.2080182643409163e-08,
+      "loss": 0.326,
+      "step": 51810
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.0474034509006716,
+      "learning_rate": 2.1926587354446084e-08,
+      "loss": 0.2959,
+      "step": 51815
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.931134150175982,
+      "learning_rate": 2.1773526981441196e-08,
+      "loss": 0.3297,
+      "step": 51820
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.214547228961579,
+      "learning_rate": 2.1621001540841346e-08,
+      "loss": 0.3123,
+      "step": 51825
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.372645717361881,
+      "learning_rate": 2.146901104903565e-08,
+      "loss": 0.3323,
+      "step": 51830
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.552758867387564,
+      "learning_rate": 2.131755552235715e-08,
+      "loss": 0.3156,
+      "step": 51835
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.718847379394675,
+      "learning_rate": 2.1166634977080046e-08,
+      "loss": 0.3071,
+      "step": 51840
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.53781423202524,
+      "learning_rate": 2.101624942942082e-08,
+      "loss": 0.3076,
+      "step": 51845
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.819087998607202,
+      "learning_rate": 2.0866398895539875e-08,
+      "loss": 0.3088,
+      "step": 51850
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.143133408222382,
+      "learning_rate": 2.0717083391538772e-08,
+      "loss": 0.3133,
+      "step": 51855
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.6909854400209765,
+      "learning_rate": 2.0568302933462458e-08,
+      "loss": 0.2913,
+      "step": 51860
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.73264316310088,
+      "learning_rate": 2.0420057537297587e-08,
+      "loss": 0.3129,
+      "step": 51865
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.33950616368151,
+      "learning_rate": 2.02723472189742e-08,
+      "loss": 0.3649,
+      "step": 51870
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.3320655853980625,
+      "learning_rate": 2.012517199436459e-08,
+      "loss": 0.2903,
+      "step": 51875
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.419131660144969,
+      "learning_rate": 1.997853187928278e-08,
+      "loss": 0.2841,
+      "step": 51880
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.511392161015789,
+      "learning_rate": 1.9832426889486724e-08,
+      "loss": 0.3031,
+      "step": 51885
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.1565660223673895,
+      "learning_rate": 1.9686857040674966e-08,
+      "loss": 0.2797,
+      "step": 51890
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.173101251395043,
+      "learning_rate": 1.9541822348490557e-08,
+      "loss": 0.2998,
+      "step": 51895
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.485837962699446,
+      "learning_rate": 1.9397322828517695e-08,
+      "loss": 0.2903,
+      "step": 51900
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.787076324755868,
+      "learning_rate": 1.9253358496283402e-08,
+      "loss": 0.3215,
+      "step": 51905
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.871186103633835,
+      "learning_rate": 1.9109929367258085e-08,
+      "loss": 0.3316,
+      "step": 51910
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.991265910841598,
+      "learning_rate": 1.8967035456852744e-08,
+      "loss": 0.3208,
+      "step": 51915
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.979550366337759,
+      "learning_rate": 1.8824676780422323e-08,
+      "loss": 0.3278,
+      "step": 51920
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.730352625739576,
+      "learning_rate": 1.868285335326403e-08,
+      "loss": 0.2971,
+      "step": 51925
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.907837546497949,
+      "learning_rate": 1.854156519061734e-08,
+      "loss": 0.3384,
+      "step": 51930
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.905632553223436,
+      "learning_rate": 1.840081230766455e-08,
+      "loss": 0.3027,
+      "step": 51935
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.252186013716569,
+      "learning_rate": 1.826059471952968e-08,
+      "loss": 0.3252,
+      "step": 51940
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.328358359509609,
+      "learning_rate": 1.8120912441280113e-08,
+      "loss": 0.3294,
+      "step": 51945
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.526319560985917,
+      "learning_rate": 1.7981765487925517e-08,
+      "loss": 0.3154,
+      "step": 51950
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.495100580443618,
+      "learning_rate": 1.7843153874416707e-08,
+      "loss": 0.315,
+      "step": 51955
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.394620423427488,
+      "learning_rate": 1.7705077615649546e-08,
+      "loss": 0.3268,
+      "step": 51960
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.330576473488113,
+      "learning_rate": 1.75675367264605e-08,
+      "loss": 0.3138,
+      "step": 51965
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.411634681174137,
+      "learning_rate": 1.7430531221628856e-08,
+      "loss": 0.2898,
+      "step": 51970
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.6029906575583235,
+      "learning_rate": 1.7294061115875615e-08,
+      "loss": 0.3093,
+      "step": 51975
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 3.941728359034263,
+      "learning_rate": 1.715812642386683e-08,
+      "loss": 0.3075,
+      "step": 51980
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.54498108787844,
+      "learning_rate": 1.7022727160207476e-08,
+      "loss": 0.3423,
+      "step": 51985
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.263460908902632,
+      "learning_rate": 1.688786333944814e-08,
+      "loss": 0.2997,
+      "step": 51990
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.327911088294999,
+      "learning_rate": 1.675353497608001e-08,
+      "loss": 0.2946,
+      "step": 51995
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.103429089545876,
+      "learning_rate": 1.6619742084537094e-08,
+      "loss": 0.3457,
+      "step": 52000
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.716292394950386,
+      "learning_rate": 1.6486484679196222e-08,
+      "loss": 0.314,
+      "step": 52005
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.6793692727390495,
+      "learning_rate": 1.63537627743765e-08,
+      "loss": 0.3264,
+      "step": 52010
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.481553135249161,
+      "learning_rate": 1.6221576384339855e-08,
+      "loss": 0.308,
+      "step": 52015
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.387374326861135,
+      "learning_rate": 1.6089925523289917e-08,
+      "loss": 0.3005,
+      "step": 52020
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.71940319027515,
+      "learning_rate": 1.5958810205372598e-08,
+      "loss": 0.3348,
+      "step": 52025
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.928486382181252,
+      "learning_rate": 1.582823044467774e-08,
+      "loss": 0.3337,
+      "step": 52030
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.455368405457012,
+      "learning_rate": 1.5698186255236338e-08,
+      "loss": 0.3363,
+      "step": 52035
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.576769667525351,
+      "learning_rate": 1.5568677651022213e-08,
+      "loss": 0.2972,
+      "step": 52040
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.150292120113141,
+      "learning_rate": 1.5439704645951458e-08,
+      "loss": 0.293,
+      "step": 52045
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.48692076773266,
+      "learning_rate": 1.5311267253882988e-08,
+      "loss": 0.3439,
+      "step": 52050
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.847504493376699,
+      "learning_rate": 1.5183365488617985e-08,
+      "loss": 0.3417,
+      "step": 52055
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.880121552097477,
+      "learning_rate": 1.5055999363899898e-08,
+      "loss": 0.3102,
+      "step": 52060
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.195617831175037,
+      "learning_rate": 1.4929168893415556e-08,
+      "loss": 0.2767,
+      "step": 52065
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.925578652706051,
+      "learning_rate": 1.4802874090792396e-08,
+      "loss": 0.3218,
+      "step": 52070
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.762875768608234,
+      "learning_rate": 1.467711496960178e-08,
+      "loss": 0.3129,
+      "step": 52075
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.669080721503346,
+      "learning_rate": 1.4551891543356788e-08,
+      "loss": 0.3249,
+      "step": 52080
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.627150354235175,
+      "learning_rate": 1.4427203825513881e-08,
+      "loss": 0.3025,
+      "step": 52085
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.725344870799214,
+      "learning_rate": 1.430305182947067e-08,
+      "loss": 0.311,
+      "step": 52090
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.652074377843964,
+      "learning_rate": 1.4179435568568156e-08,
+      "loss": 0.3137,
+      "step": 52095
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.484482365196683,
+      "learning_rate": 1.40563550560896e-08,
+      "loss": 0.3242,
+      "step": 52100
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.425388229050742,
+      "learning_rate": 1.3933810305260531e-08,
+      "loss": 0.3167,
+      "step": 52105
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.787726268688387,
+      "learning_rate": 1.3811801329248198e-08,
+      "loss": 0.318,
+      "step": 52110
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.592607809503968,
+      "learning_rate": 1.3690328141163777e-08,
+      "loss": 0.2852,
+      "step": 52115
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.853999402913513,
+      "learning_rate": 1.3569390754060163e-08,
+      "loss": 0.3311,
+      "step": 52120
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.188404494802519,
+      "learning_rate": 1.344898918093196e-08,
+      "loss": 0.2897,
+      "step": 52125
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.148936810773478,
+      "learning_rate": 1.332912343471715e-08,
+      "loss": 0.3259,
+      "step": 52130
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.511313299574433,
+      "learning_rate": 1.3209793528295988e-08,
+      "loss": 0.3132,
+      "step": 52135
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.9012202971609575,
+      "learning_rate": 1.3090999474490995e-08,
+      "loss": 0.3054,
+      "step": 52140
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.859070220167851,
+      "learning_rate": 1.2972741286066404e-08,
+      "loss": 0.2892,
+      "step": 52145
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.455208602785613,
+      "learning_rate": 1.2855018975730937e-08,
+      "loss": 0.2877,
+      "step": 52150
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.670421177321195,
+      "learning_rate": 1.2737832556133367e-08,
+      "loss": 0.313,
+      "step": 52155
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 9.17953509234675,
+      "learning_rate": 1.2621182039866398e-08,
+      "loss": 0.3415,
+      "step": 52160
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 6.275350171071507,
+      "learning_rate": 1.2505067439464446e-08,
+      "loss": 0.299,
+      "step": 52165
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 7.868990414992129,
+      "learning_rate": 1.2389488767403646e-08,
+      "loss": 0.2755,
+      "step": 52170
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.611307624850632,
+      "learning_rate": 1.227444603610517e-08,
+      "loss": 0.3087,
+      "step": 52175
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.404459277641047,
+      "learning_rate": 1.2159939257929687e-08,
+      "loss": 0.3241,
+      "step": 52180
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.927689272309279,
+      "learning_rate": 1.2045968445181799e-08,
+      "loss": 0.326,
+      "step": 52185
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.219327943777228,
+      "learning_rate": 1.1932533610107821e-08,
+      "loss": 0.3204,
+      "step": 52190
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.065673762240773,
+      "learning_rate": 1.1819634764897448e-08,
+      "loss": 0.308,
+      "step": 52195
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.773472371014076,
+      "learning_rate": 1.1707271921682084e-08,
+      "loss": 0.2988,
+      "step": 52200
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.840840010062998,
+      "learning_rate": 1.1595445092534852e-08,
+      "loss": 0.3004,
+      "step": 52205
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.817942462967672,
+      "learning_rate": 1.1484154289472805e-08,
+      "loss": 0.3087,
+      "step": 52210
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.820070402045558,
+      "learning_rate": 1.1373399524454154e-08,
+      "loss": 0.3346,
+      "step": 52215
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.512201523229725,
+      "learning_rate": 1.1263180809380492e-08,
+      "loss": 0.3409,
+      "step": 52220
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.606543231603245,
+      "learning_rate": 1.1153498156094567e-08,
+      "loss": 0.3329,
+      "step": 52225
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.819383885136055,
+      "learning_rate": 1.1044351576383061e-08,
+      "loss": 0.3189,
+      "step": 52230
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.7360271075100275,
+      "learning_rate": 1.0935741081974371e-08,
+      "loss": 0.317,
+      "step": 52235
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.647507166704835,
+      "learning_rate": 1.0827666684538051e-08,
+      "loss": 0.3307,
+      "step": 52240
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.793969473854846,
+      "learning_rate": 1.0720128395688145e-08,
+      "loss": 0.3193,
+      "step": 52245
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.185595282548315,
+      "learning_rate": 1.0613126226979853e-08,
+      "loss": 0.3176,
+      "step": 52250
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.888676946850519,
+      "learning_rate": 1.0506660189910645e-08,
+      "loss": 0.3143,
+      "step": 52255
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.972644933108963,
+      "learning_rate": 1.040073029592137e-08,
+      "loss": 0.3138,
+      "step": 52260
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.874024939395055,
+      "learning_rate": 1.0295336556394031e-08,
+      "loss": 0.2942,
+      "step": 52265
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.098285409668428,
+      "learning_rate": 1.0190478982654573e-08,
+      "loss": 0.2995,
+      "step": 52270
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.871984302664849,
+      "learning_rate": 1.0086157585970091e-08,
+      "loss": 0.2785,
+      "step": 52275
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.91481131961047,
+      "learning_rate": 9.9823723775494e-09,
+      "loss": 0.3152,
+      "step": 52280
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.44373163447765,
+      "learning_rate": 9.879123368545795e-09,
+      "loss": 0.2839,
+      "step": 52285
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.412368546087253,
+      "learning_rate": 9.776410570053186e-09,
+      "loss": 0.3269,
+      "step": 52290
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.655100645661409,
+      "learning_rate": 9.674233993109405e-09,
+      "loss": 0.3223,
+      "step": 52295
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.410426394663395,
+      "learning_rate": 9.572593648692341e-09,
+      "loss": 0.3026,
+      "step": 52300
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.617620243294774,
+      "learning_rate": 9.471489547724921e-09,
+      "loss": 0.3076,
+      "step": 52305
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.415697482325816,
+      "learning_rate": 9.370921701071233e-09,
+      "loss": 0.2923,
+      "step": 52310
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.7329934127292725,
+      "learning_rate": 9.27089011953708e-09,
+      "loss": 0.3323,
+      "step": 52315
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.441815325896142,
+      "learning_rate": 9.171394813871081e-09,
+      "loss": 0.2931,
+      "step": 52320
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.31469371107524,
+      "learning_rate": 9.072435794765244e-09,
+      "loss": 0.3393,
+      "step": 52325
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.425477438367652,
+      "learning_rate": 8.974013072852727e-09,
+      "loss": 0.3225,
+      "step": 52330
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.37568432517497,
+      "learning_rate": 8.87612665871007e-09,
+      "loss": 0.3176,
+      "step": 52335
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.736481356056593,
+      "learning_rate": 8.77877656285442e-09,
+      "loss": 0.3103,
+      "step": 52340
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.666221174640978,
+      "learning_rate": 8.681962795747955e-09,
+      "loss": 0.337,
+      "step": 52345
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.437653898208052,
+      "learning_rate": 8.585685367792362e-09,
+      "loss": 0.3232,
+      "step": 52350
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.629032618671705,
+      "learning_rate": 8.489944289333806e-09,
+      "loss": 0.2796,
+      "step": 52355
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.488751992610083,
+      "learning_rate": 8.394739570660171e-09,
+      "loss": 0.2996,
+      "step": 52360
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.709464236450522,
+      "learning_rate": 8.30007122200105e-09,
+      "loss": 0.3111,
+      "step": 52365
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.706032455750089,
+      "learning_rate": 8.205939253529416e-09,
+      "loss": 0.3112,
+      "step": 52370
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.891845461885052,
+      "learning_rate": 8.112343675359958e-09,
+      "loss": 0.2835,
+      "step": 52375
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.664060810834802,
+      "learning_rate": 8.019284497550184e-09,
+      "loss": 0.3077,
+      "step": 52380
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.362832373548632,
+      "learning_rate": 7.926761730099874e-09,
+      "loss": 0.3341,
+      "step": 52385
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.453116542371152,
+      "learning_rate": 7.834775382950522e-09,
+      "loss": 0.3521,
+      "step": 52390
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.435751272009442,
+      "learning_rate": 7.74332546598644e-09,
+      "loss": 0.2822,
+      "step": 52395
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.41740543858173,
+      "learning_rate": 7.65241198903477e-09,
+      "loss": 0.3358,
+      "step": 52400
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.461778169052866,
+      "learning_rate": 7.562034961864361e-09,
+      "loss": 0.3535,
+      "step": 52405
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.081227209607926,
+      "learning_rate": 7.472194394186338e-09,
+      "loss": 0.3222,
+      "step": 52410
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.629891181257991,
+      "learning_rate": 7.382890295654644e-09,
+      "loss": 0.3054,
+      "step": 52415
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.315021931142176,
+      "learning_rate": 7.2941226758660445e-09,
+      "loss": 0.3311,
+      "step": 52420
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.471789889424597,
+      "learning_rate": 7.205891544357912e-09,
+      "loss": 0.2736,
+      "step": 52425
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.519041621841997,
+      "learning_rate": 7.118196910612107e-09,
+      "loss": 0.2956,
+      "step": 52430
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.959552556071339,
+      "learning_rate": 7.031038784051092e-09,
+      "loss": 0.3395,
+      "step": 52435
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.820161577926311,
+      "learning_rate": 6.944417174040707e-09,
+      "loss": 0.3033,
+      "step": 52440
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.646316528705217,
+      "learning_rate": 6.858332089888509e-09,
+      "loss": 0.3119,
+      "step": 52445
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.620750086621261,
+      "learning_rate": 6.7727835408448735e-09,
+      "loss": 0.3123,
+      "step": 52450
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 6.262751220339147,
+      "learning_rate": 6.687771536103005e-09,
+      "loss": 0.3119,
+      "step": 52455
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.820043019297438,
+      "learning_rate": 6.603296084796706e-09,
+      "loss": 0.3095,
+      "step": 52460
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.497552218335965,
+      "learning_rate": 6.519357196003717e-09,
+      "loss": 0.3054,
+      "step": 52465
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.920993053052893,
+      "learning_rate": 6.435954878744044e-09,
+      "loss": 0.3158,
+      "step": 52470
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.3961457430899165,
+      "learning_rate": 6.353089141979407e-09,
+      "loss": 0.2819,
+      "step": 52475
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.767300182498846,
+      "learning_rate": 6.270759994613795e-09,
+      "loss": 0.3227,
+      "step": 52480
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.384253467656807,
+      "learning_rate": 6.188967445494021e-09,
+      "loss": 0.3101,
+      "step": 52485
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.450868333634514,
+      "learning_rate": 6.1077115034091635e-09,
+      "loss": 0.291,
+      "step": 52490
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.394315454485715,
+      "learning_rate": 6.026992177090019e-09,
+      "loss": 0.3246,
+      "step": 52495
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.2815009120238345,
+      "learning_rate": 5.946809475210758e-09,
+      "loss": 0.3052,
+      "step": 52500
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.843652894375007,
+      "learning_rate": 5.867163406387822e-09,
+      "loss": 0.303,
+      "step": 52505
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.296658960721441,
+      "learning_rate": 5.788053979178254e-09,
+      "loss": 0.3044,
+      "step": 52510
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.548167929631657,
+      "learning_rate": 5.709481202084144e-09,
+      "loss": 0.3288,
+      "step": 52515
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.280104652256016,
+      "learning_rate": 5.631445083547071e-09,
+      "loss": 0.3156,
+      "step": 52520
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.3821783390502755,
+      "learning_rate": 5.553945631953661e-09,
+      "loss": 0.3212,
+      "step": 52525
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.791121977622194,
+      "learning_rate": 5.476982855630586e-09,
+      "loss": 0.3391,
+      "step": 52530
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.748360894943294,
+      "learning_rate": 5.4005567628478975e-09,
+      "loss": 0.3202,
+      "step": 52535
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.226065304793106,
+      "learning_rate": 5.324667361819025e-09,
+      "loss": 0.345,
+      "step": 52540
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.316860487123996,
+      "learning_rate": 5.249314660697447e-09,
+      "loss": 0.3011,
+      "step": 52545
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.358530785274255,
+      "learning_rate": 5.174498667580574e-09,
+      "loss": 0.3194,
+      "step": 52550
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.555601613781942,
+      "learning_rate": 5.100219390507532e-09,
+      "loss": 0.3196,
+      "step": 52555
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.747688110343156,
+      "learning_rate": 5.02647683746027e-09,
+      "loss": 0.3059,
+      "step": 52560
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.743080291054537,
+      "learning_rate": 4.953271016362449e-09,
+      "loss": 0.3089,
+      "step": 52565
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.2473504407222915,
+      "learning_rate": 4.880601935079998e-09,
+      "loss": 0.3375,
+      "step": 52570
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.594031998831323,
+      "learning_rate": 4.808469601422783e-09,
+      "loss": 0.3303,
+      "step": 52575
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.616023760796691,
+      "learning_rate": 4.736874023140714e-09,
+      "loss": 0.2979,
+      "step": 52580
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.687626815992607,
+      "learning_rate": 4.665815207927083e-09,
+      "loss": 0.3303,
+      "step": 52585
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.847525294640633,
+      "learning_rate": 4.595293163418002e-09,
+      "loss": 0.303,
+      "step": 52590
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.721476917913135,
+      "learning_rate": 4.5253078971913e-09,
+      "loss": 0.3227,
+      "step": 52595
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.0111436246551175,
+      "learning_rate": 4.455859416766517e-09,
+      "loss": 0.32,
+      "step": 52600
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.345335916251948,
+      "learning_rate": 4.386947729606572e-09,
+      "loss": 0.3302,
+      "step": 52605
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.778534890144306,
+      "learning_rate": 4.318572843116653e-09,
+      "loss": 0.3277,
+      "step": 52610
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.9042433707296125,
+      "learning_rate": 4.25073476464366e-09,
+      "loss": 0.3049,
+      "step": 52615
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.230490503426172,
+      "learning_rate": 4.183433501477318e-09,
+      "loss": 0.3013,
+      "step": 52620
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.704792837084918,
+      "learning_rate": 4.116669060849065e-09,
+      "loss": 0.3188,
+      "step": 52625
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.393275824440584,
+      "learning_rate": 4.050441449933162e-09,
+      "loss": 0.3141,
+      "step": 52630
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.4533973250389804,
+      "learning_rate": 3.984750675846139e-09,
+      "loss": 0.2932,
+      "step": 52635
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.40402148357871,
+      "learning_rate": 3.919596745646792e-09,
+      "loss": 0.3148,
+      "step": 52640
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.35932916402804,
+      "learning_rate": 3.854979666336745e-09,
+      "loss": 0.3068,
+      "step": 52645
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.059937107282575,
+      "learning_rate": 3.7908994448582205e-09,
+      "loss": 0.2972,
+      "step": 52650
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.466068410612167,
+      "learning_rate": 3.727356088097378e-09,
+      "loss": 0.3167,
+      "step": 52655
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.0576839610472435,
+      "learning_rate": 3.664349602882644e-09,
+      "loss": 0.3411,
+      "step": 52660
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.451213919774457,
+      "learning_rate": 3.601879995984159e-09,
+      "loss": 0.3203,
+      "step": 52665
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.314255888607907,
+      "learning_rate": 3.53994727411433e-09,
+      "loss": 0.3063,
+      "step": 52670
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.765208803808081,
+      "learning_rate": 3.4785514439278357e-09,
+      "loss": 0.2994,
+      "step": 52675
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.636140889536764,
+      "learning_rate": 3.41769251202273e-09,
+      "loss": 0.3257,
+      "step": 52680
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.4474747672151915,
+      "learning_rate": 3.3573704849382273e-09,
+      "loss": 0.3077,
+      "step": 52685
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.090214735496071,
+      "learning_rate": 3.297585369155809e-09,
+      "loss": 0.2882,
+      "step": 52690
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.973183462293142,
+      "learning_rate": 3.2383371711003366e-09,
+      "loss": 0.3262,
+      "step": 52695
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.475461560725476,
+      "learning_rate": 3.1796258971372727e-09,
+      "loss": 0.292,
+      "step": 52700
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.6387364348574245,
+      "learning_rate": 3.121451553576571e-09,
+      "loss": 0.3205,
+      "step": 52705
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 7.208933111887954,
+      "learning_rate": 3.0638141466687866e-09,
+      "loss": 0.3173,
+      "step": 52710
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.362484976363597,
+      "learning_rate": 3.0067136826072986e-09,
+      "loss": 0.3412,
+      "step": 52715
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.122931067733473,
+      "learning_rate": 2.9501501675277544e-09,
+      "loss": 0.3481,
+      "step": 52720
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.30235071952696,
+      "learning_rate": 2.8941236075086256e-09,
+      "loss": 0.3356,
+      "step": 52725
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.292343359567982,
+      "learning_rate": 2.838634008569541e-09,
+      "loss": 0.3063,
+      "step": 52730
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.283475516396307,
+      "learning_rate": 2.783681376673508e-09,
+      "loss": 0.3132,
+      "step": 52735
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.893696263082588,
+      "learning_rate": 2.729265717724694e-09,
+      "loss": 0.3132,
+      "step": 52740
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.025191685581385,
+      "learning_rate": 2.675387037571753e-09,
+      "loss": 0.3127,
+      "step": 52745
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.810919273149971,
+      "learning_rate": 2.6220453420028325e-09,
+      "loss": 0.305,
+      "step": 52750
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.817797980037922,
+      "learning_rate": 2.5692406367505695e-09,
+      "loss": 0.318,
+      "step": 52755
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.628643539625244,
+      "learning_rate": 2.516972927488759e-09,
+      "loss": 0.319,
+      "step": 52760
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.781153106037272,
+      "learning_rate": 2.4652422198334638e-09,
+      "loss": 0.3039,
+      "step": 52765
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.857922136843907,
+      "learning_rate": 2.41404851934357e-09,
+      "loss": 0.3052,
+      "step": 52770
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.391891467239641,
+      "learning_rate": 2.3633918315207895e-09,
+      "loss": 0.3211,
+      "step": 52775
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.229468119515766,
+      "learning_rate": 2.313272161807434e-09,
+      "loss": 0.3046,
+      "step": 52780
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.756058296162218,
+      "learning_rate": 2.263689515589196e-09,
+      "loss": 0.3343,
+      "step": 52785
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.644018764213784,
+      "learning_rate": 2.2146438981945908e-09,
+      "loss": 0.3093,
+      "step": 52790
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.77380290534556,
+      "learning_rate": 2.1661353148932917e-09,
+      "loss": 0.3123,
+      "step": 52795
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.019618943865824,
+      "learning_rate": 2.1181637708977963e-09,
+      "loss": 0.3366,
+      "step": 52800
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.308653992855518,
+      "learning_rate": 2.0707292713634255e-09,
+      "loss": 0.3253,
+      "step": 52805
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.49799096524608,
+      "learning_rate": 2.023831821386102e-09,
+      "loss": 0.3174,
+      "step": 52810
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.347721066338038,
+      "learning_rate": 1.977471426006239e-09,
+      "loss": 0.2957,
+      "step": 52815
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.643183311459069,
+      "learning_rate": 1.9316480902048517e-09,
+      "loss": 0.3246,
+      "step": 52820
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.483571405517584,
+      "learning_rate": 1.8863618189057798e-09,
+      "loss": 0.3045,
+      "step": 52825
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.039070869095936,
+      "learning_rate": 1.8416126169762405e-09,
+      "loss": 0.3266,
+      "step": 52830
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.615760782256252,
+      "learning_rate": 1.7974004892234998e-09,
+      "loss": 0.3102,
+      "step": 52835
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.414752377335334,
+      "learning_rate": 1.753725440398757e-09,
+      "loss": 0.288,
+      "step": 52840
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.844827737767919,
+      "learning_rate": 1.7105874751949247e-09,
+      "loss": 0.2952,
+      "step": 52845
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.5092265505590605,
+      "learning_rate": 1.6679865982482946e-09,
+      "loss": 0.3163,
+      "step": 52850
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.936603015264989,
+      "learning_rate": 1.6259228141352058e-09,
+      "loss": 0.2973,
+      "step": 52855
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 6.169628484505024,
+      "learning_rate": 1.5843961273764862e-09,
+      "loss": 0.3347,
+      "step": 52860
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.294682213649803,
+      "learning_rate": 1.5434065424341227e-09,
+      "loss": 0.3081,
+      "step": 52865
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.465428353067447,
+      "learning_rate": 1.5029540637123696e-09,
+      "loss": 0.2922,
+      "step": 52870
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.152948023697802,
+      "learning_rate": 1.4630386955583054e-09,
+      "loss": 0.3113,
+      "step": 52875
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.959260029969127,
+      "learning_rate": 1.4236604422612766e-09,
+      "loss": 0.3282,
+      "step": 52880
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.470867675936348,
+      "learning_rate": 1.384819308051788e-09,
+      "loss": 0.3172,
+      "step": 52885
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.351809370846714,
+      "learning_rate": 1.3465152971048335e-09,
+      "loss": 0.2982,
+      "step": 52890
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.45804987173079,
+      "learning_rate": 1.3087484135348993e-09,
+      "loss": 0.2962,
+      "step": 52895
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.646750164507272,
+      "learning_rate": 1.271518661400961e-09,
+      "loss": 0.3302,
+      "step": 52900
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.431301545097962,
+      "learning_rate": 1.2348260447037074e-09,
+      "loss": 0.3226,
+      "step": 52905
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.682177549735444,
+      "learning_rate": 1.1986705673849851e-09,
+      "loss": 0.3384,
+      "step": 52910
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.311406088703032,
+      "learning_rate": 1.16305223333113e-09,
+      "loss": 0.2983,
+      "step": 52915
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.276341251302369,
+      "learning_rate": 1.1279710463685256e-09,
+      "loss": 0.2895,
+      "step": 52920
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.982763729718368,
+      "learning_rate": 1.0934270102674893e-09,
+      "loss": 0.3333,
+      "step": 52925
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.569834812857314,
+      "learning_rate": 1.0594201287394967e-09,
+      "loss": 0.317,
+      "step": 52930
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.413040368232051,
+      "learning_rate": 1.0259504054382918e-09,
+      "loss": 0.3102,
+      "step": 52935
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.787770120033109,
+      "learning_rate": 9.930178439615523e-10,
+      "loss": 0.3153,
+      "step": 52940
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.240075960406916,
+      "learning_rate": 9.606224478470039e-10,
+      "loss": 0.3179,
+      "step": 52945
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.030678396334085,
+      "learning_rate": 9.287642205763059e-10,
+      "loss": 0.3061,
+      "step": 52950
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 6.982543408386207,
+      "learning_rate": 8.97443165572276e-10,
+      "loss": 0.3448,
+      "step": 52955
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.616099239800316,
+      "learning_rate": 8.666592862005552e-10,
+      "loss": 0.3785,
+      "step": 52960
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.608389749842388,
+      "learning_rate": 8.36412585769053e-10,
+      "loss": 0.3058,
+      "step": 52965
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.373584522790846,
+      "learning_rate": 8.067030675285026e-10,
+      "loss": 0.2898,
+      "step": 52970
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.272426551394382,
+      "learning_rate": 7.775307346702399e-10,
+      "loss": 0.3309,
+      "step": 52975
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.963509366439007,
+      "learning_rate": 7.488955903300898e-10,
+      "loss": 0.3251,
+      "step": 52980
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.576542616274994,
+      "learning_rate": 7.207976375839254e-10,
+      "loss": 0.3009,
+      "step": 52985
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.571072696034182,
+      "learning_rate": 6.932368794521083e-10,
+      "loss": 0.2837,
+      "step": 52990
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.779269423741139,
+      "learning_rate": 6.662133188956033e-10,
+      "loss": 0.3154,
+      "step": 52995
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.412823950048739,
+      "learning_rate": 6.397269588176436e-10,
+      "loss": 0.3195,
+      "step": 53000
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.070608395115606,
+      "learning_rate": 6.137778020653962e-10,
+      "loss": 0.3061,
+      "step": 53005
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.2741498365653285,
+      "learning_rate": 5.883658514266311e-10,
+      "loss": 0.3067,
+      "step": 53010
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 6.115827711252612,
+      "learning_rate": 5.634911096319417e-10,
+      "loss": 0.3259,
+      "step": 53015
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.9834486818431705,
+      "learning_rate": 5.391535793541902e-10,
+      "loss": 0.2915,
+      "step": 53020
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.744970874173944,
+      "learning_rate": 5.153532632090619e-10,
+      "loss": 0.3187,
+      "step": 53025
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.131062627866473,
+      "learning_rate": 4.920901637534004e-10,
+      "loss": 0.3105,
+      "step": 53030
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.799085291103587,
+      "learning_rate": 4.693642834868728e-10,
+      "loss": 0.3261,
+      "step": 53035
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.027244128098909,
+      "learning_rate": 4.471756248519699e-10,
+      "loss": 0.3194,
+      "step": 53040
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.420099986937671,
+      "learning_rate": 4.255241902328955e-10,
+      "loss": 0.3221,
+      "step": 53045
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.0019439651383815,
+      "learning_rate": 4.04409981956122e-10,
+      "loss": 0.3048,
+      "step": 53050
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.583616091572451,
+      "learning_rate": 3.838330022903902e-10,
+      "loss": 0.3175,
+      "step": 53055
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.124705438210518,
+      "learning_rate": 3.6379325344670925e-10,
+      "loss": 0.2786,
+      "step": 53060
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 6.127207862512285,
+      "learning_rate": 3.4429073757835663e-10,
+      "loss": 0.3145,
+      "step": 53065
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.165205522855713,
+      "learning_rate": 3.2532545678143346e-10,
+      "loss": 0.3198,
+      "step": 53070
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.003536846773055,
+      "learning_rate": 3.06897413093199e-10,
+      "loss": 0.293,
+      "step": 53075
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.247727744195595,
+      "learning_rate": 2.8900660849429105e-10,
+      "loss": 0.3262,
+      "step": 53080
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.368846114283611,
+      "learning_rate": 2.716530449070609e-10,
+      "loss": 0.3289,
+      "step": 53085
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.713329930175754,
+      "learning_rate": 2.5483672419612806e-10,
+      "loss": 0.3188,
+      "step": 53090
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.432704527151101,
+      "learning_rate": 2.3855764816893555e-10,
+      "loss": 0.302,
+      "step": 53095
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.291525387320038,
+      "learning_rate": 2.2281581857408474e-10,
+      "loss": 0.3071,
+      "step": 53100
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.65358104887318,
+      "learning_rate": 2.0761123710300035e-10,
+      "loss": 0.3453,
+      "step": 53105
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.556930459035139,
+      "learning_rate": 1.929439053904858e-10,
+      "loss": 0.3536,
+      "step": 53110
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.5115042302969774,
+      "learning_rate": 1.7881382501139243e-10,
+      "loss": 0.3022,
+      "step": 53115
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.52677531248776,
+      "learning_rate": 1.652209974850605e-10,
+      "loss": 0.313,
+      "step": 53120
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.297324251384825,
+      "learning_rate": 1.5216542427143322e-10,
+      "loss": 0.3216,
+      "step": 53125
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.749725198149976,
+      "learning_rate": 1.3964710677383253e-10,
+      "loss": 0.3707,
+      "step": 53130
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.501658968668157,
+      "learning_rate": 1.2766604633729364e-10,
+      "loss": 0.3065,
+      "step": 53135
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.143536651636263,
+      "learning_rate": 1.162222442491201e-10,
+      "loss": 0.3079,
+      "step": 53140
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.347122517183747,
+      "learning_rate": 1.0531570173888395e-10,
+      "loss": 0.3091,
+      "step": 53145
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.578423641347769,
+      "learning_rate": 9.494641997898069e-11,
+      "loss": 0.3018,
+      "step": 53150
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 6.329463561818456,
+      "learning_rate": 8.511440008296402e-11,
+      "loss": 0.2982,
+      "step": 53155
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.7254765739441025,
+      "learning_rate": 7.581964310776624e-11,
+      "loss": 0.3123,
+      "step": 53160
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.745683841446329,
+      "learning_rate": 6.706215005203298e-11,
+      "loss": 0.3174,
+      "step": 53165
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.64765029074995,
+      "learning_rate": 5.884192185723337e-11,
+      "loss": 0.3113,
+      "step": 53170
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.49916323772985,
+      "learning_rate": 5.115895940599469e-11,
+      "loss": 0.3388,
+      "step": 53175
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.726996100697955,
+      "learning_rate": 4.401326352432289e-11,
+      "loss": 0.3108,
+      "step": 53180
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.976107560106654,
+      "learning_rate": 3.740483497993719e-11,
+      "loss": 0.3155,
+      "step": 53185
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.470688031701695,
+      "learning_rate": 3.1333674482825205e-11,
+      "loss": 0.3463,
+      "step": 53190
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.335144723373069,
+      "learning_rate": 2.579978268524297e-11,
+      "loss": 0.3014,
+      "step": 53195
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.682211886983013,
+      "learning_rate": 2.080316018227002e-11,
+      "loss": 0.3189,
+      "step": 53200
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.461975872545511,
+      "learning_rate": 1.6343807510699193e-11,
+      "loss": 0.3213,
+      "step": 53205
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.09502696077276,
+      "learning_rate": 1.2421725149591723e-11,
+      "loss": 0.2776,
+      "step": 53210
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.467293134379502,
+      "learning_rate": 9.036913520277246e-12,
+      "loss": 0.2935,
+      "step": 53215
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.5623102076298645,
+      "learning_rate": 6.1893729869089144e-12,
+      "loss": 0.3194,
+      "step": 53220
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.798586620085093,
+      "learning_rate": 3.879103854798061e-12,
+      "loss": 0.3053,
+      "step": 53225
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.714632863457232,
+      "learning_rate": 2.106106372634642e-12,
+      "loss": 0.313,
+      "step": 53230
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.561214232712145,
+      "learning_rate": 8.70380730821907e-13,
+      "loss": 0.332,
+      "step": 53235
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.398089998615461,
+      "learning_rate": 1.7192706203150722e-13,
+      "loss": 0.2983,
+      "step": 53240
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.1549195051193237,
+      "eval_runtime": 25.3927,
+      "eval_samples_per_second": 31.741,
+      "eval_steps_per_second": 3.978,
+      "step": 53244
+    },
+    {
+      "epoch": 3.0,
+      "step": 53244,
+      "total_flos": 1833005840596992.0,
+      "train_loss": 0.832810894061188,
+      "train_runtime": 128353.9377,
+      "train_samples_per_second": 6.637,
+      "train_steps_per_second": 0.415
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 53244,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 1833005840596992.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}