diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,41711 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9998487674547563,
+  "eval_steps": 500,
+  "global_step": 29754,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1782.354232835545,
+      "learning_rate": 3.3602150537634413e-09,
+      "loss": 17.1486,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1724.0339589676632,
+      "learning_rate": 1.6801075268817205e-08,
+      "loss": 16.7835,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1700.514961570837,
+      "learning_rate": 3.360215053763441e-08,
+      "loss": 16.6795,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1570.7438431775615,
+      "learning_rate": 5.040322580645161e-08,
+      "loss": 16.2985,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1602.1277018435649,
+      "learning_rate": 6.720430107526882e-08,
+      "loss": 16.4868,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1499.0619348435034,
+      "learning_rate": 8.400537634408603e-08,
+      "loss": 15.7543,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1438.121413693714,
+      "learning_rate": 1.0080645161290322e-07,
+      "loss": 14.2784,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1205.847939284157,
+      "learning_rate": 1.1760752688172043e-07,
+      "loss": 13.4497,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1081.138946537131,
+      "learning_rate": 1.3440860215053764e-07,
+      "loss": 11.4945,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1122.9535780227677,
+      "learning_rate": 1.5120967741935485e-07,
+      "loss": 10.4741,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 915.8074736571381,
+      "learning_rate": 1.6801075268817206e-07,
+      "loss": 9.3298,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 485.2026272196928,
+      "learning_rate": 1.8481182795698927e-07,
+      "loss": 7.5624,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 413.4948329618694,
+      "learning_rate": 2.0161290322580645e-07,
+      "loss": 6.5282,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 341.22189353250485,
+      "learning_rate": 2.1841397849462368e-07,
+      "loss": 5.8548,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 314.30197693693736,
+      "learning_rate": 2.3521505376344087e-07,
+      "loss": 5.4052,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 219.09034450096308,
+      "learning_rate": 2.520161290322581e-07,
+      "loss": 4.9707,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 243.8000015899658,
+      "learning_rate": 2.688172043010753e-07,
+      "loss": 4.7539,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 158.48368358483174,
+      "learning_rate": 2.856182795698925e-07,
+      "loss": 4.4491,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 177.99221919611344,
+      "learning_rate": 3.024193548387097e-07,
+      "loss": 4.0891,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 147.33085776107916,
+      "learning_rate": 3.192204301075269e-07,
+      "loss": 3.8303,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 118.84914037685353,
+      "learning_rate": 3.360215053763441e-07,
+      "loss": 3.5773,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 115.65208604223525,
+      "learning_rate": 3.528225806451614e-07,
+      "loss": 3.2914,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 117.31300158152648,
+      "learning_rate": 3.6962365591397853e-07,
+      "loss": 2.9765,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 105.60650363440705,
+      "learning_rate": 3.8642473118279574e-07,
+      "loss": 3.0645,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 97.2173603475482,
+      "learning_rate": 4.032258064516129e-07,
+      "loss": 2.6624,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 77.426266243747,
+      "learning_rate": 4.200268817204301e-07,
+      "loss": 2.689,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 76.25148368706107,
+      "learning_rate": 4.3682795698924737e-07,
+      "loss": 2.581,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 60.52769003302683,
+      "learning_rate": 4.536290322580646e-07,
+      "loss": 2.49,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 50.57180850107596,
+      "learning_rate": 4.7043010752688173e-07,
+      "loss": 2.4519,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 62.69948504309535,
+      "learning_rate": 4.872311827956989e-07,
+      "loss": 2.3989,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 47.631212194728114,
+      "learning_rate": 5.040322580645161e-07,
+      "loss": 2.2853,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 42.17283021141972,
+      "learning_rate": 5.208333333333334e-07,
+      "loss": 2.3543,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.08359958874836,
+      "learning_rate": 5.376344086021506e-07,
+      "loss": 2.1276,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 40.674430325440596,
+      "learning_rate": 5.544354838709678e-07,
+      "loss": 2.1955,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 31.948809198091773,
+      "learning_rate": 5.71236559139785e-07,
+      "loss": 2.0281,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 40.27205465031299,
+      "learning_rate": 5.880376344086022e-07,
+      "loss": 1.9945,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 30.44956418233564,
+      "learning_rate": 6.048387096774194e-07,
+      "loss": 2.0563,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 25.864985022985085,
+      "learning_rate": 6.216397849462366e-07,
+      "loss": 2.0562,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 48.57307017843553,
+      "learning_rate": 6.384408602150538e-07,
+      "loss": 2.0814,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.41894528967244,
+      "learning_rate": 6.55241935483871e-07,
+      "loss": 2.0217,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 25.687554758258678,
+      "learning_rate": 6.720430107526882e-07,
+      "loss": 1.9102,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.2588126225637,
+      "learning_rate": 6.888440860215053e-07,
+      "loss": 1.8607,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 26.021175684312663,
+      "learning_rate": 7.056451612903228e-07,
+      "loss": 1.8601,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 24.55833963633963,
+      "learning_rate": 7.224462365591399e-07,
+      "loss": 1.8857,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 24.853385573955414,
+      "learning_rate": 7.392473118279571e-07,
+      "loss": 1.9314,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 24.241953706581974,
+      "learning_rate": 7.560483870967743e-07,
+      "loss": 1.8315,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 30.963286237241633,
+      "learning_rate": 7.728494623655915e-07,
+      "loss": 1.9002,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.03379591891388,
+      "learning_rate": 7.896505376344087e-07,
+      "loss": 1.884,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 24.578647797403292,
+      "learning_rate": 8.064516129032258e-07,
+      "loss": 1.8904,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.216487036162114,
+      "learning_rate": 8.23252688172043e-07,
+      "loss": 1.8219,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.861193581911664,
+      "learning_rate": 8.400537634408602e-07,
+      "loss": 1.6311,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.299865558645777,
+      "learning_rate": 8.568548387096774e-07,
+      "loss": 1.8558,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.746355187719427,
+      "learning_rate": 8.736559139784947e-07,
+      "loss": 1.7711,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.940626478094718,
+      "learning_rate": 8.904569892473119e-07,
+      "loss": 1.6859,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.6483613759701,
+      "learning_rate": 9.072580645161292e-07,
+      "loss": 1.7343,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.89611142642267,
+      "learning_rate": 9.240591397849464e-07,
+      "loss": 1.6938,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.001859350266486,
+      "learning_rate": 9.408602150537635e-07,
+      "loss": 1.738,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.066884810710853,
+      "learning_rate": 9.576612903225808e-07,
+      "loss": 1.783,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.174269785302588,
+      "learning_rate": 9.744623655913979e-07,
+      "loss": 1.8442,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.12810335166148,
+      "learning_rate": 9.91263440860215e-07,
+      "loss": 1.7999,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.191857447339,
+      "learning_rate": 1.0080645161290323e-06,
+      "loss": 1.7081,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.815036999646132,
+      "learning_rate": 1.0248655913978496e-06,
+      "loss": 1.6522,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.835894404829006,
+      "learning_rate": 1.0416666666666667e-06,
+      "loss": 1.6562,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.63269929975477,
+      "learning_rate": 1.058467741935484e-06,
+      "loss": 1.6752,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.828660268995748,
+      "learning_rate": 1.0752688172043011e-06,
+      "loss": 1.5975,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.533688396632005,
+      "learning_rate": 1.0920698924731184e-06,
+      "loss": 1.6904,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.855998449745552,
+      "learning_rate": 1.1088709677419356e-06,
+      "loss": 1.6313,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 33.592956121234906,
+      "learning_rate": 1.1256720430107527e-06,
+      "loss": 1.7334,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.355880614952422,
+      "learning_rate": 1.14247311827957e-06,
+      "loss": 1.6071,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.581379694577958,
+      "learning_rate": 1.159274193548387e-06,
+      "loss": 1.6236,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.977879814802337,
+      "learning_rate": 1.1760752688172044e-06,
+      "loss": 1.5829,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.8664963134277,
+      "learning_rate": 1.1928763440860217e-06,
+      "loss": 1.695,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.709276914746113,
+      "learning_rate": 1.2096774193548388e-06,
+      "loss": 1.6185,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.65158447171455,
+      "learning_rate": 1.2264784946236561e-06,
+      "loss": 1.5907,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.428125598199816,
+      "learning_rate": 1.2432795698924732e-06,
+      "loss": 1.5583,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.84519660574699,
+      "learning_rate": 1.2600806451612903e-06,
+      "loss": 1.5405,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.767453102040555,
+      "learning_rate": 1.2768817204301076e-06,
+      "loss": 1.5337,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.818777203371633,
+      "learning_rate": 1.2936827956989247e-06,
+      "loss": 1.6202,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.367040069929933,
+      "learning_rate": 1.310483870967742e-06,
+      "loss": 1.5684,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.598127290678107,
+      "learning_rate": 1.3272849462365592e-06,
+      "loss": 1.5376,
+      "step": 395
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.719109719831028,
+      "learning_rate": 1.3440860215053765e-06,
+      "loss": 1.4921,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.86472842753203,
+      "learning_rate": 1.3608870967741936e-06,
+      "loss": 1.5219,
+      "step": 405
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.638436674699083,
+      "learning_rate": 1.3776881720430107e-06,
+      "loss": 1.5141,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.282678581255507,
+      "learning_rate": 1.394489247311828e-06,
+      "loss": 1.4588,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.90193197057512,
+      "learning_rate": 1.4112903225806455e-06,
+      "loss": 1.5744,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.65628343176628,
+      "learning_rate": 1.4280913978494626e-06,
+      "loss": 1.4718,
+      "step": 425
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.634571109448922,
+      "learning_rate": 1.4448924731182797e-06,
+      "loss": 1.5239,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.749287700582828,
+      "learning_rate": 1.461693548387097e-06,
+      "loss": 1.4566,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.547106603930775,
+      "learning_rate": 1.4784946236559141e-06,
+      "loss": 1.5024,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.773314653888498,
+      "learning_rate": 1.4952956989247315e-06,
+      "loss": 1.6064,
+      "step": 445
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.139400485945263,
+      "learning_rate": 1.5120967741935486e-06,
+      "loss": 1.5204,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 25.575817089248652,
+      "learning_rate": 1.5288978494623657e-06,
+      "loss": 1.4468,
+      "step": 455
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.012505506084903,
+      "learning_rate": 1.545698924731183e-06,
+      "loss": 1.4547,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.410132583317328,
+      "learning_rate": 1.5625e-06,
+      "loss": 1.446,
+      "step": 465
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.868370038078499,
+      "learning_rate": 1.5793010752688174e-06,
+      "loss": 1.5959,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.812418592555467,
+      "learning_rate": 1.5961021505376345e-06,
+      "loss": 1.4709,
+      "step": 475
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.905386132006193,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.5097,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.946674581803094,
+      "learning_rate": 1.629704301075269e-06,
+      "loss": 1.4466,
+      "step": 485
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.214144758770525,
+      "learning_rate": 1.646505376344086e-06,
+      "loss": 1.4626,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.39110267796856,
+      "learning_rate": 1.6633064516129033e-06,
+      "loss": 1.5832,
+      "step": 495
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.768822449697835,
+      "learning_rate": 1.6801075268817204e-06,
+      "loss": 1.4171,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.346879819966862,
+      "learning_rate": 1.6969086021505377e-06,
+      "loss": 1.4682,
+      "step": 505
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.776476285237981,
+      "learning_rate": 1.7137096774193548e-06,
+      "loss": 1.4112,
+      "step": 510
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.468416925660526,
+      "learning_rate": 1.7305107526881724e-06,
+      "loss": 1.3957,
+      "step": 515
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.81879140532043,
+      "learning_rate": 1.7473118279569895e-06,
+      "loss": 1.4286,
+      "step": 520
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 22.268339884303696,
+      "learning_rate": 1.7641129032258068e-06,
+      "loss": 1.4274,
+      "step": 525
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.974470923302114,
+      "learning_rate": 1.7809139784946239e-06,
+      "loss": 1.5043,
+      "step": 530
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.01147151337686,
+      "learning_rate": 1.797715053763441e-06,
+      "loss": 1.4032,
+      "step": 535
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 22.10572306421554,
+      "learning_rate": 1.8145161290322583e-06,
+      "loss": 1.4333,
+      "step": 540
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.640677257163176,
+      "learning_rate": 1.8313172043010754e-06,
+      "loss": 1.4681,
+      "step": 545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 26.564187554827637,
+      "learning_rate": 1.8481182795698927e-06,
+      "loss": 1.4594,
+      "step": 550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.08193320266031,
+      "learning_rate": 1.8649193548387098e-06,
+      "loss": 1.4362,
+      "step": 555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 34.685212861993335,
+      "learning_rate": 1.881720430107527e-06,
+      "loss": 1.471,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 43.100869084203225,
+      "learning_rate": 1.8985215053763442e-06,
+      "loss": 1.4642,
+      "step": 565
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.559894937589828,
+      "learning_rate": 1.9153225806451616e-06,
+      "loss": 1.3963,
+      "step": 570
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 38.2485410412665,
+      "learning_rate": 1.9321236559139787e-06,
+      "loss": 1.4386,
+      "step": 575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 30.95124448720418,
+      "learning_rate": 1.9489247311827958e-06,
+      "loss": 1.4151,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 51.70839528402327,
+      "learning_rate": 1.965725806451613e-06,
+      "loss": 1.4209,
+      "step": 585
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 23.196005284612998,
+      "learning_rate": 1.98252688172043e-06,
+      "loss": 1.4012,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 27.1622839764402,
+      "learning_rate": 1.9993279569892475e-06,
+      "loss": 1.3261,
+      "step": 595
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.43452977396187,
+      "learning_rate": 2.0161290322580646e-06,
+      "loss": 1.3766,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.033787281630863,
+      "learning_rate": 2.032930107526882e-06,
+      "loss": 1.4099,
+      "step": 605
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.955820585990788,
+      "learning_rate": 2.0497311827956992e-06,
+      "loss": 1.3794,
+      "step": 610
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.461768151074137,
+      "learning_rate": 2.0665322580645163e-06,
+      "loss": 1.4485,
+      "step": 615
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.64480417285225,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 1.3828,
+      "step": 620
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.057917363440225,
+      "learning_rate": 2.100134408602151e-06,
+      "loss": 1.4158,
+      "step": 625
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 23.390563850096793,
+      "learning_rate": 2.116935483870968e-06,
+      "loss": 1.4126,
+      "step": 630
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 27.856058877726337,
+      "learning_rate": 2.133736559139785e-06,
+      "loss": 1.3153,
+      "step": 635
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.627432697331793,
+      "learning_rate": 2.1505376344086023e-06,
+      "loss": 1.3693,
+      "step": 640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.018250112690833,
+      "learning_rate": 2.1673387096774194e-06,
+      "loss": 1.3771,
+      "step": 645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 24.363765549799407,
+      "learning_rate": 2.184139784946237e-06,
+      "loss": 1.4001,
+      "step": 650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.8886205337284,
+      "learning_rate": 2.200940860215054e-06,
+      "loss": 1.3851,
+      "step": 655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.200499403924693,
+      "learning_rate": 2.217741935483871e-06,
+      "loss": 1.3914,
+      "step": 660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 22.506933751518403,
+      "learning_rate": 2.234543010752688e-06,
+      "loss": 1.3736,
+      "step": 665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.168953266789856,
+      "learning_rate": 2.2513440860215053e-06,
+      "loss": 1.349,
+      "step": 670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 58.48769280628009,
+      "learning_rate": 2.268145161290323e-06,
+      "loss": 1.4582,
+      "step": 675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 34.01636176936607,
+      "learning_rate": 2.28494623655914e-06,
+      "loss": 1.4758,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 25.025355540731983,
+      "learning_rate": 2.301747311827957e-06,
+      "loss": 1.3476,
+      "step": 685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.751839601928431,
+      "learning_rate": 2.318548387096774e-06,
+      "loss": 1.3785,
+      "step": 690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.071805821551914,
+      "learning_rate": 2.3353494623655917e-06,
+      "loss": 1.3635,
+      "step": 695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.602958992356196,
+      "learning_rate": 2.3521505376344088e-06,
+      "loss": 1.3845,
+      "step": 700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.385890285083427,
+      "learning_rate": 2.3689516129032263e-06,
+      "loss": 1.2857,
+      "step": 705
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.215016267194583,
+      "learning_rate": 2.3857526881720434e-06,
+      "loss": 1.3015,
+      "step": 710
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 33.48226663382944,
+      "learning_rate": 2.4025537634408605e-06,
+      "loss": 1.4045,
+      "step": 715
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 23.942731192443755,
+      "learning_rate": 2.4193548387096776e-06,
+      "loss": 1.3649,
+      "step": 720
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.558052499204832,
+      "learning_rate": 2.4361559139784947e-06,
+      "loss": 1.3811,
+      "step": 725
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.69964334303326,
+      "learning_rate": 2.4529569892473122e-06,
+      "loss": 1.3416,
+      "step": 730
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.177544261859945,
+      "learning_rate": 2.4697580645161293e-06,
+      "loss": 1.3243,
+      "step": 735
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.803122016869308,
+      "learning_rate": 2.4865591397849464e-06,
+      "loss": 1.371,
+      "step": 740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.51671956073884,
+      "learning_rate": 2.503360215053764e-06,
+      "loss": 1.3652,
+      "step": 745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.72972602267457,
+      "learning_rate": 2.5201612903225806e-06,
+      "loss": 1.3826,
+      "step": 750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.865572453001107,
+      "learning_rate": 2.536962365591398e-06,
+      "loss": 1.387,
+      "step": 755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.938389282777255,
+      "learning_rate": 2.5537634408602153e-06,
+      "loss": 1.3031,
+      "step": 760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.036068148740473,
+      "learning_rate": 2.570564516129033e-06,
+      "loss": 1.392,
+      "step": 765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.200040828281905,
+      "learning_rate": 2.5873655913978495e-06,
+      "loss": 1.4077,
+      "step": 770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.776949710823484,
+      "learning_rate": 2.604166666666667e-06,
+      "loss": 1.3833,
+      "step": 775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.947572172197686,
+      "learning_rate": 2.620967741935484e-06,
+      "loss": 1.4015,
+      "step": 780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 76.93992013037158,
+      "learning_rate": 2.6377688172043016e-06,
+      "loss": 1.4051,
+      "step": 785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.942371317725181,
+      "learning_rate": 2.6545698924731183e-06,
+      "loss": 1.348,
+      "step": 790
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.086369840086844,
+      "learning_rate": 2.671370967741936e-06,
+      "loss": 1.3453,
+      "step": 795
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 35.89028186175731,
+      "learning_rate": 2.688172043010753e-06,
+      "loss": 1.3807,
+      "step": 800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 29.63422412243611,
+      "learning_rate": 2.70497311827957e-06,
+      "loss": 1.3625,
+      "step": 805
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 55.32993232323523,
+      "learning_rate": 2.721774193548387e-06,
+      "loss": 1.4005,
+      "step": 810
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.984066997357118,
+      "learning_rate": 2.7385752688172047e-06,
+      "loss": 1.3381,
+      "step": 815
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 25.472146032597063,
+      "learning_rate": 2.7553763440860214e-06,
+      "loss": 1.4245,
+      "step": 820
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.457358515963906,
+      "learning_rate": 2.772177419354839e-06,
+      "loss": 1.3625,
+      "step": 825
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.759070196363787,
+      "learning_rate": 2.788978494623656e-06,
+      "loss": 1.448,
+      "step": 830
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 40.79061945114763,
+      "learning_rate": 2.8057795698924735e-06,
+      "loss": 1.3464,
+      "step": 835
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 25.161756708332057,
+      "learning_rate": 2.822580645161291e-06,
+      "loss": 1.3722,
+      "step": 840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.188019881756727,
+      "learning_rate": 2.8393817204301077e-06,
+      "loss": 1.2966,
+      "step": 845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.235137203610662,
+      "learning_rate": 2.8561827956989252e-06,
+      "loss": 1.3905,
+      "step": 850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 36.617132909309895,
+      "learning_rate": 2.872983870967742e-06,
+      "loss": 1.3609,
+      "step": 855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 29.49169037802908,
+      "learning_rate": 2.8897849462365594e-06,
+      "loss": 1.3619,
+      "step": 860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 46.26563028512804,
+      "learning_rate": 2.9065860215053765e-06,
+      "loss": 1.3855,
+      "step": 865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.0915541742763,
+      "learning_rate": 2.923387096774194e-06,
+      "loss": 1.4284,
+      "step": 870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 35.015078004933514,
+      "learning_rate": 2.9401881720430108e-06,
+      "loss": 1.3365,
+      "step": 875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 95.54066941374312,
+      "learning_rate": 2.9569892473118283e-06,
+      "loss": 1.3615,
+      "step": 880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 69.29704172941422,
+      "learning_rate": 2.9737903225806454e-06,
+      "loss": 1.3591,
+      "step": 885
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 28.783068659669784,
+      "learning_rate": 2.990591397849463e-06,
+      "loss": 1.3271,
+      "step": 890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.165110584313982,
+      "learning_rate": 3.0073924731182796e-06,
+      "loss": 1.2884,
+      "step": 895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 41.205307526496874,
+      "learning_rate": 3.024193548387097e-06,
+      "loss": 1.3995,
+      "step": 900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 40.001454194930204,
+      "learning_rate": 3.0409946236559142e-06,
+      "loss": 1.3513,
+      "step": 905
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.56121470148507,
+      "learning_rate": 3.0577956989247313e-06,
+      "loss": 1.2726,
+      "step": 910
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.766228304604502,
+      "learning_rate": 3.0745967741935484e-06,
+      "loss": 1.3304,
+      "step": 915
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.262090744904206,
+      "learning_rate": 3.091397849462366e-06,
+      "loss": 1.3634,
+      "step": 920
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.491180802304278,
+      "learning_rate": 3.1081989247311826e-06,
+      "loss": 1.3582,
+      "step": 925
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 44.89752185269729,
+      "learning_rate": 3.125e-06,
+      "loss": 1.3173,
+      "step": 930
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 35.91291644151354,
+      "learning_rate": 3.1418010752688177e-06,
+      "loss": 1.3348,
+      "step": 935
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.238116458693074,
+      "learning_rate": 3.1586021505376348e-06,
+      "loss": 1.3507,
+      "step": 940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 66.27238623554678,
+      "learning_rate": 3.1754032258064523e-06,
+      "loss": 1.4242,
+      "step": 945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 23.33319086370242,
+      "learning_rate": 3.192204301075269e-06,
+      "loss": 1.3167,
+      "step": 950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 27.96170215869396,
+      "learning_rate": 3.2090053763440865e-06,
+      "loss": 1.3044,
+      "step": 955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 24.96629022725591,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 1.3122,
+      "step": 960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 31.97296302595778,
+      "learning_rate": 3.2426075268817207e-06,
+      "loss": 1.3263,
+      "step": 965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 40.56869211119147,
+      "learning_rate": 3.259408602150538e-06,
+      "loss": 1.3216,
+      "step": 970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 48.86733327981514,
+      "learning_rate": 3.2762096774193553e-06,
+      "loss": 1.3852,
+      "step": 975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 48.8589827825722,
+      "learning_rate": 3.293010752688172e-06,
+      "loss": 1.3942,
+      "step": 980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 105.92172275966337,
+      "learning_rate": 3.3098118279569895e-06,
+      "loss": 1.331,
+      "step": 985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 79.25673823625935,
+      "learning_rate": 3.3266129032258067e-06,
+      "loss": 1.4146,
+      "step": 990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 22.753954756085893,
+      "learning_rate": 3.343413978494624e-06,
+      "loss": 1.3425,
+      "step": 995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.938951839801916,
+      "learning_rate": 3.360215053763441e-06,
+      "loss": 1.2828,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 40.93281890650386,
+      "learning_rate": 3.3770161290322584e-06,
+      "loss": 1.4161,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 23.016726816711316,
+      "learning_rate": 3.3938172043010755e-06,
+      "loss": 1.342,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 19.694780338872466,
+      "learning_rate": 3.4106182795698926e-06,
+      "loss": 1.2934,
+      "step": 1015
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.232488558864302,
+      "learning_rate": 3.4274193548387097e-06,
+      "loss": 1.4007,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.56739533203392,
+      "learning_rate": 3.4442204301075272e-06,
+      "loss": 1.2597,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.55384322451927,
+      "learning_rate": 3.4610215053763447e-06,
+      "loss": 1.3139,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.51771613755045,
+      "learning_rate": 3.4778225806451614e-06,
+      "loss": 1.2369,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.547628388697412,
+      "learning_rate": 3.494623655913979e-06,
+      "loss": 1.3077,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 21.253857468288704,
+      "learning_rate": 3.511424731182796e-06,
+      "loss": 1.2945,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 32.2978178380615,
+      "learning_rate": 3.5282258064516136e-06,
+      "loss": 1.3445,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.768307513613498,
+      "learning_rate": 3.5450268817204303e-06,
+      "loss": 1.3081,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.387095517003354,
+      "learning_rate": 3.5618279569892478e-06,
+      "loss": 1.2361,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.17899433065435,
+      "learning_rate": 3.578629032258065e-06,
+      "loss": 1.307,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 19.757988892716817,
+      "learning_rate": 3.595430107526882e-06,
+      "loss": 1.3318,
+      "step": 1070
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.618417336002926,
+      "learning_rate": 3.612231182795699e-06,
+      "loss": 1.311,
+      "step": 1075
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.71596301767083,
+      "learning_rate": 3.6290322580645166e-06,
+      "loss": 1.226,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.068963240572275,
+      "learning_rate": 3.6458333333333333e-06,
+      "loss": 1.2774,
+      "step": 1085
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.802355495211414,
+      "learning_rate": 3.662634408602151e-06,
+      "loss": 1.3052,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.684098983890863,
+      "learning_rate": 3.679435483870968e-06,
+      "loss": 1.2614,
+      "step": 1095
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.099864032228766,
+      "learning_rate": 3.6962365591397855e-06,
+      "loss": 1.2375,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 44.01005527420453,
+      "learning_rate": 3.713037634408602e-06,
+      "loss": 1.269,
+      "step": 1105
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.69498598023566,
+      "learning_rate": 3.7298387096774197e-06,
+      "loss": 1.3259,
+      "step": 1110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 22.87277001637021,
+      "learning_rate": 3.7466397849462368e-06,
+      "loss": 1.2521,
+      "step": 1115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.733526400008255,
+      "learning_rate": 3.763440860215054e-06,
+      "loss": 1.3009,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.430350713728973,
+      "learning_rate": 3.7802419354838714e-06,
+      "loss": 1.2388,
+      "step": 1125
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.85124486151365,
+      "learning_rate": 3.7970430107526885e-06,
+      "loss": 1.3255,
+      "step": 1130
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.2973639532908,
+      "learning_rate": 3.813844086021506e-06,
+      "loss": 1.2868,
+      "step": 1135
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.044475090556347,
+      "learning_rate": 3.830645161290323e-06,
+      "loss": 1.2921,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.419443385517743,
+      "learning_rate": 3.84744623655914e-06,
+      "loss": 1.3027,
+      "step": 1145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.408727766295128,
+      "learning_rate": 3.864247311827957e-06,
+      "loss": 1.2636,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 28.714229113569143,
+      "learning_rate": 3.8810483870967744e-06,
+      "loss": 1.3094,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.468610643519202,
+      "learning_rate": 3.8978494623655915e-06,
+      "loss": 1.295,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 22.147649353738434,
+      "learning_rate": 3.914650537634409e-06,
+      "loss": 1.268,
+      "step": 1165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 21.815639864910924,
+      "learning_rate": 3.931451612903226e-06,
+      "loss": 1.3022,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.131219829800257,
+      "learning_rate": 3.948252688172044e-06,
+      "loss": 1.3033,
+      "step": 1175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 40.422268139360455,
+      "learning_rate": 3.96505376344086e-06,
+      "loss": 1.2916,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.18996572855082,
+      "learning_rate": 3.981854838709678e-06,
+      "loss": 1.2089,
+      "step": 1185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 34.97991531805703,
+      "learning_rate": 3.998655913978495e-06,
+      "loss": 1.25,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 24.84307661290493,
+      "learning_rate": 4.015456989247312e-06,
+      "loss": 1.3503,
+      "step": 1195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 23.814660056097257,
+      "learning_rate": 4.032258064516129e-06,
+      "loss": 1.2976,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 48.66684356663973,
+      "learning_rate": 4.049059139784946e-06,
+      "loss": 1.278,
+      "step": 1205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.743111074822284,
+      "learning_rate": 4.065860215053764e-06,
+      "loss": 1.2592,
+      "step": 1210
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 24.74596115728974,
+      "learning_rate": 4.082661290322581e-06,
+      "loss": 1.2825,
+      "step": 1215
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 34.36014978790517,
+      "learning_rate": 4.0994623655913985e-06,
+      "loss": 1.2459,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 20.999547137931017,
+      "learning_rate": 4.1162634408602156e-06,
+      "loss": 1.2121,
+      "step": 1225
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.868497455108407,
+      "learning_rate": 4.133064516129033e-06,
+      "loss": 1.2768,
+      "step": 1230
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 30.434809468480765,
+      "learning_rate": 4.14986559139785e-06,
+      "loss": 1.2675,
+      "step": 1235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.616266135748583,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 1.2608,
+      "step": 1240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.532446205798287,
+      "learning_rate": 4.183467741935484e-06,
+      "loss": 1.3375,
+      "step": 1245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.560371038383625,
+      "learning_rate": 4.200268817204302e-06,
+      "loss": 1.2329,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.724212892297027,
+      "learning_rate": 4.217069892473118e-06,
+      "loss": 1.2697,
+      "step": 1255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 22.2529932746503,
+      "learning_rate": 4.233870967741936e-06,
+      "loss": 1.2645,
+      "step": 1260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.558716472596469,
+      "learning_rate": 4.250672043010753e-06,
+      "loss": 1.2197,
+      "step": 1265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 29.247532948847013,
+      "learning_rate": 4.26747311827957e-06,
+      "loss": 1.2919,
+      "step": 1270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.804159591275033,
+      "learning_rate": 4.2842741935483874e-06,
+      "loss": 1.3064,
+      "step": 1275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.409715922052758,
+      "learning_rate": 4.3010752688172045e-06,
+      "loss": 1.2718,
+      "step": 1280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.77842104466719,
+      "learning_rate": 4.317876344086022e-06,
+      "loss": 1.3095,
+      "step": 1285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.685031119948517,
+      "learning_rate": 4.334677419354839e-06,
+      "loss": 1.2832,
+      "step": 1290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.385894858494375,
+      "learning_rate": 4.351478494623656e-06,
+      "loss": 1.3237,
+      "step": 1295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.723338453835758,
+      "learning_rate": 4.368279569892474e-06,
+      "loss": 1.2666,
+      "step": 1300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.439591311170961,
+      "learning_rate": 4.385080645161291e-06,
+      "loss": 1.2909,
+      "step": 1305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.514215609608113,
+      "learning_rate": 4.401881720430108e-06,
+      "loss": 1.2327,
+      "step": 1310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.845827804092984,
+      "learning_rate": 4.418682795698925e-06,
+      "loss": 1.2617,
+      "step": 1315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.823360940605603,
+      "learning_rate": 4.435483870967742e-06,
+      "loss": 1.2915,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.472975718436162,
+      "learning_rate": 4.452284946236559e-06,
+      "loss": 1.302,
+      "step": 1325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.284002557459615,
+      "learning_rate": 4.469086021505376e-06,
+      "loss": 1.2788,
+      "step": 1330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 38.27911520285089,
+      "learning_rate": 4.485887096774194e-06,
+      "loss": 1.2552,
+      "step": 1335
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.432535268516432,
+      "learning_rate": 4.502688172043011e-06,
+      "loss": 1.2759,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.98747211347199,
+      "learning_rate": 4.5194892473118286e-06,
+      "loss": 1.2844,
+      "step": 1345
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.511423262582975,
+      "learning_rate": 4.536290322580646e-06,
+      "loss": 1.2283,
+      "step": 1350
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 20.514439149634352,
+      "learning_rate": 4.553091397849463e-06,
+      "loss": 1.2353,
+      "step": 1355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 16.015991119714123,
+      "learning_rate": 4.56989247311828e-06,
+      "loss": 1.2959,
+      "step": 1360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.106660893981587,
+      "learning_rate": 4.586693548387097e-06,
+      "loss": 1.2239,
+      "step": 1365
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 29.853730062938347,
+      "learning_rate": 4.603494623655914e-06,
+      "loss": 1.3194,
+      "step": 1370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.685737996052852,
+      "learning_rate": 4.620295698924732e-06,
+      "loss": 1.2148,
+      "step": 1375
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.877331149990967,
+      "learning_rate": 4.637096774193548e-06,
+      "loss": 1.2045,
+      "step": 1380
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.008218814296683,
+      "learning_rate": 4.653897849462366e-06,
+      "loss": 1.2477,
+      "step": 1385
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 22.743570797984013,
+      "learning_rate": 4.670698924731183e-06,
+      "loss": 1.2775,
+      "step": 1390
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 29.524974297593605,
+      "learning_rate": 4.6875000000000004e-06,
+      "loss": 1.299,
+      "step": 1395
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.37930037376844,
+      "learning_rate": 4.7043010752688175e-06,
+      "loss": 1.3329,
+      "step": 1400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.598571626831996,
+      "learning_rate": 4.721102150537635e-06,
+      "loss": 1.2942,
+      "step": 1405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.752457322963501,
+      "learning_rate": 4.737903225806453e-06,
+      "loss": 1.2408,
+      "step": 1410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.210288312721424,
+      "learning_rate": 4.754704301075269e-06,
+      "loss": 1.2762,
+      "step": 1415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 15.028431784981668,
+      "learning_rate": 4.771505376344087e-06,
+      "loss": 1.2817,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.993762858927266,
+      "learning_rate": 4.788306451612904e-06,
+      "loss": 1.2521,
+      "step": 1425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 26.11557096134634,
+      "learning_rate": 4.805107526881721e-06,
+      "loss": 1.2275,
+      "step": 1430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 23.808664429913215,
+      "learning_rate": 4.821908602150538e-06,
+      "loss": 1.2729,
+      "step": 1435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.214083146280673,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 1.2103,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.950068491419977,
+      "learning_rate": 4.855510752688172e-06,
+      "loss": 1.2407,
+      "step": 1445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 49.77700236183305,
+      "learning_rate": 4.872311827956989e-06,
+      "loss": 1.2599,
+      "step": 1450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.2907307607193,
+      "learning_rate": 4.8891129032258065e-06,
+      "loss": 1.2365,
+      "step": 1455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 35.24775519440709,
+      "learning_rate": 4.9059139784946245e-06,
+      "loss": 1.2727,
+      "step": 1460
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.209125211904805,
+      "learning_rate": 4.922715053763441e-06,
+      "loss": 1.2516,
+      "step": 1465
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.378215907461524,
+      "learning_rate": 4.939516129032259e-06,
+      "loss": 1.2488,
+      "step": 1470
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 17.6633476819897,
+      "learning_rate": 4.956317204301076e-06,
+      "loss": 1.2668,
+      "step": 1475
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.740924281703688,
+      "learning_rate": 4.973118279569893e-06,
+      "loss": 1.249,
+      "step": 1480
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.726705345231204,
+      "learning_rate": 4.98991935483871e-06,
+      "loss": 1.3307,
+      "step": 1485
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.59599163185456,
+      "learning_rate": 5.006720430107528e-06,
+      "loss": 1.2854,
+      "step": 1490
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 41.090146378069534,
+      "learning_rate": 5.023521505376344e-06,
+      "loss": 1.291,
+      "step": 1495
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.10786448778689,
+      "learning_rate": 5.040322580645161e-06,
+      "loss": 1.2843,
+      "step": 1500
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.353102002707647,
+      "learning_rate": 5.057123655913979e-06,
+      "loss": 1.2556,
+      "step": 1505
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.89417888688282,
+      "learning_rate": 5.073924731182796e-06,
+      "loss": 1.2582,
+      "step": 1510
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 38.41649367763352,
+      "learning_rate": 5.090725806451613e-06,
+      "loss": 1.3358,
+      "step": 1515
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.107954803233616,
+      "learning_rate": 5.1075268817204305e-06,
+      "loss": 1.28,
+      "step": 1520
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.825179318943986,
+      "learning_rate": 5.124327956989248e-06,
+      "loss": 1.2516,
+      "step": 1525
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 26.492711678315192,
+      "learning_rate": 5.141129032258066e-06,
+      "loss": 1.2652,
+      "step": 1530
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 17.328853885466696,
+      "learning_rate": 5.157930107526882e-06,
+      "loss": 1.2965,
+      "step": 1535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 17.50828288344384,
+      "learning_rate": 5.174731182795699e-06,
+      "loss": 1.2399,
+      "step": 1540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.950092920119129,
+      "learning_rate": 5.191532258064517e-06,
+      "loss": 1.3137,
+      "step": 1545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 20.052669655145927,
+      "learning_rate": 5.208333333333334e-06,
+      "loss": 1.3046,
+      "step": 1550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 53.370358081765325,
+      "learning_rate": 5.22513440860215e-06,
+      "loss": 1.2716,
+      "step": 1555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 23.64171905195358,
+      "learning_rate": 5.241935483870968e-06,
+      "loss": 1.2413,
+      "step": 1560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.501262717248586,
+      "learning_rate": 5.258736559139785e-06,
+      "loss": 1.3243,
+      "step": 1565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 46.31409454126866,
+      "learning_rate": 5.275537634408603e-06,
+      "loss": 1.2501,
+      "step": 1570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 24.436697787509363,
+      "learning_rate": 5.2923387096774195e-06,
+      "loss": 1.2446,
+      "step": 1575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 15.211955398526683,
+      "learning_rate": 5.309139784946237e-06,
+      "loss": 1.2485,
+      "step": 1580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 51.17545641781786,
+      "learning_rate": 5.3259408602150546e-06,
+      "loss": 1.2664,
+      "step": 1585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.73729158273849,
+      "learning_rate": 5.342741935483872e-06,
+      "loss": 1.2712,
+      "step": 1590
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.40335244214581,
+      "learning_rate": 5.359543010752689e-06,
+      "loss": 1.2636,
+      "step": 1595
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.214516983760586,
+      "learning_rate": 5.376344086021506e-06,
+      "loss": 1.2158,
+      "step": 1600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.34966113388582,
+      "learning_rate": 5.393145161290323e-06,
+      "loss": 1.2513,
+      "step": 1605
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 27.012632212205652,
+      "learning_rate": 5.40994623655914e-06,
+      "loss": 1.2153,
+      "step": 1610
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 66.26220201569467,
+      "learning_rate": 5.426747311827958e-06,
+      "loss": 1.2941,
+      "step": 1615
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.447517864939536,
+      "learning_rate": 5.443548387096774e-06,
+      "loss": 1.2104,
+      "step": 1620
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 53.03873905700024,
+      "learning_rate": 5.460349462365591e-06,
+      "loss": 1.2912,
+      "step": 1625
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 82.47953703104959,
+      "learning_rate": 5.477150537634409e-06,
+      "loss": 1.2963,
+      "step": 1630
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 59.35832189434322,
+      "learning_rate": 5.4939516129032264e-06,
+      "loss": 1.299,
+      "step": 1635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.23456282221694,
+      "learning_rate": 5.510752688172043e-06,
+      "loss": 1.2955,
+      "step": 1640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.208909499987218,
+      "learning_rate": 5.527553763440861e-06,
+      "loss": 1.2827,
+      "step": 1645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.814579537509662,
+      "learning_rate": 5.544354838709678e-06,
+      "loss": 1.2782,
+      "step": 1650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 44.84766228129635,
+      "learning_rate": 5.561155913978496e-06,
+      "loss": 1.3388,
+      "step": 1655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 80.02157305544218,
+      "learning_rate": 5.577956989247312e-06,
+      "loss": 1.3387,
+      "step": 1660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 47.30693371027271,
+      "learning_rate": 5.594758064516129e-06,
+      "loss": 1.2853,
+      "step": 1665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 31.111971822450105,
+      "learning_rate": 5.611559139784947e-06,
+      "loss": 1.3207,
+      "step": 1670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 29.323472551959778,
+      "learning_rate": 5.628360215053764e-06,
+      "loss": 1.2522,
+      "step": 1675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 40.92036524504599,
+      "learning_rate": 5.645161290322582e-06,
+      "loss": 1.3008,
+      "step": 1680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 33.7231465176679,
+      "learning_rate": 5.661962365591398e-06,
+      "loss": 1.2796,
+      "step": 1685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.506027765246968,
+      "learning_rate": 5.678763440860215e-06,
+      "loss": 1.2689,
+      "step": 1690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.309943493618878,
+      "learning_rate": 5.6955645161290325e-06,
+      "loss": 1.2803,
+      "step": 1695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.260279103971566,
+      "learning_rate": 5.7123655913978505e-06,
+      "loss": 1.2479,
+      "step": 1700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.417965761132848,
+      "learning_rate": 5.729166666666667e-06,
+      "loss": 1.2659,
+      "step": 1705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.069359196115766,
+      "learning_rate": 5.745967741935484e-06,
+      "loss": 1.258,
+      "step": 1710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 31.862010501733916,
+      "learning_rate": 5.762768817204302e-06,
+      "loss": 1.2745,
+      "step": 1715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.171126284799604,
+      "learning_rate": 5.779569892473119e-06,
+      "loss": 1.2201,
+      "step": 1720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.839283951323784,
+      "learning_rate": 5.796370967741935e-06,
+      "loss": 1.2394,
+      "step": 1725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 28.093728374613843,
+      "learning_rate": 5.813172043010753e-06,
+      "loss": 1.2638,
+      "step": 1730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 44.58712437099615,
+      "learning_rate": 5.82997311827957e-06,
+      "loss": 1.263,
+      "step": 1735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.611237724562416,
+      "learning_rate": 5.846774193548388e-06,
+      "loss": 1.2396,
+      "step": 1740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.721134129684025,
+      "learning_rate": 5.863575268817204e-06,
+      "loss": 1.2612,
+      "step": 1745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 28.526923438499946,
+      "learning_rate": 5.8803763440860215e-06,
+      "loss": 1.2348,
+      "step": 1750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 28.344913656755676,
+      "learning_rate": 5.8971774193548394e-06,
+      "loss": 1.2826,
+      "step": 1755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.471422006525902,
+      "learning_rate": 5.9139784946236566e-06,
+      "loss": 1.2943,
+      "step": 1760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.434390641845038,
+      "learning_rate": 5.930779569892473e-06,
+      "loss": 1.2229,
+      "step": 1765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.75539696310857,
+      "learning_rate": 5.947580645161291e-06,
+      "loss": 1.2526,
+      "step": 1770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.3925293626586,
+      "learning_rate": 5.964381720430108e-06,
+      "loss": 1.2394,
+      "step": 1775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.786185248449616,
+      "learning_rate": 5.981182795698926e-06,
+      "loss": 1.2927,
+      "step": 1780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 28.326800741035534,
+      "learning_rate": 5.997983870967743e-06,
+      "loss": 1.3106,
+      "step": 1785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.276235173795225,
+      "learning_rate": 6.014784946236559e-06,
+      "loss": 1.2011,
+      "step": 1790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.043491768486806,
+      "learning_rate": 6.031586021505377e-06,
+      "loss": 1.31,
+      "step": 1795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.982912118702751,
+      "learning_rate": 6.048387096774194e-06,
+      "loss": 1.1834,
+      "step": 1800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.46029378615589,
+      "learning_rate": 6.065188172043011e-06,
+      "loss": 1.2058,
+      "step": 1805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.56457697533073,
+      "learning_rate": 6.0819892473118284e-06,
+      "loss": 1.2209,
+      "step": 1810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.007818323470907,
+      "learning_rate": 6.0987903225806455e-06,
+      "loss": 1.2084,
+      "step": 1815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 30.888852345480156,
+      "learning_rate": 6.115591397849463e-06,
+      "loss": 1.2878,
+      "step": 1820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.206797920592592,
+      "learning_rate": 6.132392473118281e-06,
+      "loss": 1.256,
+      "step": 1825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 41.77653719952749,
+      "learning_rate": 6.149193548387097e-06,
+      "loss": 1.2143,
+      "step": 1830
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 22.773443987679865,
+      "learning_rate": 6.165994623655914e-06,
+      "loss": 1.2212,
+      "step": 1835
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.003527818874442,
+      "learning_rate": 6.182795698924732e-06,
+      "loss": 1.1451,
+      "step": 1840
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.832295518728609,
+      "learning_rate": 6.199596774193549e-06,
+      "loss": 1.2371,
+      "step": 1845
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.106996556562182,
+      "learning_rate": 6.216397849462365e-06,
+      "loss": 1.2564,
+      "step": 1850
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.6321798945971,
+      "learning_rate": 6.233198924731183e-06,
+      "loss": 1.235,
+      "step": 1855
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.246030421588882,
+      "learning_rate": 6.25e-06,
+      "loss": 1.3035,
+      "step": 1860
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.37717478205693,
+      "learning_rate": 6.266801075268818e-06,
+      "loss": 1.2778,
+      "step": 1865
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.24027128069132,
+      "learning_rate": 6.283602150537635e-06,
+      "loss": 1.272,
+      "step": 1870
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 34.70380358345755,
+      "learning_rate": 6.300403225806452e-06,
+      "loss": 1.2512,
+      "step": 1875
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 46.86598098882556,
+      "learning_rate": 6.3172043010752696e-06,
+      "loss": 1.2413,
+      "step": 1880
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 24.5265225085656,
+      "learning_rate": 6.334005376344087e-06,
+      "loss": 1.2549,
+      "step": 1885
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.176692968830055,
+      "learning_rate": 6.350806451612905e-06,
+      "loss": 1.2243,
+      "step": 1890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.623066798462252,
+      "learning_rate": 6.367607526881721e-06,
+      "loss": 1.2203,
+      "step": 1895
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.506001827294856,
+      "learning_rate": 6.384408602150538e-06,
+      "loss": 1.2448,
+      "step": 1900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 24.346786364555726,
+      "learning_rate": 6.401209677419356e-06,
+      "loss": 1.1681,
+      "step": 1905
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 22.37322556152353,
+      "learning_rate": 6.418010752688173e-06,
+      "loss": 1.2679,
+      "step": 1910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 18.538859354210924,
+      "learning_rate": 6.434811827956989e-06,
+      "loss": 1.225,
+      "step": 1915
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 34.54548609395492,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 1.2446,
+      "step": 1920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.82382489549032,
+      "learning_rate": 6.468413978494624e-06,
+      "loss": 1.2507,
+      "step": 1925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.144353601820267,
+      "learning_rate": 6.4852150537634414e-06,
+      "loss": 1.2938,
+      "step": 1930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 29.58265058763573,
+      "learning_rate": 6.502016129032258e-06,
+      "loss": 1.2952,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.721343926876305,
+      "learning_rate": 6.518817204301076e-06,
+      "loss": 1.1828,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.514261400823013,
+      "learning_rate": 6.535618279569893e-06,
+      "loss": 1.2053,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.533730854979197,
+      "learning_rate": 6.552419354838711e-06,
+      "loss": 1.2568,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.386926044621365,
+      "learning_rate": 6.569220430107528e-06,
+      "loss": 1.2712,
+      "step": 1955
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.035649299385856,
+      "learning_rate": 6.586021505376344e-06,
+      "loss": 1.1855,
+      "step": 1960
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.69163264701267,
+      "learning_rate": 6.602822580645162e-06,
+      "loss": 1.2761,
+      "step": 1965
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 37.0944311318195,
+      "learning_rate": 6.619623655913979e-06,
+      "loss": 1.2577,
+      "step": 1970
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.865249614590498,
+      "learning_rate": 6.636424731182797e-06,
+      "loss": 1.26,
+      "step": 1975
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.904200524254145,
+      "learning_rate": 6.653225806451613e-06,
+      "loss": 1.2764,
+      "step": 1980
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.933867176670313,
+      "learning_rate": 6.67002688172043e-06,
+      "loss": 1.2721,
+      "step": 1985
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 29.639862278436,
+      "learning_rate": 6.686827956989248e-06,
+      "loss": 1.2208,
+      "step": 1990
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 25.54302227116381,
+      "learning_rate": 6.7036290322580655e-06,
+      "loss": 1.2142,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.655438913627023,
+      "learning_rate": 6.720430107526882e-06,
+      "loss": 1.2235,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.24343043881259,
+      "learning_rate": 6.7372311827957e-06,
+      "loss": 1.2186,
+      "step": 2005
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.287545958285396,
+      "learning_rate": 6.754032258064517e-06,
+      "loss": 1.2351,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 27.61706184500475,
+      "learning_rate": 6.770833333333334e-06,
+      "loss": 1.219,
+      "step": 2015
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.006823598473522,
+      "learning_rate": 6.787634408602151e-06,
+      "loss": 1.2106,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 24.899259234998013,
+      "learning_rate": 6.804435483870968e-06,
+      "loss": 1.2471,
+      "step": 2025
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 108.13820601181524,
+      "learning_rate": 6.821236559139785e-06,
+      "loss": 1.3069,
+      "step": 2030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 51.41482122176862,
+      "learning_rate": 6.838037634408603e-06,
+      "loss": 1.274,
+      "step": 2035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 27.969160983681736,
+      "learning_rate": 6.854838709677419e-06,
+      "loss": 1.2421,
+      "step": 2040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 116.52262746448919,
+      "learning_rate": 6.8716397849462365e-06,
+      "loss": 1.209,
+      "step": 2045
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 28.9605266075955,
+      "learning_rate": 6.8884408602150544e-06,
+      "loss": 1.2727,
+      "step": 2050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 20.262447992796503,
+      "learning_rate": 6.9052419354838715e-06,
+      "loss": 1.2644,
+      "step": 2055
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.79893422071457,
+      "learning_rate": 6.9220430107526895e-06,
+      "loss": 1.2442,
+      "step": 2060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 23.597737721499232,
+      "learning_rate": 6.938844086021506e-06,
+      "loss": 1.241,
+      "step": 2065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 24.518983087782463,
+      "learning_rate": 6.955645161290323e-06,
+      "loss": 1.2307,
+      "step": 2070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 20.560253477289546,
+      "learning_rate": 6.972446236559141e-06,
+      "loss": 1.2691,
+      "step": 2075
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 25.848118116745248,
+      "learning_rate": 6.989247311827958e-06,
+      "loss": 1.2724,
+      "step": 2080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.914019924510955,
+      "learning_rate": 7.006048387096774e-06,
+      "loss": 1.2748,
+      "step": 2085
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.925737866573701,
+      "learning_rate": 7.022849462365592e-06,
+      "loss": 1.2468,
+      "step": 2090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.000718814137622,
+      "learning_rate": 7.039650537634409e-06,
+      "loss": 1.2004,
+      "step": 2095
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.595339324971437,
+      "learning_rate": 7.056451612903227e-06,
+      "loss": 1.2976,
+      "step": 2100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.463529208058105,
+      "learning_rate": 7.073252688172043e-06,
+      "loss": 1.2366,
+      "step": 2105
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 29.45323360640321,
+      "learning_rate": 7.0900537634408605e-06,
+      "loss": 1.2691,
+      "step": 2110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.03636155828545,
+      "learning_rate": 7.1068548387096785e-06,
+      "loss": 1.2626,
+      "step": 2115
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 41.205485411736404,
+      "learning_rate": 7.1236559139784956e-06,
+      "loss": 1.2255,
+      "step": 2120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 19.8354928734948,
+      "learning_rate": 7.140456989247312e-06,
+      "loss": 1.2627,
+      "step": 2125
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.143130418752412,
+      "learning_rate": 7.15725806451613e-06,
+      "loss": 1.2636,
+      "step": 2130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 26.329639735822322,
+      "learning_rate": 7.174059139784947e-06,
+      "loss": 1.2913,
+      "step": 2135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.53974742117649,
+      "learning_rate": 7.190860215053764e-06,
+      "loss": 1.1796,
+      "step": 2140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.81150442809384,
+      "learning_rate": 7.207661290322582e-06,
+      "loss": 1.2697,
+      "step": 2145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 41.27431702908246,
+      "learning_rate": 7.224462365591398e-06,
+      "loss": 1.2155,
+      "step": 2150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.502773503666184,
+      "learning_rate": 7.241263440860215e-06,
+      "loss": 1.2091,
+      "step": 2155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.51563458004467,
+      "learning_rate": 7.258064516129033e-06,
+      "loss": 1.2635,
+      "step": 2160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.575762923082294,
+      "learning_rate": 7.27486559139785e-06,
+      "loss": 1.221,
+      "step": 2165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 21.278627675604866,
+      "learning_rate": 7.291666666666667e-06,
+      "loss": 1.2411,
+      "step": 2170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.04465876557275,
+      "learning_rate": 7.3084677419354845e-06,
+      "loss": 1.2134,
+      "step": 2175
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.011104679604054,
+      "learning_rate": 7.325268817204302e-06,
+      "loss": 1.2422,
+      "step": 2180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.353363056360596,
+      "learning_rate": 7.34206989247312e-06,
+      "loss": 1.2202,
+      "step": 2185
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.87880319538541,
+      "learning_rate": 7.358870967741936e-06,
+      "loss": 1.2097,
+      "step": 2190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.533815983023235,
+      "learning_rate": 7.375672043010753e-06,
+      "loss": 1.2266,
+      "step": 2195
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 55.96590550959414,
+      "learning_rate": 7.392473118279571e-06,
+      "loss": 1.2446,
+      "step": 2200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 27.6433561291497,
+      "learning_rate": 7.409274193548388e-06,
+      "loss": 1.2762,
+      "step": 2205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 49.82007468136901,
+      "learning_rate": 7.426075268817204e-06,
+      "loss": 1.2283,
+      "step": 2210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.84247480602059,
+      "learning_rate": 7.442876344086022e-06,
+      "loss": 1.2353,
+      "step": 2215
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.413311841272398,
+      "learning_rate": 7.459677419354839e-06,
+      "loss": 1.2622,
+      "step": 2220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 40.40763721338431,
+      "learning_rate": 7.476478494623656e-06,
+      "loss": 1.2936,
+      "step": 2225
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 82.94010690130378,
+      "learning_rate": 7.4932795698924735e-06,
+      "loss": 1.2799,
+      "step": 2230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 40.39126638765991,
+      "learning_rate": 7.510080645161291e-06,
+      "loss": 1.3345,
+      "step": 2235
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 47.39627347260159,
+      "learning_rate": 7.526881720430108e-06,
+      "loss": 1.228,
+      "step": 2240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.629121665795358,
+      "learning_rate": 7.543682795698926e-06,
+      "loss": 1.2331,
+      "step": 2245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.306221861015311,
+      "learning_rate": 7.560483870967743e-06,
+      "loss": 1.2713,
+      "step": 2250
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 24.542736233239204,
+      "learning_rate": 7.577284946236559e-06,
+      "loss": 1.2892,
+      "step": 2255
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 21.07591070120843,
+      "learning_rate": 7.594086021505377e-06,
+      "loss": 1.2692,
+      "step": 2260
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 29.56250640158638,
+      "learning_rate": 7.610887096774194e-06,
+      "loss": 1.2271,
+      "step": 2265
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 46.59319817819226,
+      "learning_rate": 7.627688172043012e-06,
+      "loss": 1.2268,
+      "step": 2270
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 84.07647878086206,
+      "learning_rate": 7.644489247311827e-06,
+      "loss": 1.2491,
+      "step": 2275
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 42.82187183604961,
+      "learning_rate": 7.661290322580646e-06,
+      "loss": 1.2686,
+      "step": 2280
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 63.93417339646263,
+      "learning_rate": 7.678091397849463e-06,
+      "loss": 1.2721,
+      "step": 2285
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 33.91029033440181,
+      "learning_rate": 7.69489247311828e-06,
+      "loss": 1.2616,
+      "step": 2290
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.582134518943782,
+      "learning_rate": 7.711693548387098e-06,
+      "loss": 1.2684,
+      "step": 2295
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 24.52154465326118,
+      "learning_rate": 7.728494623655915e-06,
+      "loss": 1.268,
+      "step": 2300
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 39.68127286315339,
+      "learning_rate": 7.745295698924732e-06,
+      "loss": 1.2984,
+      "step": 2305
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 64.21209850893193,
+      "learning_rate": 7.762096774193549e-06,
+      "loss": 1.3044,
+      "step": 2310
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 51.574901468650154,
+      "learning_rate": 7.778897849462366e-06,
+      "loss": 1.2065,
+      "step": 2315
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 29.1347166115766,
+      "learning_rate": 7.795698924731183e-06,
+      "loss": 1.2155,
+      "step": 2320
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.815362147149694,
+      "learning_rate": 7.8125e-06,
+      "loss": 1.2377,
+      "step": 2325
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 31.023853624949705,
+      "learning_rate": 7.829301075268817e-06,
+      "loss": 1.2757,
+      "step": 2330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.259718591127633,
+      "learning_rate": 7.846102150537636e-06,
+      "loss": 1.2178,
+      "step": 2335
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.77824590902307,
+      "learning_rate": 7.862903225806451e-06,
+      "loss": 1.27,
+      "step": 2340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 26.383993695953727,
+      "learning_rate": 7.879704301075269e-06,
+      "loss": 1.2079,
+      "step": 2345
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 26.736318983340148,
+      "learning_rate": 7.896505376344087e-06,
+      "loss": 1.2078,
+      "step": 2350
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 27.43805184607868,
+      "learning_rate": 7.913306451612904e-06,
+      "loss": 1.2061,
+      "step": 2355
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.12846401107268,
+      "learning_rate": 7.93010752688172e-06,
+      "loss": 1.2931,
+      "step": 2360
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.723056959950561,
+      "learning_rate": 7.946908602150539e-06,
+      "loss": 1.1951,
+      "step": 2365
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.066035570374572,
+      "learning_rate": 7.963709677419356e-06,
+      "loss": 1.1917,
+      "step": 2370
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.15126653319653,
+      "learning_rate": 7.980510752688173e-06,
+      "loss": 1.2501,
+      "step": 2375
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.825376500667879,
+      "learning_rate": 7.99731182795699e-06,
+      "loss": 1.2303,
+      "step": 2380
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.19258369455789,
+      "learning_rate": 8.014112903225807e-06,
+      "loss": 1.2147,
+      "step": 2385
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.237466445781678,
+      "learning_rate": 8.030913978494624e-06,
+      "loss": 1.1802,
+      "step": 2390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 28.40527681526738,
+      "learning_rate": 8.047715053763441e-06,
+      "loss": 1.262,
+      "step": 2395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.153614338557704,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 1.2535,
+      "step": 2400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.013184380961132,
+      "learning_rate": 8.081317204301075e-06,
+      "loss": 1.3406,
+      "step": 2405
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.551706178969816,
+      "learning_rate": 8.098118279569893e-06,
+      "loss": 1.2798,
+      "step": 2410
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 29.301817149313784,
+      "learning_rate": 8.114919354838711e-06,
+      "loss": 1.2021,
+      "step": 2415
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.604552659024082,
+      "learning_rate": 8.131720430107529e-06,
+      "loss": 1.2193,
+      "step": 2420
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.661579135545585,
+      "learning_rate": 8.148521505376344e-06,
+      "loss": 1.1753,
+      "step": 2425
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.732871735188713,
+      "learning_rate": 8.165322580645163e-06,
+      "loss": 1.1896,
+      "step": 2430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.94264730965134,
+      "learning_rate": 8.18212365591398e-06,
+      "loss": 1.294,
+      "step": 2435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.98862683040616,
+      "learning_rate": 8.198924731182797e-06,
+      "loss": 1.3597,
+      "step": 2440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.348562191116157,
+      "learning_rate": 8.215725806451614e-06,
+      "loss": 1.1981,
+      "step": 2445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.382309784539512,
+      "learning_rate": 8.232526881720431e-06,
+      "loss": 1.197,
+      "step": 2450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.802639889690079,
+      "learning_rate": 8.249327956989248e-06,
+      "loss": 1.2518,
+      "step": 2455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.105120008248589,
+      "learning_rate": 8.266129032258065e-06,
+      "loss": 1.2412,
+      "step": 2460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 22.5005191369109,
+      "learning_rate": 8.282930107526882e-06,
+      "loss": 1.2095,
+      "step": 2465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 55.06496286963062,
+      "learning_rate": 8.2997311827957e-06,
+      "loss": 1.2292,
+      "step": 2470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 25.40056685381794,
+      "learning_rate": 8.316532258064517e-06,
+      "loss": 1.2229,
+      "step": 2475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.75732050971706,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.276,
+      "step": 2480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.259188424158435,
+      "learning_rate": 8.35013440860215e-06,
+      "loss": 1.2295,
+      "step": 2485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 48.034449754252755,
+      "learning_rate": 8.366935483870968e-06,
+      "loss": 1.2694,
+      "step": 2490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 27.913204135569178,
+      "learning_rate": 8.383736559139785e-06,
+      "loss": 1.2747,
+      "step": 2495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 34.71178286331895,
+      "learning_rate": 8.400537634408604e-06,
+      "loss": 1.2141,
+      "step": 2500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.01748344569644,
+      "learning_rate": 8.41733870967742e-06,
+      "loss": 1.2222,
+      "step": 2505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.43993883230521,
+      "learning_rate": 8.434139784946236e-06,
+      "loss": 1.258,
+      "step": 2510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 38.56997728761633,
+      "learning_rate": 8.450940860215055e-06,
+      "loss": 1.2001,
+      "step": 2515
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 22.647653948196986,
+      "learning_rate": 8.467741935483872e-06,
+      "loss": 1.2565,
+      "step": 2520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 31.804400232690394,
+      "learning_rate": 8.48454301075269e-06,
+      "loss": 1.2916,
+      "step": 2525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.568518363297464,
+      "learning_rate": 8.501344086021506e-06,
+      "loss": 1.2009,
+      "step": 2530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.26372201231781,
+      "learning_rate": 8.518145161290324e-06,
+      "loss": 1.209,
+      "step": 2535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 46.54550273923697,
+      "learning_rate": 8.53494623655914e-06,
+      "loss": 1.245,
+      "step": 2540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.219462432954233,
+      "learning_rate": 8.551747311827958e-06,
+      "loss": 1.2544,
+      "step": 2545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 44.110924602079656,
+      "learning_rate": 8.568548387096775e-06,
+      "loss": 1.2368,
+      "step": 2550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 43.35815365202579,
+      "learning_rate": 8.585349462365592e-06,
+      "loss": 1.2318,
+      "step": 2555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.019119762296853,
+      "learning_rate": 8.602150537634409e-06,
+      "loss": 1.2579,
+      "step": 2560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.597771671863768,
+      "learning_rate": 8.618951612903226e-06,
+      "loss": 1.255,
+      "step": 2565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 50.13693425901624,
+      "learning_rate": 8.635752688172043e-06,
+      "loss": 1.256,
+      "step": 2570
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.904065398841818,
+      "learning_rate": 8.65255376344086e-06,
+      "loss": 1.2039,
+      "step": 2575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 37.45772381933171,
+      "learning_rate": 8.669354838709677e-06,
+      "loss": 1.2229,
+      "step": 2580
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 39.27057013113243,
+      "learning_rate": 8.686155913978496e-06,
+      "loss": 1.2191,
+      "step": 2585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 70.60558760616007,
+      "learning_rate": 8.702956989247312e-06,
+      "loss": 1.3142,
+      "step": 2590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 28.19659398732191,
+      "learning_rate": 8.719758064516129e-06,
+      "loss": 1.2259,
+      "step": 2595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 23.9566064510682,
+      "learning_rate": 8.736559139784948e-06,
+      "loss": 1.2543,
+      "step": 2600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.671051560916663,
+      "learning_rate": 8.753360215053765e-06,
+      "loss": 1.2193,
+      "step": 2605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 29.838824406491774,
+      "learning_rate": 8.770161290322582e-06,
+      "loss": 1.2308,
+      "step": 2610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.987977176453866,
+      "learning_rate": 8.786962365591399e-06,
+      "loss": 1.2298,
+      "step": 2615
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.52133772623778,
+      "learning_rate": 8.803763440860216e-06,
+      "loss": 1.2272,
+      "step": 2620
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.693819052666122,
+      "learning_rate": 8.820564516129033e-06,
+      "loss": 1.2678,
+      "step": 2625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.076796566784243,
+      "learning_rate": 8.83736559139785e-06,
+      "loss": 1.1967,
+      "step": 2630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 29.791845399978822,
+      "learning_rate": 8.854166666666667e-06,
+      "loss": 1.2163,
+      "step": 2635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 47.53203127521539,
+      "learning_rate": 8.870967741935484e-06,
+      "loss": 1.2407,
+      "step": 2640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.00670809300897,
+      "learning_rate": 8.887768817204302e-06,
+      "loss": 1.238,
+      "step": 2645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.108573334947934,
+      "learning_rate": 8.904569892473119e-06,
+      "loss": 1.2901,
+      "step": 2650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.094412129400425,
+      "learning_rate": 8.921370967741936e-06,
+      "loss": 1.1964,
+      "step": 2655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.235105946719376,
+      "learning_rate": 8.938172043010753e-06,
+      "loss": 1.245,
+      "step": 2660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 41.31431089609605,
+      "learning_rate": 8.95497311827957e-06,
+      "loss": 1.2624,
+      "step": 2665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.115421164459278,
+      "learning_rate": 8.971774193548389e-06,
+      "loss": 1.231,
+      "step": 2670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 55.68637807130677,
+      "learning_rate": 8.988575268817204e-06,
+      "loss": 1.2537,
+      "step": 2675
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 44.42132523032244,
+      "learning_rate": 9.005376344086021e-06,
+      "loss": 1.1876,
+      "step": 2680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 27.949890360806133,
+      "learning_rate": 9.02217741935484e-06,
+      "loss": 1.2117,
+      "step": 2685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 40.761220118311314,
+      "learning_rate": 9.038978494623657e-06,
+      "loss": 1.2633,
+      "step": 2690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.224082082802013,
+      "learning_rate": 9.055779569892473e-06,
+      "loss": 1.2541,
+      "step": 2695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.094096668262384,
+      "learning_rate": 9.072580645161291e-06,
+      "loss": 1.2348,
+      "step": 2700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 60.1962371836327,
+      "learning_rate": 9.089381720430108e-06,
+      "loss": 1.2331,
+      "step": 2705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 55.94215746637798,
+      "learning_rate": 9.106182795698926e-06,
+      "loss": 1.2352,
+      "step": 2710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 32.57124388066799,
+      "learning_rate": 9.122983870967743e-06,
+      "loss": 1.2323,
+      "step": 2715
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.941148434081548,
+      "learning_rate": 9.13978494623656e-06,
+      "loss": 1.2388,
+      "step": 2720
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.389045889903844,
+      "learning_rate": 9.156586021505377e-06,
+      "loss": 1.2326,
+      "step": 2725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.3662470293946,
+      "learning_rate": 9.173387096774194e-06,
+      "loss": 1.2977,
+      "step": 2730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.143682290674668,
+      "learning_rate": 9.190188172043013e-06,
+      "loss": 1.2299,
+      "step": 2735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.488199757167923,
+      "learning_rate": 9.206989247311828e-06,
+      "loss": 1.1949,
+      "step": 2740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.543690391067857,
+      "learning_rate": 9.223790322580645e-06,
+      "loss": 1.285,
+      "step": 2745
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 37.70309866093175,
+      "learning_rate": 9.240591397849464e-06,
+      "loss": 1.2336,
+      "step": 2750
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 62.688716746409945,
+      "learning_rate": 9.257392473118281e-06,
+      "loss": 1.2581,
+      "step": 2755
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 37.72106543644531,
+      "learning_rate": 9.274193548387097e-06,
+      "loss": 1.2729,
+      "step": 2760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 56.0751796884353,
+      "learning_rate": 9.290994623655915e-06,
+      "loss": 1.2609,
+      "step": 2765
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 25.60819582600334,
+      "learning_rate": 9.307795698924732e-06,
+      "loss": 1.22,
+      "step": 2770
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.017430243129686,
+      "learning_rate": 9.32459677419355e-06,
+      "loss": 1.2838,
+      "step": 2775
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 21.030906077378877,
+      "learning_rate": 9.341397849462367e-06,
+      "loss": 1.2826,
+      "step": 2780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 28.879404465573838,
+      "learning_rate": 9.358198924731184e-06,
+      "loss": 1.252,
+      "step": 2785
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.44653324996414,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.2714,
+      "step": 2790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.247990145821044,
+      "learning_rate": 9.391801075268818e-06,
+      "loss": 1.2259,
+      "step": 2795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 24.325834288611475,
+      "learning_rate": 9.408602150537635e-06,
+      "loss": 1.2544,
+      "step": 2800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.962373850237869,
+      "learning_rate": 9.425403225806452e-06,
+      "loss": 1.2342,
+      "step": 2805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 20.103796795195294,
+      "learning_rate": 9.44220430107527e-06,
+      "loss": 1.2428,
+      "step": 2810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 65.72278478898748,
+      "learning_rate": 9.459005376344086e-06,
+      "loss": 1.2779,
+      "step": 2815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 34.57590700018222,
+      "learning_rate": 9.475806451612905e-06,
+      "loss": 1.2214,
+      "step": 2820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 29.87845279871248,
+      "learning_rate": 9.49260752688172e-06,
+      "loss": 1.2607,
+      "step": 2825
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.99531248344474,
+      "learning_rate": 9.509408602150538e-06,
+      "loss": 1.215,
+      "step": 2830
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 40.937307170959855,
+      "learning_rate": 9.526209677419356e-06,
+      "loss": 1.2376,
+      "step": 2835
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 24.801802543062614,
+      "learning_rate": 9.543010752688174e-06,
+      "loss": 1.2416,
+      "step": 2840
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.843620697232266,
+      "learning_rate": 9.559811827956989e-06,
+      "loss": 1.2758,
+      "step": 2845
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 25.767959275553633,
+      "learning_rate": 9.576612903225808e-06,
+      "loss": 1.2924,
+      "step": 2850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 24.835210656953024,
+      "learning_rate": 9.593413978494625e-06,
+      "loss": 1.3021,
+      "step": 2855
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.811252228116862,
+      "learning_rate": 9.610215053763442e-06,
+      "loss": 1.1958,
+      "step": 2860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.84580463225306,
+      "learning_rate": 9.627016129032259e-06,
+      "loss": 1.2299,
+      "step": 2865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 35.50525850048963,
+      "learning_rate": 9.643817204301076e-06,
+      "loss": 1.2543,
+      "step": 2870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.821644675146317,
+      "learning_rate": 9.660618279569893e-06,
+      "loss": 1.2512,
+      "step": 2875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 25.16120140943554,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 1.2644,
+      "step": 2880
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 26.443037158380964,
+      "learning_rate": 9.694220430107528e-06,
+      "loss": 1.2873,
+      "step": 2885
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 32.58614186448375,
+      "learning_rate": 9.711021505376345e-06,
+      "loss": 1.2056,
+      "step": 2890
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 28.009823108590062,
+      "learning_rate": 9.727822580645162e-06,
+      "loss": 1.2312,
+      "step": 2895
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 29.593172996117087,
+      "learning_rate": 9.744623655913979e-06,
+      "loss": 1.2385,
+      "step": 2900
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 46.463353284382364,
+      "learning_rate": 9.761424731182798e-06,
+      "loss": 1.2203,
+      "step": 2905
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.522172617728398,
+      "learning_rate": 9.778225806451613e-06,
+      "loss": 1.2851,
+      "step": 2910
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 21.8644130205428,
+      "learning_rate": 9.79502688172043e-06,
+      "loss": 1.3051,
+      "step": 2915
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 60.03291649051756,
+      "learning_rate": 9.811827956989249e-06,
+      "loss": 1.2196,
+      "step": 2920
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 60.594682600646706,
+      "learning_rate": 9.828629032258066e-06,
+      "loss": 1.2187,
+      "step": 2925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.381452691629116,
+      "learning_rate": 9.845430107526881e-06,
+      "loss": 1.2154,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.090385181932243,
+      "learning_rate": 9.8622311827957e-06,
+      "loss": 1.2283,
+      "step": 2935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.85732363275535,
+      "learning_rate": 9.879032258064517e-06,
+      "loss": 1.3106,
+      "step": 2940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 39.278581764228036,
+      "learning_rate": 9.895833333333334e-06,
+      "loss": 1.2291,
+      "step": 2945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 44.68729775499523,
+      "learning_rate": 9.912634408602152e-06,
+      "loss": 1.2338,
+      "step": 2950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 43.45808222006517,
+      "learning_rate": 9.929435483870969e-06,
+      "loss": 1.3069,
+      "step": 2955
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.02806886416131,
+      "learning_rate": 9.946236559139786e-06,
+      "loss": 1.2378,
+      "step": 2960
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 35.73834294646412,
+      "learning_rate": 9.963037634408603e-06,
+      "loss": 1.2714,
+      "step": 2965
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 24.782915964923706,
+      "learning_rate": 9.97983870967742e-06,
+      "loss": 1.2359,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 70.72793777002627,
+      "learning_rate": 9.996639784946237e-06,
+      "loss": 1.3251,
+      "step": 2975
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 57.16942627102183,
+      "learning_rate": 9.999999449441523e-06,
+      "loss": 1.3049,
+      "step": 2980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 94.53740449854857,
+      "learning_rate": 9.999997212797916e-06,
+      "loss": 1.2459,
+      "step": 2985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 111.95977469538862,
+      "learning_rate": 9.999993255660043e-06,
+      "loss": 1.3163,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.031804726431162,
+      "learning_rate": 9.999987578029268e-06,
+      "loss": 1.3024,
+      "step": 2995
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 71.96322221490063,
+      "learning_rate": 9.999980179907541e-06,
+      "loss": 1.299,
+      "step": 3000
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 51.906863459767855,
+      "learning_rate": 9.999971061297411e-06,
+      "loss": 1.3545,
+      "step": 3005
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 41.68798761739742,
+      "learning_rate": 9.999960222202014e-06,
+      "loss": 1.2937,
+      "step": 3010
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 46.196693267941605,
+      "learning_rate": 9.99994766262508e-06,
+      "loss": 1.291,
+      "step": 3015
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 83.24411008845641,
+      "learning_rate": 9.99993338257093e-06,
+      "loss": 1.3184,
+      "step": 3020
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 79.47916536696067,
+      "learning_rate": 9.999917382044479e-06,
+      "loss": 1.3597,
+      "step": 3025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 47.364807117651814,
+      "learning_rate": 9.999899661051232e-06,
+      "loss": 1.2771,
+      "step": 3030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.949929395733275,
+      "learning_rate": 9.999880219597287e-06,
+      "loss": 1.2681,
+      "step": 3035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 36.05513683071444,
+      "learning_rate": 9.999859057689336e-06,
+      "loss": 1.2577,
+      "step": 3040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 49.1723693913278,
+      "learning_rate": 9.999836175334657e-06,
+      "loss": 1.2624,
+      "step": 3045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 19.818863595628407,
+      "learning_rate": 9.999811572541125e-06,
+      "loss": 1.252,
+      "step": 3050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 65.3095406563796,
+      "learning_rate": 9.999785249317207e-06,
+      "loss": 1.2392,
+      "step": 3055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 33.44323503436429,
+      "learning_rate": 9.999757205671959e-06,
+      "loss": 1.2445,
+      "step": 3060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 54.9536579150963,
+      "learning_rate": 9.999727441615032e-06,
+      "loss": 1.3022,
+      "step": 3065
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.625118717845428,
+      "learning_rate": 9.99969595715667e-06,
+      "loss": 1.2172,
+      "step": 3070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 53.04058016138585,
+      "learning_rate": 9.999662752307702e-06,
+      "loss": 1.2891,
+      "step": 3075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 41.0790648049123,
+      "learning_rate": 9.999627827079557e-06,
+      "loss": 1.2606,
+      "step": 3080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.949311683171679,
+      "learning_rate": 9.999591181484251e-06,
+      "loss": 1.2433,
+      "step": 3085
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.606680721081727,
+      "learning_rate": 9.999552815534396e-06,
+      "loss": 1.2767,
+      "step": 3090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 23.548287110125504,
+      "learning_rate": 9.999512729243191e-06,
+      "loss": 1.2329,
+      "step": 3095
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.659213956234366,
+      "learning_rate": 9.999470922624432e-06,
+      "loss": 1.2118,
+      "step": 3100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.109026198379695,
+      "learning_rate": 9.999427395692502e-06,
+      "loss": 1.271,
+      "step": 3105
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 16.205675455931562,
+      "learning_rate": 9.999382148462382e-06,
+      "loss": 1.2228,
+      "step": 3110
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.84615719190485,
+      "learning_rate": 9.99933518094964e-06,
+      "loss": 1.2587,
+      "step": 3115
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.473676379339645,
+      "learning_rate": 9.999286493170435e-06,
+      "loss": 1.2854,
+      "step": 3120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.029530584732344,
+      "learning_rate": 9.999236085141523e-06,
+      "loss": 1.2422,
+      "step": 3125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.574237532934099,
+      "learning_rate": 9.999183956880252e-06,
+      "loss": 1.1645,
+      "step": 3130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.224399893537099,
+      "learning_rate": 9.999130108404553e-06,
+      "loss": 1.2224,
+      "step": 3135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.540594477353714,
+      "learning_rate": 9.99907453973296e-06,
+      "loss": 1.238,
+      "step": 3140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 25.11320923926726,
+      "learning_rate": 9.999017250884591e-06,
+      "loss": 1.2602,
+      "step": 3145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.434342173799088,
+      "learning_rate": 9.998958241879163e-06,
+      "loss": 1.2327,
+      "step": 3150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 23.92510726176057,
+      "learning_rate": 9.998897512736977e-06,
+      "loss": 1.2771,
+      "step": 3155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.27586016590783,
+      "learning_rate": 9.998835063478928e-06,
+      "loss": 1.2186,
+      "step": 3160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.206208499134267,
+      "learning_rate": 9.998770894126513e-06,
+      "loss": 1.244,
+      "step": 3165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.071587368592734,
+      "learning_rate": 9.998705004701805e-06,
+      "loss": 1.2746,
+      "step": 3170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.851985458784839,
+      "learning_rate": 9.998637395227481e-06,
+      "loss": 1.2293,
+      "step": 3175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.125235158488461,
+      "learning_rate": 9.998568065726804e-06,
+      "loss": 1.2395,
+      "step": 3180
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.459385418108235,
+      "learning_rate": 9.998497016223628e-06,
+      "loss": 1.2507,
+      "step": 3185
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.723762738270437,
+      "learning_rate": 9.998424246742403e-06,
+      "loss": 1.2461,
+      "step": 3190
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.560316300456964,
+      "learning_rate": 9.99834975730817e-06,
+      "loss": 1.1793,
+      "step": 3195
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.330824182062663,
+      "learning_rate": 9.998273547946557e-06,
+      "loss": 1.2471,
+      "step": 3200
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.256225012715516,
+      "learning_rate": 9.998195618683793e-06,
+      "loss": 1.2439,
+      "step": 3205
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.292485899881653,
+      "learning_rate": 9.99811596954669e-06,
+      "loss": 1.1854,
+      "step": 3210
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.509099993425679,
+      "learning_rate": 9.998034600562654e-06,
+      "loss": 1.2467,
+      "step": 3215
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.822666642474209,
+      "learning_rate": 9.997951511759686e-06,
+      "loss": 1.321,
+      "step": 3220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.367649827222804,
+      "learning_rate": 9.997866703166376e-06,
+      "loss": 1.2055,
+      "step": 3225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.915246360335449,
+      "learning_rate": 9.997780174811908e-06,
+      "loss": 1.2468,
+      "step": 3230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.3235819296260605,
+      "learning_rate": 9.997691926726055e-06,
+      "loss": 1.2301,
+      "step": 3235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.881549658254508,
+      "learning_rate": 9.997601958939185e-06,
+      "loss": 1.2476,
+      "step": 3240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.774689870813962,
+      "learning_rate": 9.997510271482252e-06,
+      "loss": 1.2587,
+      "step": 3245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 23.119656881210187,
+      "learning_rate": 9.997416864386808e-06,
+      "loss": 1.2944,
+      "step": 3250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 44.43248294796607,
+      "learning_rate": 9.997321737684994e-06,
+      "loss": 1.2861,
+      "step": 3255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 47.999482127159396,
+      "learning_rate": 9.997224891409542e-06,
+      "loss": 1.2578,
+      "step": 3260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 70.59257075033224,
+      "learning_rate": 9.99712632559378e-06,
+      "loss": 1.2717,
+      "step": 3265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 23.71641327422162,
+      "learning_rate": 9.997026040271623e-06,
+      "loss": 1.2947,
+      "step": 3270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 22.5621081905156,
+      "learning_rate": 9.996924035477575e-06,
+      "loss": 1.2861,
+      "step": 3275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.337136579692809,
+      "learning_rate": 9.996820311246741e-06,
+      "loss": 1.2452,
+      "step": 3280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 38.92682377208474,
+      "learning_rate": 9.99671486761481e-06,
+      "loss": 1.227,
+      "step": 3285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 39.12070470753764,
+      "learning_rate": 9.996607704618067e-06,
+      "loss": 1.2175,
+      "step": 3290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 55.06863641189487,
+      "learning_rate": 9.996498822293383e-06,
+      "loss": 1.261,
+      "step": 3295
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 53.5917064046568,
+      "learning_rate": 9.996388220678226e-06,
+      "loss": 1.267,
+      "step": 3300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 22.7001638237988,
+      "learning_rate": 9.996275899810657e-06,
+      "loss": 1.2679,
+      "step": 3305
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 26.240695732587145,
+      "learning_rate": 9.996161859729322e-06,
+      "loss": 1.2276,
+      "step": 3310
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 28.864109624837717,
+      "learning_rate": 9.99604610047346e-06,
+      "loss": 1.2622,
+      "step": 3315
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.181745192592954,
+      "learning_rate": 9.99592862208291e-06,
+      "loss": 1.2353,
+      "step": 3320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 31.135738640073715,
+      "learning_rate": 9.99580942459809e-06,
+      "loss": 1.2259,
+      "step": 3325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.694094175623361,
+      "learning_rate": 9.995688508060022e-06,
+      "loss": 1.2638,
+      "step": 3330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.336896391555797,
+      "learning_rate": 9.995565872510306e-06,
+      "loss": 1.2791,
+      "step": 3335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 67.85777726137141,
+      "learning_rate": 9.995441517991149e-06,
+      "loss": 1.2656,
+      "step": 3340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 49.402079926275604,
+      "learning_rate": 9.995315444545332e-06,
+      "loss": 1.2348,
+      "step": 3345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 104.59190696680739,
+      "learning_rate": 9.995187652216245e-06,
+      "loss": 1.3082,
+      "step": 3350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 48.01718012292942,
+      "learning_rate": 9.995058141047857e-06,
+      "loss": 1.2673,
+      "step": 3355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 33.64789850910811,
+      "learning_rate": 9.994926911084734e-06,
+      "loss": 1.2607,
+      "step": 3360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 31.926457677825184,
+      "learning_rate": 9.994793962372032e-06,
+      "loss": 1.3047,
+      "step": 3365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 36.43634702069511,
+      "learning_rate": 9.994659294955496e-06,
+      "loss": 1.2635,
+      "step": 3370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.716948822292796,
+      "learning_rate": 9.994522908881468e-06,
+      "loss": 1.2718,
+      "step": 3375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.93812477910786,
+      "learning_rate": 9.994384804196877e-06,
+      "loss": 1.2071,
+      "step": 3380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.4307138152681285,
+      "learning_rate": 9.994244980949246e-06,
+      "loss": 1.2283,
+      "step": 3385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 26.716814730179262,
+      "learning_rate": 9.994103439186686e-06,
+      "loss": 1.2311,
+      "step": 3390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.57535376565661,
+      "learning_rate": 9.993960178957904e-06,
+      "loss": 1.2416,
+      "step": 3395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.489410995507601,
+      "learning_rate": 9.993815200312194e-06,
+      "loss": 1.2401,
+      "step": 3400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.963757789326618,
+      "learning_rate": 9.993668503299443e-06,
+      "loss": 1.2223,
+      "step": 3405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.731516475386249,
+      "learning_rate": 9.993520087970128e-06,
+      "loss": 1.2622,
+      "step": 3410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.199217760337424,
+      "learning_rate": 9.993369954375322e-06,
+      "loss": 1.2187,
+      "step": 3415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 20.61144183377057,
+      "learning_rate": 9.993218102566683e-06,
+      "loss": 1.2648,
+      "step": 3420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.007496284136545,
+      "learning_rate": 9.993064532596464e-06,
+      "loss": 1.2809,
+      "step": 3425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.051987396263192,
+      "learning_rate": 9.992909244517507e-06,
+      "loss": 1.2683,
+      "step": 3430
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.8353867234768675,
+      "learning_rate": 9.99275223838325e-06,
+      "loss": 1.2712,
+      "step": 3435
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.769209591212741,
+      "learning_rate": 9.992593514247716e-06,
+      "loss": 1.2244,
+      "step": 3440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.723137752045876,
+      "learning_rate": 9.992433072165521e-06,
+      "loss": 1.2055,
+      "step": 3445
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.580879719381715,
+      "learning_rate": 9.992270912191875e-06,
+      "loss": 1.214,
+      "step": 3450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.348028440545775,
+      "learning_rate": 9.992107034382576e-06,
+      "loss": 1.2512,
+      "step": 3455
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.119279921782843,
+      "learning_rate": 9.991941438794016e-06,
+      "loss": 1.2072,
+      "step": 3460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.524058585005298,
+      "learning_rate": 9.991774125483173e-06,
+      "loss": 1.2263,
+      "step": 3465
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.444283873220677,
+      "learning_rate": 9.991605094507621e-06,
+      "loss": 1.2625,
+      "step": 3470
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.837669847558407,
+      "learning_rate": 9.991434345925525e-06,
+      "loss": 1.2565,
+      "step": 3475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 27.594282860281425,
+      "learning_rate": 9.991261879795637e-06,
+      "loss": 1.2006,
+      "step": 3480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.791241576434466,
+      "learning_rate": 9.991087696177304e-06,
+      "loss": 1.2963,
+      "step": 3485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 30.203839734682212,
+      "learning_rate": 9.990911795130461e-06,
+      "loss": 1.2561,
+      "step": 3490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 98.23200362278807,
+      "learning_rate": 9.990734176715635e-06,
+      "loss": 1.2461,
+      "step": 3495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 27.271840312622874,
+      "learning_rate": 9.990554840993948e-06,
+      "loss": 1.3156,
+      "step": 3500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 61.93216058953879,
+      "learning_rate": 9.990373788027105e-06,
+      "loss": 1.3108,
+      "step": 3505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.847753788646035,
+      "learning_rate": 9.990191017877408e-06,
+      "loss": 1.2783,
+      "step": 3510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.410406057080326,
+      "learning_rate": 9.99000653060775e-06,
+      "loss": 1.2319,
+      "step": 3515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.31825759414051,
+      "learning_rate": 9.989820326281608e-06,
+      "loss": 1.2944,
+      "step": 3520
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 24.686023979520524,
+      "learning_rate": 9.989632404963058e-06,
+      "loss": 1.2311,
+      "step": 3525
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.056066379950447,
+      "learning_rate": 9.989442766716766e-06,
+      "loss": 1.2642,
+      "step": 3530
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.193503192165913,
+      "learning_rate": 9.98925141160798e-06,
+      "loss": 1.229,
+      "step": 3535
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.672413486305313,
+      "learning_rate": 9.98905833970255e-06,
+      "loss": 1.2523,
+      "step": 3540
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.013438018268829,
+      "learning_rate": 9.98886355106691e-06,
+      "loss": 1.2177,
+      "step": 3545
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.739984602942702,
+      "learning_rate": 9.988667045768088e-06,
+      "loss": 1.3048,
+      "step": 3550
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 14.313811608656064,
+      "learning_rate": 9.988468823873701e-06,
+      "loss": 1.2297,
+      "step": 3555
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.080105167095407,
+      "learning_rate": 9.988268885451954e-06,
+      "loss": 1.2716,
+      "step": 3560
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.052177241976224,
+      "learning_rate": 9.988067230571648e-06,
+      "loss": 1.2084,
+      "step": 3565
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 43.26819159375231,
+      "learning_rate": 9.987863859302174e-06,
+      "loss": 1.2811,
+      "step": 3570
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 31.906852111691876,
+      "learning_rate": 9.987658771713508e-06,
+      "loss": 1.2278,
+      "step": 3575
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 32.0926758251754,
+      "learning_rate": 9.987451967876222e-06,
+      "loss": 1.1654,
+      "step": 3580
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.778055781974132,
+      "learning_rate": 9.987243447861479e-06,
+      "loss": 1.2623,
+      "step": 3585
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 26.73272187671943,
+      "learning_rate": 9.987033211741028e-06,
+      "loss": 1.1936,
+      "step": 3590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 23.712272553679398,
+      "learning_rate": 9.986821259587214e-06,
+      "loss": 1.2293,
+      "step": 3595
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 25.72299647520127,
+      "learning_rate": 9.986607591472966e-06,
+      "loss": 1.2311,
+      "step": 3600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.615736190580805,
+      "learning_rate": 9.98639220747181e-06,
+      "loss": 1.3012,
+      "step": 3605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.95268313491177,
+      "learning_rate": 9.986175107657855e-06,
+      "loss": 1.2166,
+      "step": 3610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.109081494313262,
+      "learning_rate": 9.985956292105809e-06,
+      "loss": 1.21,
+      "step": 3615
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.46050026278034,
+      "learning_rate": 9.985735760890966e-06,
+      "loss": 1.2424,
+      "step": 3620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.548313419301463,
+      "learning_rate": 9.985513514089209e-06,
+      "loss": 1.2031,
+      "step": 3625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.367090000488474,
+      "learning_rate": 9.985289551777014e-06,
+      "loss": 1.2292,
+      "step": 3630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.359290128056417,
+      "learning_rate": 9.985063874031444e-06,
+      "loss": 1.2018,
+      "step": 3635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 25.69776468028513,
+      "learning_rate": 9.984836480930157e-06,
+      "loss": 1.246,
+      "step": 3640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 26.738006219526483,
+      "learning_rate": 9.9846073725514e-06,
+      "loss": 1.232,
+      "step": 3645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 27.98023302821429,
+      "learning_rate": 9.984376548974005e-06,
+      "loss": 1.1696,
+      "step": 3650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.861073552576952,
+      "learning_rate": 9.984144010277402e-06,
+      "loss": 1.2144,
+      "step": 3655
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.373906157122731,
+      "learning_rate": 9.983909756541603e-06,
+      "loss": 1.2201,
+      "step": 3660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.88970548441138,
+      "learning_rate": 9.983673787847218e-06,
+      "loss": 1.2754,
+      "step": 3665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.941159431577093,
+      "learning_rate": 9.983436104275443e-06,
+      "loss": 1.1831,
+      "step": 3670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.114563134319768,
+      "learning_rate": 9.983196705908063e-06,
+      "loss": 1.3016,
+      "step": 3675
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.319121830353334,
+      "learning_rate": 9.982955592827456e-06,
+      "loss": 1.1637,
+      "step": 3680
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.652938083479674,
+      "learning_rate": 9.98271276511659e-06,
+      "loss": 1.2296,
+      "step": 3685
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.455205422152094,
+      "learning_rate": 9.982468222859021e-06,
+      "loss": 1.2656,
+      "step": 3690
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.94661713597918,
+      "learning_rate": 9.982221966138895e-06,
+      "loss": 1.2311,
+      "step": 3695
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.960873556702204,
+      "learning_rate": 9.981973995040948e-06,
+      "loss": 1.2198,
+      "step": 3700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 22.09218480259108,
+      "learning_rate": 9.981724309650507e-06,
+      "loss": 1.323,
+      "step": 3705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.0080144446023,
+      "learning_rate": 9.981472910053492e-06,
+      "loss": 1.2343,
+      "step": 3710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 21.145227700843645,
+      "learning_rate": 9.981219796336403e-06,
+      "loss": 1.2497,
+      "step": 3715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 19.79162580341351,
+      "learning_rate": 9.980964968586342e-06,
+      "loss": 1.2426,
+      "step": 3720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.497505379032013,
+      "learning_rate": 9.980708426890993e-06,
+      "loss": 1.1945,
+      "step": 3725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.618953280914997,
+      "learning_rate": 9.98045017133863e-06,
+      "loss": 1.2648,
+      "step": 3730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.349019252093978,
+      "learning_rate": 9.980190202018121e-06,
+      "loss": 1.2245,
+      "step": 3735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.344019064583094,
+      "learning_rate": 9.979928519018919e-06,
+      "loss": 1.2064,
+      "step": 3740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.945729051308529,
+      "learning_rate": 9.97966512243107e-06,
+      "loss": 1.1961,
+      "step": 3745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.531189055193323,
+      "learning_rate": 9.97940001234521e-06,
+      "loss": 1.2092,
+      "step": 3750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 23.213046385730106,
+      "learning_rate": 9.97913318885256e-06,
+      "loss": 1.2197,
+      "step": 3755
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 32.725060083495215,
+      "learning_rate": 9.978864652044936e-06,
+      "loss": 1.2902,
+      "step": 3760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 43.98473480275498,
+      "learning_rate": 9.97859440201474e-06,
+      "loss": 1.2233,
+      "step": 3765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 42.96571294466646,
+      "learning_rate": 9.978322438854966e-06,
+      "loss": 1.2434,
+      "step": 3770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 67.18105533450216,
+      "learning_rate": 9.978048762659195e-06,
+      "loss": 1.2261,
+      "step": 3775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.791734699201156,
+      "learning_rate": 9.977773373521601e-06,
+      "loss": 1.2142,
+      "step": 3780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.897995141075134,
+      "learning_rate": 9.977496271536943e-06,
+      "loss": 1.2249,
+      "step": 3785
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.497407187298952,
+      "learning_rate": 9.977217456800572e-06,
+      "loss": 1.2123,
+      "step": 3790
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.736211949879045,
+      "learning_rate": 9.976936929408427e-06,
+      "loss": 1.2156,
+      "step": 3795
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.303746441739822,
+      "learning_rate": 9.976654689457038e-06,
+      "loss": 1.2863,
+      "step": 3800
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.829497062940318,
+      "learning_rate": 9.976370737043525e-06,
+      "loss": 1.2618,
+      "step": 3805
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 25.263446092793245,
+      "learning_rate": 9.976085072265593e-06,
+      "loss": 1.2526,
+      "step": 3810
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 35.3377923043493,
+      "learning_rate": 9.975797695221542e-06,
+      "loss": 1.2453,
+      "step": 3815
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.464297585102695,
+      "learning_rate": 9.975508606010254e-06,
+      "loss": 1.2531,
+      "step": 3820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 23.692828808280154,
+      "learning_rate": 9.975217804731208e-06,
+      "loss": 1.2235,
+      "step": 3825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.668925213714026,
+      "learning_rate": 9.974925291484468e-06,
+      "loss": 1.2151,
+      "step": 3830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.866655427075113,
+      "learning_rate": 9.974631066370685e-06,
+      "loss": 1.2726,
+      "step": 3835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.841814770895773,
+      "learning_rate": 9.974335129491107e-06,
+      "loss": 1.27,
+      "step": 3840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 23.100541426504538,
+      "learning_rate": 9.974037480947558e-06,
+      "loss": 1.2535,
+      "step": 3845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 64.29530761193791,
+      "learning_rate": 9.973738120842465e-06,
+      "loss": 1.2538,
+      "step": 3850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 41.529956070053096,
+      "learning_rate": 9.973437049278833e-06,
+      "loss": 1.2561,
+      "step": 3855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.71277426315383,
+      "learning_rate": 9.973134266360265e-06,
+      "loss": 1.2695,
+      "step": 3860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.312591268093923,
+      "learning_rate": 9.972829772190944e-06,
+      "loss": 1.236,
+      "step": 3865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.681689247726292,
+      "learning_rate": 9.972523566875649e-06,
+      "loss": 1.2536,
+      "step": 3870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 26.235793994940313,
+      "learning_rate": 9.972215650519743e-06,
+      "loss": 1.3028,
+      "step": 3875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.132056462810064,
+      "learning_rate": 9.97190602322918e-06,
+      "loss": 1.1961,
+      "step": 3880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 32.44650553471024,
+      "learning_rate": 9.971594685110507e-06,
+      "loss": 1.3053,
+      "step": 3885
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.110785173854836,
+      "learning_rate": 9.971281636270847e-06,
+      "loss": 1.2593,
+      "step": 3890
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.444171464337078,
+      "learning_rate": 9.970966876817928e-06,
+      "loss": 1.2561,
+      "step": 3895
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.17838705462199,
+      "learning_rate": 9.970650406860051e-06,
+      "loss": 1.2029,
+      "step": 3900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.115984906641157,
+      "learning_rate": 9.970332226506118e-06,
+      "loss": 1.2493,
+      "step": 3905
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.70047927862016,
+      "learning_rate": 9.970012335865612e-06,
+      "loss": 1.1869,
+      "step": 3910
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 18.038800651626424,
+      "learning_rate": 9.96969073504861e-06,
+      "loss": 1.2424,
+      "step": 3915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.15861746144962,
+      "learning_rate": 9.96936742416577e-06,
+      "loss": 1.1838,
+      "step": 3920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.484113761097328,
+      "learning_rate": 9.969042403328348e-06,
+      "loss": 1.2171,
+      "step": 3925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.110356553423042,
+      "learning_rate": 9.96871567264818e-06,
+      "loss": 1.2518,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.73363097364326,
+      "learning_rate": 9.968387232237695e-06,
+      "loss": 1.2789,
+      "step": 3935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 27.686523594105953,
+      "learning_rate": 9.968057082209909e-06,
+      "loss": 1.244,
+      "step": 3940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.820108672992458,
+      "learning_rate": 9.967725222678426e-06,
+      "loss": 1.2913,
+      "step": 3945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.67650247073177,
+      "learning_rate": 9.967391653757438e-06,
+      "loss": 1.2102,
+      "step": 3950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 18.73171602233947,
+      "learning_rate": 9.967056375561726e-06,
+      "loss": 1.1982,
+      "step": 3955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 24.675528714132533,
+      "learning_rate": 9.966719388206661e-06,
+      "loss": 1.2475,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 45.25930029609521,
+      "learning_rate": 9.966380691808197e-06,
+      "loss": 1.2594,
+      "step": 3965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 87.67463131497611,
+      "learning_rate": 9.96604028648288e-06,
+      "loss": 1.2446,
+      "step": 3970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.474598522395306,
+      "learning_rate": 9.965698172347843e-06,
+      "loss": 1.2846,
+      "step": 3975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 56.77411416314017,
+      "learning_rate": 9.96535434952081e-06,
+      "loss": 1.252,
+      "step": 3980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 46.8997659190866,
+      "learning_rate": 9.965008818120088e-06,
+      "loss": 1.2369,
+      "step": 3985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 27.77528218311658,
+      "learning_rate": 9.964661578264571e-06,
+      "loss": 1.2456,
+      "step": 3990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 18.546895946518553,
+      "learning_rate": 9.964312630073749e-06,
+      "loss": 1.2298,
+      "step": 3995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.279947297055188,
+      "learning_rate": 9.963961973667691e-06,
+      "loss": 1.1957,
+      "step": 4000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 51.66062157374361,
+      "learning_rate": 9.96360960916706e-06,
+      "loss": 1.2408,
+      "step": 4005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 38.003334883771586,
+      "learning_rate": 9.963255536693103e-06,
+      "loss": 1.2566,
+      "step": 4010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 36.49216818644186,
+      "learning_rate": 9.962899756367657e-06,
+      "loss": 1.2388,
+      "step": 4015
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 37.691359368487895,
+      "learning_rate": 9.962542268313144e-06,
+      "loss": 1.2522,
+      "step": 4020
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.338664665110976,
+      "learning_rate": 9.962183072652577e-06,
+      "loss": 1.2367,
+      "step": 4025
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.087845748300886,
+      "learning_rate": 9.961822169509552e-06,
+      "loss": 1.2158,
+      "step": 4030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 38.570589682996335,
+      "learning_rate": 9.96145955900826e-06,
+      "loss": 1.2555,
+      "step": 4035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 27.693672064187652,
+      "learning_rate": 9.96109524127347e-06,
+      "loss": 1.206,
+      "step": 4040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.133340119549086,
+      "learning_rate": 9.960729216430549e-06,
+      "loss": 1.269,
+      "step": 4045
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 47.127690639583996,
+      "learning_rate": 9.96036148460544e-06,
+      "loss": 1.255,
+      "step": 4050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.301073562064612,
+      "learning_rate": 9.959992045924682e-06,
+      "loss": 1.2739,
+      "step": 4055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.609836658516967,
+      "learning_rate": 9.9596209005154e-06,
+      "loss": 1.3207,
+      "step": 4060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 33.995550626776534,
+      "learning_rate": 9.959248048505304e-06,
+      "loss": 1.2085,
+      "step": 4065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 59.88588971982497,
+      "learning_rate": 9.958873490022688e-06,
+      "loss": 1.2102,
+      "step": 4070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 30.261967435183816,
+      "learning_rate": 9.958497225196441e-06,
+      "loss": 1.2009,
+      "step": 4075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.236600833942322,
+      "learning_rate": 9.958119254156036e-06,
+      "loss": 1.2641,
+      "step": 4080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.097576830169164,
+      "learning_rate": 9.95773957703153e-06,
+      "loss": 1.2568,
+      "step": 4085
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 28.322236699356623,
+      "learning_rate": 9.957358193953573e-06,
+      "loss": 1.2436,
+      "step": 4090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.21633921097602,
+      "learning_rate": 9.956975105053395e-06,
+      "loss": 1.2931,
+      "step": 4095
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.400387277003981,
+      "learning_rate": 9.956590310462817e-06,
+      "loss": 1.2489,
+      "step": 4100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.919282255738858,
+      "learning_rate": 9.956203810314248e-06,
+      "loss": 1.2455,
+      "step": 4105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.852815217381016,
+      "learning_rate": 9.955815604740682e-06,
+      "loss": 1.2237,
+      "step": 4110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.280350390136551,
+      "learning_rate": 9.955425693875699e-06,
+      "loss": 1.2531,
+      "step": 4115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.206007829007667,
+      "learning_rate": 9.955034077853466e-06,
+      "loss": 1.1968,
+      "step": 4120
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.565179622370264,
+      "learning_rate": 9.954640756808743e-06,
+      "loss": 1.2529,
+      "step": 4125
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.6317617587089,
+      "learning_rate": 9.954245730876866e-06,
+      "loss": 1.2129,
+      "step": 4130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.47417695536017,
+      "learning_rate": 9.953849000193764e-06,
+      "loss": 1.233,
+      "step": 4135
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.358928325643685,
+      "learning_rate": 9.953450564895955e-06,
+      "loss": 1.2425,
+      "step": 4140
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 38.85271406815992,
+      "learning_rate": 9.953050425120537e-06,
+      "loss": 1.2653,
+      "step": 4145
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 26.5549459705775,
+      "learning_rate": 9.952648581005196e-06,
+      "loss": 1.2167,
+      "step": 4150
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.9806651207861155,
+      "learning_rate": 9.95224503268821e-06,
+      "loss": 1.2159,
+      "step": 4155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.204088379198158,
+      "learning_rate": 9.951839780308439e-06,
+      "loss": 1.1693,
+      "step": 4160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 34.843347127473734,
+      "learning_rate": 9.951432824005328e-06,
+      "loss": 1.2514,
+      "step": 4165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.431225825049113,
+      "learning_rate": 9.951024163918913e-06,
+      "loss": 1.2578,
+      "step": 4170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.273926092048018,
+      "learning_rate": 9.95061380018981e-06,
+      "loss": 1.1902,
+      "step": 4175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.601147365831144,
+      "learning_rate": 9.950201732959228e-06,
+      "loss": 1.2412,
+      "step": 4180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 39.646460393382455,
+      "learning_rate": 9.949787962368957e-06,
+      "loss": 1.2466,
+      "step": 4185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.64410488897537,
+      "learning_rate": 9.949372488561377e-06,
+      "loss": 1.1957,
+      "step": 4190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.449811055571177,
+      "learning_rate": 9.94895531167945e-06,
+      "loss": 1.228,
+      "step": 4195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.967857762143987,
+      "learning_rate": 9.948536431866726e-06,
+      "loss": 1.2249,
+      "step": 4200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.977410140447304,
+      "learning_rate": 9.948115849267344e-06,
+      "loss": 1.2819,
+      "step": 4205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.49390147390754,
+      "learning_rate": 9.947693564026025e-06,
+      "loss": 1.252,
+      "step": 4210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.579440419526714,
+      "learning_rate": 9.947269576288074e-06,
+      "loss": 1.2551,
+      "step": 4215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.073349590631008,
+      "learning_rate": 9.946843886199387e-06,
+      "loss": 1.2799,
+      "step": 4220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 33.080214220661595,
+      "learning_rate": 9.946416493906445e-06,
+      "loss": 1.219,
+      "step": 4225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.340182315915886,
+      "learning_rate": 9.94598739955631e-06,
+      "loss": 1.2418,
+      "step": 4230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.732259768036213,
+      "learning_rate": 9.945556603296636e-06,
+      "loss": 1.2616,
+      "step": 4235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.220849733614314,
+      "learning_rate": 9.945124105275658e-06,
+      "loss": 1.2264,
+      "step": 4240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.415288492518843,
+      "learning_rate": 9.9446899056422e-06,
+      "loss": 1.2292,
+      "step": 4245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.51497427087494,
+      "learning_rate": 9.944254004545666e-06,
+      "loss": 1.2158,
+      "step": 4250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 45.84960290897499,
+      "learning_rate": 9.943816402136053e-06,
+      "loss": 1.185,
+      "step": 4255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 51.5481831854867,
+      "learning_rate": 9.943377098563936e-06,
+      "loss": 1.2144,
+      "step": 4260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 26.0885083725824,
+      "learning_rate": 9.942936093980482e-06,
+      "loss": 1.2406,
+      "step": 4265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 52.86487745850491,
+      "learning_rate": 9.94249338853744e-06,
+      "loss": 1.2343,
+      "step": 4270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 32.085728312012584,
+      "learning_rate": 9.942048982387142e-06,
+      "loss": 1.2485,
+      "step": 4275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.147507968619474,
+      "learning_rate": 9.94160287568251e-06,
+      "loss": 1.2447,
+      "step": 4280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.746530392136881,
+      "learning_rate": 9.941155068577049e-06,
+      "loss": 1.2367,
+      "step": 4285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 38.852565860595654,
+      "learning_rate": 9.940705561224847e-06,
+      "loss": 1.2278,
+      "step": 4290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 51.92516409644444,
+      "learning_rate": 9.940254353780581e-06,
+      "loss": 1.2651,
+      "step": 4295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 43.46149206212037,
+      "learning_rate": 9.939801446399511e-06,
+      "loss": 1.2728,
+      "step": 4300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 26.44387192644677,
+      "learning_rate": 9.93934683923748e-06,
+      "loss": 1.1896,
+      "step": 4305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 37.746036203691254,
+      "learning_rate": 9.93889053245092e-06,
+      "loss": 1.2588,
+      "step": 4310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.520173237581492,
+      "learning_rate": 9.938432526196844e-06,
+      "loss": 1.1869,
+      "step": 4315
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.467561168011299,
+      "learning_rate": 9.937972820632854e-06,
+      "loss": 1.2904,
+      "step": 4320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.609641849366643,
+      "learning_rate": 9.937511415917132e-06,
+      "loss": 1.2455,
+      "step": 4325
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.00022249781794,
+      "learning_rate": 9.937048312208448e-06,
+      "loss": 1.2635,
+      "step": 4330
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.548844065392103,
+      "learning_rate": 9.936583509666154e-06,
+      "loss": 1.2521,
+      "step": 4335
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.740443268777065,
+      "learning_rate": 9.93611700845019e-06,
+      "loss": 1.214,
+      "step": 4340
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.354451049628311,
+      "learning_rate": 9.93564880872108e-06,
+      "loss": 1.2474,
+      "step": 4345
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.354353870738024,
+      "learning_rate": 9.935178910639927e-06,
+      "loss": 1.2202,
+      "step": 4350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.461666361813926,
+      "learning_rate": 9.934707314368424e-06,
+      "loss": 1.1891,
+      "step": 4355
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 26.26375051175625,
+      "learning_rate": 9.934234020068847e-06,
+      "loss": 1.2721,
+      "step": 4360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.66123236538432,
+      "learning_rate": 9.933759027904058e-06,
+      "loss": 1.1961,
+      "step": 4365
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.849872593058937,
+      "learning_rate": 9.9332823380375e-06,
+      "loss": 1.191,
+      "step": 4370
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.52585365597777,
+      "learning_rate": 9.932803950633199e-06,
+      "loss": 1.2106,
+      "step": 4375
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 23.505146587479103,
+      "learning_rate": 9.93232386585577e-06,
+      "loss": 1.2883,
+      "step": 4380
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.704178651242195,
+      "learning_rate": 9.931842083870413e-06,
+      "loss": 1.2569,
+      "step": 4385
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.662070835592704,
+      "learning_rate": 9.931358604842902e-06,
+      "loss": 1.2289,
+      "step": 4390
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.059661247937026,
+      "learning_rate": 9.930873428939607e-06,
+      "loss": 1.2457,
+      "step": 4395
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.51351005431564,
+      "learning_rate": 9.930386556327475e-06,
+      "loss": 1.2639,
+      "step": 4400
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 25.232301350217114,
+      "learning_rate": 9.929897987174038e-06,
+      "loss": 1.1957,
+      "step": 4405
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.350420590665204,
+      "learning_rate": 9.92940772164741e-06,
+      "loss": 1.2223,
+      "step": 4410
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.311765119921033,
+      "learning_rate": 9.928915759916295e-06,
+      "loss": 1.2363,
+      "step": 4415
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.127480756243877,
+      "learning_rate": 9.928422102149974e-06,
+      "loss": 1.2429,
+      "step": 4420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.699872791778428,
+      "learning_rate": 9.927926748518316e-06,
+      "loss": 1.2488,
+      "step": 4425
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.510205983564731,
+      "learning_rate": 9.927429699191768e-06,
+      "loss": 1.2004,
+      "step": 4430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 31.498092704192338,
+      "learning_rate": 9.926930954341368e-06,
+      "loss": 1.2013,
+      "step": 4435
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 23.41388595460819,
+      "learning_rate": 9.926430514138734e-06,
+      "loss": 1.2054,
+      "step": 4440
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 33.179614627446334,
+      "learning_rate": 9.925928378756064e-06,
+      "loss": 1.2418,
+      "step": 4445
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.854169811315563,
+      "learning_rate": 9.925424548366142e-06,
+      "loss": 1.2694,
+      "step": 4450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 22.130365073256336,
+      "learning_rate": 9.92491902314234e-06,
+      "loss": 1.2199,
+      "step": 4455
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 26.12787060881701,
+      "learning_rate": 9.924411803258604e-06,
+      "loss": 1.2743,
+      "step": 4460
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 44.41777719215755,
+      "learning_rate": 9.923902888889472e-06,
+      "loss": 1.2235,
+      "step": 4465
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 68.4105816881735,
+      "learning_rate": 9.923392280210056e-06,
+      "loss": 1.2475,
+      "step": 4470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 65.17654692737699,
+      "learning_rate": 9.922879977396062e-06,
+      "loss": 1.2378,
+      "step": 4475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.007706083704576,
+      "learning_rate": 9.922365980623768e-06,
+      "loss": 1.3025,
+      "step": 4480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 33.37928601370577,
+      "learning_rate": 9.92185029007004e-06,
+      "loss": 1.2351,
+      "step": 4485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 49.204195129373645,
+      "learning_rate": 9.92133290591233e-06,
+      "loss": 1.2301,
+      "step": 4490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 20.865406533566258,
+      "learning_rate": 9.920813828328668e-06,
+      "loss": 1.2425,
+      "step": 4495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.484545945850263,
+      "learning_rate": 9.920293057497665e-06,
+      "loss": 1.2036,
+      "step": 4500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.112272089623518,
+      "learning_rate": 9.919770593598523e-06,
+      "loss": 1.1751,
+      "step": 4505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.091124986461764,
+      "learning_rate": 9.919246436811017e-06,
+      "loss": 1.1866,
+      "step": 4510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.14013552151064,
+      "learning_rate": 9.918720587315512e-06,
+      "loss": 1.2246,
+      "step": 4515
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.0306005576408,
+      "learning_rate": 9.918193045292949e-06,
+      "loss": 1.2514,
+      "step": 4520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.245379854317187,
+      "learning_rate": 9.917663810924858e-06,
+      "loss": 1.1635,
+      "step": 4525
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.92733969341995,
+      "learning_rate": 9.917132884393346e-06,
+      "loss": 1.2316,
+      "step": 4530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.617987539724332,
+      "learning_rate": 9.916600265881104e-06,
+      "loss": 1.2483,
+      "step": 4535
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.6733268929963545,
+      "learning_rate": 9.916065955571408e-06,
+      "loss": 1.2179,
+      "step": 4540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.242155444345359,
+      "learning_rate": 9.915529953648111e-06,
+      "loss": 1.2364,
+      "step": 4545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.546232213228472,
+      "learning_rate": 9.914992260295653e-06,
+      "loss": 1.2006,
+      "step": 4550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.412672944813984,
+      "learning_rate": 9.914452875699053e-06,
+      "loss": 1.2184,
+      "step": 4555
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.453619543099387,
+      "learning_rate": 9.91391180004391e-06,
+      "loss": 1.2225,
+      "step": 4560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.707388871143154,
+      "learning_rate": 9.913369033516412e-06,
+      "loss": 1.2034,
+      "step": 4565
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.46315709092553,
+      "learning_rate": 9.912824576303321e-06,
+      "loss": 1.2387,
+      "step": 4570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.270494960714121,
+      "learning_rate": 9.912278428591986e-06,
+      "loss": 1.2646,
+      "step": 4575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.417488967738182,
+      "learning_rate": 9.911730590570335e-06,
+      "loss": 1.2333,
+      "step": 4580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 28.781749944510707,
+      "learning_rate": 9.91118106242688e-06,
+      "loss": 1.2128,
+      "step": 4585
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 33.17270676630123,
+      "learning_rate": 9.910629844350711e-06,
+      "loss": 1.2591,
+      "step": 4590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 27.21735457760835,
+      "learning_rate": 9.910076936531503e-06,
+      "loss": 1.1947,
+      "step": 4595
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 44.91419841319632,
+      "learning_rate": 9.909522339159511e-06,
+      "loss": 1.2802,
+      "step": 4600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.381704300692164,
+      "learning_rate": 9.908966052425573e-06,
+      "loss": 1.2412,
+      "step": 4605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 133.4976979325375,
+      "learning_rate": 9.908408076521104e-06,
+      "loss": 1.2974,
+      "step": 4610
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 104.40804477999271,
+      "learning_rate": 9.907848411638102e-06,
+      "loss": 1.2562,
+      "step": 4615
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 43.1718486242428,
+      "learning_rate": 9.90728705796915e-06,
+      "loss": 1.2858,
+      "step": 4620
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.430405281503546,
+      "learning_rate": 9.90672401570741e-06,
+      "loss": 1.2349,
+      "step": 4625
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 28.408859457680723,
+      "learning_rate": 9.906159285046622e-06,
+      "loss": 1.2663,
+      "step": 4630
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 35.80208086924776,
+      "learning_rate": 9.905592866181108e-06,
+      "loss": 1.2238,
+      "step": 4635
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 55.64747717171744,
+      "learning_rate": 9.905024759305777e-06,
+      "loss": 1.2932,
+      "step": 4640
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 25.653632126635838,
+      "learning_rate": 9.904454964616108e-06,
+      "loss": 1.2248,
+      "step": 4645
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 89.7506396373891,
+      "learning_rate": 9.903883482308172e-06,
+      "loss": 1.2288,
+      "step": 4650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 84.92599413538284,
+      "learning_rate": 9.903310312578613e-06,
+      "loss": 1.2751,
+      "step": 4655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 40.33396768812129,
+      "learning_rate": 9.902735455624658e-06,
+      "loss": 1.2513,
+      "step": 4660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 44.693246410015014,
+      "learning_rate": 9.902158911644117e-06,
+      "loss": 1.222,
+      "step": 4665
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 36.40728587256452,
+      "learning_rate": 9.901580680835377e-06,
+      "loss": 1.247,
+      "step": 4670
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.767226773863197,
+      "learning_rate": 9.901000763397404e-06,
+      "loss": 1.2207,
+      "step": 4675
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 33.84826992392133,
+      "learning_rate": 9.900419159529751e-06,
+      "loss": 1.2387,
+      "step": 4680
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.084937111619066,
+      "learning_rate": 9.899835869432544e-06,
+      "loss": 1.2599,
+      "step": 4685
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.903378794691731,
+      "learning_rate": 9.899250893306496e-06,
+      "loss": 1.2094,
+      "step": 4690
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.651747205568551,
+      "learning_rate": 9.898664231352894e-06,
+      "loss": 1.278,
+      "step": 4695
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 18.91091257408942,
+      "learning_rate": 9.898075883773609e-06,
+      "loss": 1.2486,
+      "step": 4700
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.666073931133331,
+      "learning_rate": 9.897485850771092e-06,
+      "loss": 1.2789,
+      "step": 4705
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.233336917533183,
+      "learning_rate": 9.89689413254837e-06,
+      "loss": 1.2588,
+      "step": 4710
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 48.842048881189655,
+      "learning_rate": 9.896300729309051e-06,
+      "loss": 1.1803,
+      "step": 4715
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 48.32400165414345,
+      "learning_rate": 9.895705641257332e-06,
+      "loss": 1.2072,
+      "step": 4720
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 15.576826950538285,
+      "learning_rate": 9.895108868597976e-06,
+      "loss": 1.2672,
+      "step": 4725
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 57.056727168245324,
+      "learning_rate": 9.894510411536335e-06,
+      "loss": 1.2345,
+      "step": 4730
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 45.82864349568885,
+      "learning_rate": 9.893910270278335e-06,
+      "loss": 1.2594,
+      "step": 4735
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 53.472664284359006,
+      "learning_rate": 9.893308445030485e-06,
+      "loss": 1.2658,
+      "step": 4740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 51.09707449624854,
+      "learning_rate": 9.892704935999872e-06,
+      "loss": 1.2381,
+      "step": 4745
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 52.06497551634407,
+      "learning_rate": 9.892099743394165e-06,
+      "loss": 1.2305,
+      "step": 4750
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.917155144998816,
+      "learning_rate": 9.89149286742161e-06,
+      "loss": 1.248,
+      "step": 4755
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 33.93858162026787,
+      "learning_rate": 9.89088430829103e-06,
+      "loss": 1.1996,
+      "step": 4760
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 24.056306625196044,
+      "learning_rate": 9.890274066211828e-06,
+      "loss": 1.2323,
+      "step": 4765
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 36.916877403006154,
+      "learning_rate": 9.889662141393994e-06,
+      "loss": 1.2677,
+      "step": 4770
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 49.26865152084359,
+      "learning_rate": 9.889048534048088e-06,
+      "loss": 1.2934,
+      "step": 4775
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 36.128655809311724,
+      "learning_rate": 9.888433244385248e-06,
+      "loss": 1.2324,
+      "step": 4780
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 33.15784894308101,
+      "learning_rate": 9.8878162726172e-06,
+      "loss": 1.2033,
+      "step": 4785
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.0821031794577,
+      "learning_rate": 9.887197618956239e-06,
+      "loss": 1.2938,
+      "step": 4790
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.63229833800865,
+      "learning_rate": 9.886577283615247e-06,
+      "loss": 1.2084,
+      "step": 4795
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 34.363919504880776,
+      "learning_rate": 9.885955266807677e-06,
+      "loss": 1.2019,
+      "step": 4800
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 23.32077401063675,
+      "learning_rate": 9.885331568747569e-06,
+      "loss": 1.2544,
+      "step": 4805
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 27.802670446738542,
+      "learning_rate": 9.884706189649532e-06,
+      "loss": 1.1967,
+      "step": 4810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.71809098143742,
+      "learning_rate": 9.88407912972876e-06,
+      "loss": 1.2215,
+      "step": 4815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 14.621038663696202,
+      "learning_rate": 9.883450389201025e-06,
+      "loss": 1.2621,
+      "step": 4820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 34.66887135697982,
+      "learning_rate": 9.882819968282678e-06,
+      "loss": 1.2593,
+      "step": 4825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.02489044655485,
+      "learning_rate": 9.88218786719064e-06,
+      "loss": 1.2477,
+      "step": 4830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.21532524157301,
+      "learning_rate": 9.881554086142422e-06,
+      "loss": 1.278,
+      "step": 4835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 39.71229395992522,
+      "learning_rate": 9.880918625356103e-06,
+      "loss": 1.2432,
+      "step": 4840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 38.8748764564693,
+      "learning_rate": 9.88028148505035e-06,
+      "loss": 1.24,
+      "step": 4845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.839325112997184,
+      "learning_rate": 9.879642665444397e-06,
+      "loss": 1.2117,
+      "step": 4850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 79.07900220844354,
+      "learning_rate": 9.879002166758065e-06,
+      "loss": 1.2215,
+      "step": 4855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 23.67166694685957,
+      "learning_rate": 9.878359989211747e-06,
+      "loss": 1.2823,
+      "step": 4860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.273654423274305,
+      "learning_rate": 9.877716133026415e-06,
+      "loss": 1.3155,
+      "step": 4865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.417875140663172,
+      "learning_rate": 9.87707059842362e-06,
+      "loss": 1.187,
+      "step": 4870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.901533696869453,
+      "learning_rate": 9.876423385625492e-06,
+      "loss": 1.2881,
+      "step": 4875
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.667402310523359,
+      "learning_rate": 9.875774494854734e-06,
+      "loss": 1.263,
+      "step": 4880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.67718694052718,
+      "learning_rate": 9.875123926334629e-06,
+      "loss": 1.2246,
+      "step": 4885
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.615445503902517,
+      "learning_rate": 9.874471680289037e-06,
+      "loss": 1.2496,
+      "step": 4890
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.202333722795672,
+      "learning_rate": 9.873817756942396e-06,
+      "loss": 1.2405,
+      "step": 4895
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.941122551177786,
+      "learning_rate": 9.873162156519718e-06,
+      "loss": 1.2547,
+      "step": 4900
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.333362690140278,
+      "learning_rate": 9.872504879246598e-06,
+      "loss": 1.257,
+      "step": 4905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 22.836006708807865,
+      "learning_rate": 9.871845925349204e-06,
+      "loss": 1.2134,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 29.536939423968438,
+      "learning_rate": 9.871185295054278e-06,
+      "loss": 1.217,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 25.05820903087772,
+      "learning_rate": 9.870522988589146e-06,
+      "loss": 1.2439,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.98113010207636,
+      "learning_rate": 9.869859006181705e-06,
+      "loss": 1.1887,
+      "step": 4925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.53745644035191,
+      "learning_rate": 9.86919334806043e-06,
+      "loss": 1.233,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.311372888712697,
+      "learning_rate": 9.868526014454377e-06,
+      "loss": 1.2674,
+      "step": 4935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.363781317584587,
+      "learning_rate": 9.867857005593172e-06,
+      "loss": 1.2111,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.179319863792337,
+      "learning_rate": 9.86718632170702e-06,
+      "loss": 1.2123,
+      "step": 4945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.505244786041763,
+      "learning_rate": 9.866513963026703e-06,
+      "loss": 1.2404,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.854134017392649,
+      "learning_rate": 9.865839929783578e-06,
+      "loss": 1.2601,
+      "step": 4955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.191892520579293,
+      "learning_rate": 9.865164222209583e-06,
+      "loss": 1.2635,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.751273585672897,
+      "learning_rate": 9.864486840537225e-06,
+      "loss": 1.1951,
+      "step": 4965
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 44.06145173052812,
+      "learning_rate": 9.863807784999591e-06,
+      "loss": 1.2202,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.736756270544875,
+      "learning_rate": 9.863127055830343e-06,
+      "loss": 1.2326,
+      "step": 4975
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.145462317727578,
+      "learning_rate": 9.862444653263723e-06,
+      "loss": 1.2715,
+      "step": 4980
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.699736433964649,
+      "learning_rate": 9.861760577534538e-06,
+      "loss": 1.2768,
+      "step": 4985
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.020798950138495,
+      "learning_rate": 9.861074828878184e-06,
+      "loss": 1.254,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.66731748752754,
+      "learning_rate": 9.860387407530625e-06,
+      "loss": 1.2259,
+      "step": 4995
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.659673761001347,
+      "learning_rate": 9.859698313728399e-06,
+      "loss": 1.2653,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.695606595819575,
+      "learning_rate": 9.859007547708626e-06,
+      "loss": 1.2337,
+      "step": 5005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.328135911399114,
+      "learning_rate": 9.858315109708998e-06,
+      "loss": 1.2404,
+      "step": 5010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.11731390695412,
+      "learning_rate": 9.857620999967778e-06,
+      "loss": 1.2627,
+      "step": 5015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 20.37594952930069,
+      "learning_rate": 9.856925218723814e-06,
+      "loss": 1.2325,
+      "step": 5020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 23.392796017990797,
+      "learning_rate": 9.85622776621652e-06,
+      "loss": 1.2049,
+      "step": 5025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 46.26122007063707,
+      "learning_rate": 9.85552864268589e-06,
+      "loss": 1.2259,
+      "step": 5030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.075205946208943,
+      "learning_rate": 9.854827848372492e-06,
+      "loss": 1.2915,
+      "step": 5035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.041877762390996,
+      "learning_rate": 9.854125383517468e-06,
+      "loss": 1.2339,
+      "step": 5040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 18.90087269271391,
+      "learning_rate": 9.853421248362536e-06,
+      "loss": 1.2436,
+      "step": 5045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.086319639257932,
+      "learning_rate": 9.852715443149988e-06,
+      "loss": 1.2494,
+      "step": 5050
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.108264527082493,
+      "learning_rate": 9.852007968122691e-06,
+      "loss": 1.213,
+      "step": 5055
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.833632396667184,
+      "learning_rate": 9.851298823524086e-06,
+      "loss": 1.2191,
+      "step": 5060
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.616780902137627,
+      "learning_rate": 9.85058800959819e-06,
+      "loss": 1.2347,
+      "step": 5065
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 53.27961975081988,
+      "learning_rate": 9.849875526589592e-06,
+      "loss": 1.2327,
+      "step": 5070
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.000337199179235,
+      "learning_rate": 9.849161374743456e-06,
+      "loss": 1.2462,
+      "step": 5075
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.697995505646016,
+      "learning_rate": 9.848445554305525e-06,
+      "loss": 1.255,
+      "step": 5080
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.537173940270334,
+      "learning_rate": 9.847728065522109e-06,
+      "loss": 1.2691,
+      "step": 5085
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.881764103267859,
+      "learning_rate": 9.847008908640096e-06,
+      "loss": 1.2048,
+      "step": 5090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.444448898642385,
+      "learning_rate": 9.846288083906945e-06,
+      "loss": 1.2537,
+      "step": 5095
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 23.459583080559593,
+      "learning_rate": 9.845565591570696e-06,
+      "loss": 1.2382,
+      "step": 5100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.53518476827716,
+      "learning_rate": 9.844841431879953e-06,
+      "loss": 1.2616,
+      "step": 5105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.742081524029113,
+      "learning_rate": 9.8441156050839e-06,
+      "loss": 1.2426,
+      "step": 5110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.190738080721399,
+      "learning_rate": 9.843388111432295e-06,
+      "loss": 1.2427,
+      "step": 5115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.59878231203815,
+      "learning_rate": 9.842658951175468e-06,
+      "loss": 1.2566,
+      "step": 5120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.645578107701345,
+      "learning_rate": 9.84192812456432e-06,
+      "loss": 1.1824,
+      "step": 5125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.559822039639393,
+      "learning_rate": 9.841195631850329e-06,
+      "loss": 1.2168,
+      "step": 5130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.02789966316085,
+      "learning_rate": 9.840461473285545e-06,
+      "loss": 1.2199,
+      "step": 5135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.593998708996688,
+      "learning_rate": 9.83972564912259e-06,
+      "loss": 1.2275,
+      "step": 5140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.418060402702537,
+      "learning_rate": 9.838988159614662e-06,
+      "loss": 1.2532,
+      "step": 5145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.902976645073543,
+      "learning_rate": 9.838249005015532e-06,
+      "loss": 1.2601,
+      "step": 5150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.003923073859536,
+      "learning_rate": 9.837508185579539e-06,
+      "loss": 1.2247,
+      "step": 5155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 27.50304902900661,
+      "learning_rate": 9.8367657015616e-06,
+      "loss": 1.318,
+      "step": 5160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.691673740264005,
+      "learning_rate": 9.836021553217204e-06,
+      "loss": 1.2053,
+      "step": 5165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 39.807525373760335,
+      "learning_rate": 9.835275740802407e-06,
+      "loss": 1.2445,
+      "step": 5170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 54.93907310262677,
+      "learning_rate": 9.834528264573848e-06,
+      "loss": 1.2496,
+      "step": 5175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 29.959308679991757,
+      "learning_rate": 9.833779124788732e-06,
+      "loss": 1.213,
+      "step": 5180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 30.341779174554887,
+      "learning_rate": 9.833028321704833e-06,
+      "loss": 1.241,
+      "step": 5185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.11403595375024,
+      "learning_rate": 9.832275855580506e-06,
+      "loss": 1.2206,
+      "step": 5190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.211458334912457,
+      "learning_rate": 9.831521726674673e-06,
+      "loss": 1.2358,
+      "step": 5195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 55.873465004227356,
+      "learning_rate": 9.83076593524683e-06,
+      "loss": 1.2454,
+      "step": 5200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.95501751361242,
+      "learning_rate": 9.830008481557039e-06,
+      "loss": 1.265,
+      "step": 5205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 26.831116982855633,
+      "learning_rate": 9.829249365865944e-06,
+      "loss": 1.1869,
+      "step": 5210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.272876439024357,
+      "learning_rate": 9.828488588434755e-06,
+      "loss": 1.2621,
+      "step": 5215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.744067112409168,
+      "learning_rate": 9.827726149525254e-06,
+      "loss": 1.2149,
+      "step": 5220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.587763110752242,
+      "learning_rate": 9.826962049399797e-06,
+      "loss": 1.1707,
+      "step": 5225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 46.80257048909383,
+      "learning_rate": 9.826196288321308e-06,
+      "loss": 1.2159,
+      "step": 5230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 30.462258854027745,
+      "learning_rate": 9.825428866553286e-06,
+      "loss": 1.2172,
+      "step": 5235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.872961826576143,
+      "learning_rate": 9.824659784359801e-06,
+      "loss": 1.2454,
+      "step": 5240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.107407756464607,
+      "learning_rate": 9.823889042005491e-06,
+      "loss": 1.2975,
+      "step": 5245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 32.772802733577855,
+      "learning_rate": 9.82311663975557e-06,
+      "loss": 1.2728,
+      "step": 5250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.778002093988857,
+      "learning_rate": 9.822342577875818e-06,
+      "loss": 1.2828,
+      "step": 5255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 20.179978174318997,
+      "learning_rate": 9.821566856632591e-06,
+      "loss": 1.2224,
+      "step": 5260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.651830319926578,
+      "learning_rate": 9.820789476292815e-06,
+      "loss": 1.2582,
+      "step": 5265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 23.417263418666796,
+      "learning_rate": 9.820010437123985e-06,
+      "loss": 1.2399,
+      "step": 5270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.117470831131186,
+      "learning_rate": 9.819229739394167e-06,
+      "loss": 1.2411,
+      "step": 5275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 28.815770897493955,
+      "learning_rate": 9.818447383371999e-06,
+      "loss": 1.2659,
+      "step": 5280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.109159903975375,
+      "learning_rate": 9.817663369326687e-06,
+      "loss": 1.2177,
+      "step": 5285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.63369205165027,
+      "learning_rate": 9.816877697528011e-06,
+      "loss": 1.2485,
+      "step": 5290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.89167586242028,
+      "learning_rate": 9.81609036824632e-06,
+      "loss": 1.2576,
+      "step": 5295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 36.62989419884296,
+      "learning_rate": 9.815301381752535e-06,
+      "loss": 1.2026,
+      "step": 5300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.802536810868858,
+      "learning_rate": 9.814510738318142e-06,
+      "loss": 1.2953,
+      "step": 5305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.092988919751438,
+      "learning_rate": 9.813718438215202e-06,
+      "loss": 1.2312,
+      "step": 5310
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.4301077730165,
+      "learning_rate": 9.812924481716347e-06,
+      "loss": 1.2257,
+      "step": 5315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.0347139548726805,
+      "learning_rate": 9.812128869094773e-06,
+      "loss": 1.2851,
+      "step": 5320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.166163833725394,
+      "learning_rate": 9.811331600624251e-06,
+      "loss": 1.1624,
+      "step": 5325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.087763297173032,
+      "learning_rate": 9.810532676579123e-06,
+      "loss": 1.2585,
+      "step": 5330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.94304547276539,
+      "learning_rate": 9.809732097234295e-06,
+      "loss": 1.2302,
+      "step": 5335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.76779948855747,
+      "learning_rate": 9.808929862865244e-06,
+      "loss": 1.2467,
+      "step": 5340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 27.146992171033162,
+      "learning_rate": 9.808125973748021e-06,
+      "loss": 1.239,
+      "step": 5345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 69.52604053781388,
+      "learning_rate": 9.807320430159245e-06,
+      "loss": 1.2823,
+      "step": 5350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 22.185999195630405,
+      "learning_rate": 9.806513232376097e-06,
+      "loss": 1.2242,
+      "step": 5355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 39.44914138709512,
+      "learning_rate": 9.805704380676338e-06,
+      "loss": 1.2085,
+      "step": 5360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.255961127537361,
+      "learning_rate": 9.804893875338292e-06,
+      "loss": 1.1902,
+      "step": 5365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 31.670780533498856,
+      "learning_rate": 9.804081716640852e-06,
+      "loss": 1.2313,
+      "step": 5370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.131983864366022,
+      "learning_rate": 9.803267904863483e-06,
+      "loss": 1.215,
+      "step": 5375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.312170515341023,
+      "learning_rate": 9.802452440286215e-06,
+      "loss": 1.2263,
+      "step": 5380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.20883072214561,
+      "learning_rate": 9.801635323189648e-06,
+      "loss": 1.3343,
+      "step": 5385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 21.616483837568865,
+      "learning_rate": 9.800816553854952e-06,
+      "loss": 1.2645,
+      "step": 5390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.544994897236405,
+      "learning_rate": 9.799996132563867e-06,
+      "loss": 1.2552,
+      "step": 5395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 22.379944292734105,
+      "learning_rate": 9.799174059598697e-06,
+      "loss": 1.2214,
+      "step": 5400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 28.820684171774843,
+      "learning_rate": 9.798350335242318e-06,
+      "loss": 1.2612,
+      "step": 5405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 17.531294494626554,
+      "learning_rate": 9.797524959778169e-06,
+      "loss": 1.2082,
+      "step": 5410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.5919105729720595,
+      "learning_rate": 9.796697933490265e-06,
+      "loss": 1.2545,
+      "step": 5415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.756273626218609,
+      "learning_rate": 9.795869256663183e-06,
+      "loss": 1.2734,
+      "step": 5420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 29.31089089420677,
+      "learning_rate": 9.79503892958207e-06,
+      "loss": 1.2371,
+      "step": 5425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 19.237086210719905,
+      "learning_rate": 9.79420695253264e-06,
+      "loss": 1.1942,
+      "step": 5430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 43.10425467271122,
+      "learning_rate": 9.79337332580118e-06,
+      "loss": 1.2834,
+      "step": 5435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.207577647989709,
+      "learning_rate": 9.792538049674536e-06,
+      "loss": 1.2432,
+      "step": 5440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.049823339098804,
+      "learning_rate": 9.791701124440123e-06,
+      "loss": 1.3175,
+      "step": 5445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 44.32935270103052,
+      "learning_rate": 9.790862550385933e-06,
+      "loss": 1.2502,
+      "step": 5450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 36.82462495372997,
+      "learning_rate": 9.790022327800515e-06,
+      "loss": 1.2129,
+      "step": 5455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 26.76749530527562,
+      "learning_rate": 9.789180456972989e-06,
+      "loss": 1.2581,
+      "step": 5460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 71.981834353475,
+      "learning_rate": 9.788336938193041e-06,
+      "loss": 1.2926,
+      "step": 5465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 69.81416038234114,
+      "learning_rate": 9.787491771750925e-06,
+      "loss": 1.3102,
+      "step": 5470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 53.959547812668724,
+      "learning_rate": 9.786644957937466e-06,
+      "loss": 1.3059,
+      "step": 5475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 124.99102839739543,
+      "learning_rate": 9.785796497044047e-06,
+      "loss": 1.2209,
+      "step": 5480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 35.57188832728497,
+      "learning_rate": 9.784946389362624e-06,
+      "loss": 1.248,
+      "step": 5485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 98.88287778453991,
+      "learning_rate": 9.784094635185718e-06,
+      "loss": 1.3008,
+      "step": 5490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 120.06635252889072,
+      "learning_rate": 9.783241234806417e-06,
+      "loss": 1.2917,
+      "step": 5495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 47.268439703258316,
+      "learning_rate": 9.782386188518378e-06,
+      "loss": 1.3227,
+      "step": 5500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.03822977666455,
+      "learning_rate": 9.781529496615819e-06,
+      "loss": 1.2413,
+      "step": 5505
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 17.75968374946111,
+      "learning_rate": 9.780671159393525e-06,
+      "loss": 1.2753,
+      "step": 5510
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.653797529576986,
+      "learning_rate": 9.779811177146854e-06,
+      "loss": 1.1932,
+      "step": 5515
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.42135697910021,
+      "learning_rate": 9.778949550171719e-06,
+      "loss": 1.2838,
+      "step": 5520
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.442380055841458,
+      "learning_rate": 9.77808627876461e-06,
+      "loss": 1.2718,
+      "step": 5525
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 25.571013974374363,
+      "learning_rate": 9.777221363222576e-06,
+      "loss": 1.2893,
+      "step": 5530
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.107924256021452,
+      "learning_rate": 9.776354803843233e-06,
+      "loss": 1.223,
+      "step": 5535
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 22.823538370422042,
+      "learning_rate": 9.775486600924765e-06,
+      "loss": 1.205,
+      "step": 5540
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.079573929956554,
+      "learning_rate": 9.774616754765918e-06,
+      "loss": 1.2629,
+      "step": 5545
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.370464473154634,
+      "learning_rate": 9.773745265666006e-06,
+      "loss": 1.2299,
+      "step": 5550
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.160721101979744,
+      "learning_rate": 9.772872133924907e-06,
+      "loss": 1.2204,
+      "step": 5555
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 24.96434652482283,
+      "learning_rate": 9.771997359843066e-06,
+      "loss": 1.2244,
+      "step": 5560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 27.007797934357004,
+      "learning_rate": 9.771120943721492e-06,
+      "loss": 1.2095,
+      "step": 5565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.687110756133366,
+      "learning_rate": 9.770242885861757e-06,
+      "loss": 1.1947,
+      "step": 5570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.129469968585911,
+      "learning_rate": 9.769363186566e-06,
+      "loss": 1.2495,
+      "step": 5575
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 15.770422506749258,
+      "learning_rate": 9.76848184613693e-06,
+      "loss": 1.2225,
+      "step": 5580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 22.39907626509249,
+      "learning_rate": 9.767598864877808e-06,
+      "loss": 1.1946,
+      "step": 5585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 23.564102543690783,
+      "learning_rate": 9.76671424309247e-06,
+      "loss": 1.2236,
+      "step": 5590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.498061833519296,
+      "learning_rate": 9.765827981085314e-06,
+      "loss": 1.2342,
+      "step": 5595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 14.157141212968256,
+      "learning_rate": 9.764940079161302e-06,
+      "loss": 1.265,
+      "step": 5600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.234968885987945,
+      "learning_rate": 9.76405053762596e-06,
+      "loss": 1.2341,
+      "step": 5605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.741718694760706,
+      "learning_rate": 9.76315935678538e-06,
+      "loss": 1.1982,
+      "step": 5610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.162597237447422,
+      "learning_rate": 9.76226653694621e-06,
+      "loss": 1.2361,
+      "step": 5615
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 21.02779574214582,
+      "learning_rate": 9.761372078415675e-06,
+      "loss": 1.2287,
+      "step": 5620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 18.851378631057123,
+      "learning_rate": 9.760475981501558e-06,
+      "loss": 1.2333,
+      "step": 5625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.78763285674776,
+      "learning_rate": 9.7595782465122e-06,
+      "loss": 1.2298,
+      "step": 5630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.69984737716517,
+      "learning_rate": 9.758678873756515e-06,
+      "loss": 1.2666,
+      "step": 5635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.042149705146223,
+      "learning_rate": 9.757777863543973e-06,
+      "loss": 1.2436,
+      "step": 5640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.11977096613918,
+      "learning_rate": 9.756875216184614e-06,
+      "loss": 1.2466,
+      "step": 5645
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.664705978775967,
+      "learning_rate": 9.755970931989035e-06,
+      "loss": 1.2387,
+      "step": 5650
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.252407580225176,
+      "learning_rate": 9.755065011268401e-06,
+      "loss": 1.2514,
+      "step": 5655
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.59180227816643,
+      "learning_rate": 9.754157454334439e-06,
+      "loss": 1.2036,
+      "step": 5660
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.266010042715896,
+      "learning_rate": 9.753248261499437e-06,
+      "loss": 1.2018,
+      "step": 5665
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 19.68909605489335,
+      "learning_rate": 9.752337433076248e-06,
+      "loss": 1.2048,
+      "step": 5670
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 44.27561117531063,
+      "learning_rate": 9.751424969378286e-06,
+      "loss": 1.2434,
+      "step": 5675
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.319303141711027,
+      "learning_rate": 9.750510870719532e-06,
+      "loss": 1.1879,
+      "step": 5680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 25.15578221338594,
+      "learning_rate": 9.749595137414525e-06,
+      "loss": 1.2388,
+      "step": 5685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.321282031004424,
+      "learning_rate": 9.748677769778368e-06,
+      "loss": 1.2015,
+      "step": 5690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 27.291663057795933,
+      "learning_rate": 9.747758768126724e-06,
+      "loss": 1.2821,
+      "step": 5695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.251920040152498,
+      "learning_rate": 9.746838132775823e-06,
+      "loss": 1.1904,
+      "step": 5700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.673801034324477,
+      "learning_rate": 9.745915864042455e-06,
+      "loss": 1.212,
+      "step": 5705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.009051061546915,
+      "learning_rate": 9.744991962243971e-06,
+      "loss": 1.2011,
+      "step": 5710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.415166435895298,
+      "learning_rate": 9.744066427698285e-06,
+      "loss": 1.2115,
+      "step": 5715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.864972467275857,
+      "learning_rate": 9.743139260723871e-06,
+      "loss": 1.2226,
+      "step": 5720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.83282651933836,
+      "learning_rate": 9.74221046163977e-06,
+      "loss": 1.3015,
+      "step": 5725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.517980595445845,
+      "learning_rate": 9.741280030765576e-06,
+      "loss": 1.2031,
+      "step": 5730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.334819861708224,
+      "learning_rate": 9.740347968421453e-06,
+      "loss": 1.2512,
+      "step": 5735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.394363447308308,
+      "learning_rate": 9.739414274928121e-06,
+      "loss": 1.3025,
+      "step": 5740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.250141257436102,
+      "learning_rate": 9.738478950606864e-06,
+      "loss": 1.2673,
+      "step": 5745
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 25.409589436587748,
+      "learning_rate": 9.737541995779526e-06,
+      "loss": 1.1938,
+      "step": 5750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.857813684638622,
+      "learning_rate": 9.736603410768513e-06,
+      "loss": 1.1694,
+      "step": 5755
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.8243029109924285,
+      "learning_rate": 9.735663195896789e-06,
+      "loss": 1.1978,
+      "step": 5760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.954461871946666,
+      "learning_rate": 9.734721351487881e-06,
+      "loss": 1.2346,
+      "step": 5765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.316762828559021,
+      "learning_rate": 9.73377787786588e-06,
+      "loss": 1.2735,
+      "step": 5770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.68701531419754,
+      "learning_rate": 9.732832775355434e-06,
+      "loss": 1.2395,
+      "step": 5775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 21.106624359950015,
+      "learning_rate": 9.731886044281748e-06,
+      "loss": 1.2553,
+      "step": 5780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 26.419039922771095,
+      "learning_rate": 9.730937684970594e-06,
+      "loss": 1.1864,
+      "step": 5785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 14.47606213050339,
+      "learning_rate": 9.729987697748303e-06,
+      "loss": 1.2524,
+      "step": 5790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.793629060293263,
+      "learning_rate": 9.72903608294176e-06,
+      "loss": 1.2426,
+      "step": 5795
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.5793849168018435,
+      "learning_rate": 9.72808284087842e-06,
+      "loss": 1.2307,
+      "step": 5800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.128228531886656,
+      "learning_rate": 9.727127971886289e-06,
+      "loss": 1.2282,
+      "step": 5805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.432836679461422,
+      "learning_rate": 9.726171476293937e-06,
+      "loss": 1.2244,
+      "step": 5810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 19.641558757565917,
+      "learning_rate": 9.725213354430496e-06,
+      "loss": 1.2558,
+      "step": 5815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 26.33319425625301,
+      "learning_rate": 9.724253606625651e-06,
+      "loss": 1.2333,
+      "step": 5820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 51.359947539006086,
+      "learning_rate": 9.723292233209653e-06,
+      "loss": 1.2724,
+      "step": 5825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.29157944242356,
+      "learning_rate": 9.72232923451331e-06,
+      "loss": 1.2224,
+      "step": 5830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.222712003517096,
+      "learning_rate": 9.721364610867988e-06,
+      "loss": 1.2052,
+      "step": 5835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.545359419012799,
+      "learning_rate": 9.72039836260561e-06,
+      "loss": 1.2601,
+      "step": 5840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.516808421680913,
+      "learning_rate": 9.719430490058666e-06,
+      "loss": 1.2155,
+      "step": 5845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 22.072674421278542,
+      "learning_rate": 9.718460993560197e-06,
+      "loss": 1.2381,
+      "step": 5850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 26.640128221824423,
+      "learning_rate": 9.717489873443807e-06,
+      "loss": 1.2868,
+      "step": 5855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.269708168787664,
+      "learning_rate": 9.716517130043658e-06,
+      "loss": 1.2279,
+      "step": 5860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 16.005739462293075,
+      "learning_rate": 9.715542763694469e-06,
+      "loss": 1.2376,
+      "step": 5865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.050689181115803,
+      "learning_rate": 9.714566774731518e-06,
+      "loss": 1.2021,
+      "step": 5870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.528858851920072,
+      "learning_rate": 9.713589163490645e-06,
+      "loss": 1.2167,
+      "step": 5875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.533475071584181,
+      "learning_rate": 9.71260993030824e-06,
+      "loss": 1.2132,
+      "step": 5880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.651551555020754,
+      "learning_rate": 9.71162907552126e-06,
+      "loss": 1.2411,
+      "step": 5885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.091995680595648,
+      "learning_rate": 9.710646599467215e-06,
+      "loss": 1.3136,
+      "step": 5890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.703057172358534,
+      "learning_rate": 9.709662502484177e-06,
+      "loss": 1.2139,
+      "step": 5895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.568655647470476,
+      "learning_rate": 9.708676784910767e-06,
+      "loss": 1.2019,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.499941306187813,
+      "learning_rate": 9.707689447086174e-06,
+      "loss": 1.2124,
+      "step": 5905
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.596113261143561,
+      "learning_rate": 9.706700489350137e-06,
+      "loss": 1.2569,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.712435257088414,
+      "learning_rate": 9.705709912042959e-06,
+      "loss": 1.2312,
+      "step": 5915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.448047445432763,
+      "learning_rate": 9.704717715505494e-06,
+      "loss": 1.2094,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.309107814440093,
+      "learning_rate": 9.703723900079156e-06,
+      "loss": 1.2249,
+      "step": 5925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.665381140645497,
+      "learning_rate": 9.702728466105918e-06,
+      "loss": 1.2193,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.949250005352762,
+      "learning_rate": 9.701731413928305e-06,
+      "loss": 1.2466,
+      "step": 5935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 44.720747917861885,
+      "learning_rate": 9.700732743889402e-06,
+      "loss": 1.2538,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 37.26154099664765,
+      "learning_rate": 9.699732456332855e-06,
+      "loss": 1.2931,
+      "step": 5945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 60.81777030005192,
+      "learning_rate": 9.698730551602857e-06,
+      "loss": 1.229,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 32.24160414710352,
+      "learning_rate": 9.697727030044165e-06,
+      "loss": 1.2347,
+      "step": 5955
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.549838623118922,
+      "learning_rate": 9.696721892002086e-06,
+      "loss": 1.2633,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.840685661279561,
+      "learning_rate": 9.695715137822491e-06,
+      "loss": 1.2155,
+      "step": 5965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 45.466885900243135,
+      "learning_rate": 9.694706767851803e-06,
+      "loss": 1.2183,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 47.74491389796742,
+      "learning_rate": 9.693696782436999e-06,
+      "loss": 1.2244,
+      "step": 5975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 38.78516801788493,
+      "learning_rate": 9.692685181925616e-06,
+      "loss": 1.2745,
+      "step": 5980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.448991128639108,
+      "learning_rate": 9.691671966665743e-06,
+      "loss": 1.2304,
+      "step": 5985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.417128118229789,
+      "learning_rate": 9.690657137006028e-06,
+      "loss": 1.2243,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.355053182742228,
+      "learning_rate": 9.689640693295673e-06,
+      "loss": 1.2213,
+      "step": 5995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.028209869242007,
+      "learning_rate": 9.688622635884434e-06,
+      "loss": 1.2339,
+      "step": 6000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.668657050541347,
+      "learning_rate": 9.687602965122624e-06,
+      "loss": 1.2313,
+      "step": 6005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.902714837352619,
+      "learning_rate": 9.686581681361112e-06,
+      "loss": 1.2749,
+      "step": 6010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.917042572282254,
+      "learning_rate": 9.685558784951318e-06,
+      "loss": 1.1797,
+      "step": 6015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.440952998521315,
+      "learning_rate": 9.684534276245222e-06,
+      "loss": 1.2392,
+      "step": 6020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 28.600345521150423,
+      "learning_rate": 9.683508155595355e-06,
+      "loss": 1.2236,
+      "step": 6025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 55.11319802817678,
+      "learning_rate": 9.682480423354805e-06,
+      "loss": 1.206,
+      "step": 6030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 60.53880107700804,
+      "learning_rate": 9.681451079877214e-06,
+      "loss": 1.2893,
+      "step": 6035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 21.035566080355625,
+      "learning_rate": 9.680420125516779e-06,
+      "loss": 1.2402,
+      "step": 6040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.617150535642468,
+      "learning_rate": 9.679387560628247e-06,
+      "loss": 1.206,
+      "step": 6045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.093646024119007,
+      "learning_rate": 9.678353385566926e-06,
+      "loss": 1.2368,
+      "step": 6050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.733954738163668,
+      "learning_rate": 9.677317600688674e-06,
+      "loss": 1.2561,
+      "step": 6055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.341747092645615,
+      "learning_rate": 9.676280206349902e-06,
+      "loss": 1.2127,
+      "step": 6060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.062867324416539,
+      "learning_rate": 9.675241202907577e-06,
+      "loss": 1.2512,
+      "step": 6065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.630871536778502,
+      "learning_rate": 9.674200590719219e-06,
+      "loss": 1.1599,
+      "step": 6070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.54550227669717,
+      "learning_rate": 9.673158370142902e-06,
+      "loss": 1.158,
+      "step": 6075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 17.44815630750628,
+      "learning_rate": 9.672114541537255e-06,
+      "loss": 1.2572,
+      "step": 6080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.579969741515313,
+      "learning_rate": 9.671069105261457e-06,
+      "loss": 1.1934,
+      "step": 6085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.899566769739347,
+      "learning_rate": 9.67002206167524e-06,
+      "loss": 1.26,
+      "step": 6090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.65903029035593,
+      "learning_rate": 9.66897341113889e-06,
+      "loss": 1.2176,
+      "step": 6095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.406093802458411,
+      "learning_rate": 9.667923154013252e-06,
+      "loss": 1.2592,
+      "step": 6100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.11693527850338,
+      "learning_rate": 9.666871290659715e-06,
+      "loss": 1.219,
+      "step": 6105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.848587868111656,
+      "learning_rate": 9.665817821440223e-06,
+      "loss": 1.2077,
+      "step": 6110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.238908843851462,
+      "learning_rate": 9.664762746717274e-06,
+      "loss": 1.2573,
+      "step": 6115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.335079682624038,
+      "learning_rate": 9.66370606685392e-06,
+      "loss": 1.2615,
+      "step": 6120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.274431760147126,
+      "learning_rate": 9.662647782213763e-06,
+      "loss": 1.2101,
+      "step": 6125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.259435803547051,
+      "learning_rate": 9.661587893160957e-06,
+      "loss": 1.2227,
+      "step": 6130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 18.325180356389684,
+      "learning_rate": 9.66052640006021e-06,
+      "loss": 1.2377,
+      "step": 6135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.132396290240596,
+      "learning_rate": 9.659463303276779e-06,
+      "loss": 1.2625,
+      "step": 6140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.713191427068931,
+      "learning_rate": 9.658398603176478e-06,
+      "loss": 1.2974,
+      "step": 6145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.338937889405493,
+      "learning_rate": 9.657332300125665e-06,
+      "loss": 1.2185,
+      "step": 6150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.81946959153514,
+      "learning_rate": 9.656264394491256e-06,
+      "loss": 1.1804,
+      "step": 6155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.97123339745943,
+      "learning_rate": 9.655194886640715e-06,
+      "loss": 1.2721,
+      "step": 6160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 20.83391204567327,
+      "learning_rate": 9.654123776942061e-06,
+      "loss": 1.2028,
+      "step": 6165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 18.73028944994978,
+      "learning_rate": 9.653051065763862e-06,
+      "loss": 1.3421,
+      "step": 6170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 53.7545975218364,
+      "learning_rate": 9.651976753475234e-06,
+      "loss": 1.2165,
+      "step": 6175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 77.10926771492795,
+      "learning_rate": 9.650900840445848e-06,
+      "loss": 1.2407,
+      "step": 6180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.79236867444254,
+      "learning_rate": 9.649823327045924e-06,
+      "loss": 1.2756,
+      "step": 6185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.08348453792601,
+      "learning_rate": 9.648744213646236e-06,
+      "loss": 1.2452,
+      "step": 6190
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 37.571486457997665,
+      "learning_rate": 9.647663500618105e-06,
+      "loss": 1.2712,
+      "step": 6195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 62.85049842589145,
+      "learning_rate": 9.6465811883334e-06,
+      "loss": 1.2964,
+      "step": 6200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 38.57779119365313,
+      "learning_rate": 9.645497277164547e-06,
+      "loss": 1.2494,
+      "step": 6205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 20.42096778530192,
+      "learning_rate": 9.644411767484518e-06,
+      "loss": 1.2838,
+      "step": 6210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 22.478502839401152,
+      "learning_rate": 9.643324659666835e-06,
+      "loss": 1.2418,
+      "step": 6215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 22.76077350421602,
+      "learning_rate": 9.642235954085572e-06,
+      "loss": 1.2485,
+      "step": 6220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.03143192056874,
+      "learning_rate": 9.641145651115353e-06,
+      "loss": 1.2521,
+      "step": 6225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.175516763267444,
+      "learning_rate": 9.640053751131346e-06,
+      "loss": 1.2065,
+      "step": 6230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 28.09344221326846,
+      "learning_rate": 9.638960254509275e-06,
+      "loss": 1.2107,
+      "step": 6235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.04588546999759,
+      "learning_rate": 9.637865161625413e-06,
+      "loss": 1.2309,
+      "step": 6240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.280353040372706,
+      "learning_rate": 9.636768472856576e-06,
+      "loss": 1.2752,
+      "step": 6245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.312458044057468,
+      "learning_rate": 9.635670188580137e-06,
+      "loss": 1.2841,
+      "step": 6250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 18.780242029323798,
+      "learning_rate": 9.634570309174014e-06,
+      "loss": 1.265,
+      "step": 6255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.44977312785592,
+      "learning_rate": 9.633468835016675e-06,
+      "loss": 1.2838,
+      "step": 6260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.267491130984855,
+      "learning_rate": 9.632365766487135e-06,
+      "loss": 1.2211,
+      "step": 6265
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 35.17393142764635,
+      "learning_rate": 9.631261103964958e-06,
+      "loss": 1.2804,
+      "step": 6270
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 37.53819962798613,
+      "learning_rate": 9.63015484783026e-06,
+      "loss": 1.2865,
+      "step": 6275
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 34.51336920403124,
+      "learning_rate": 9.6290469984637e-06,
+      "loss": 1.2361,
+      "step": 6280
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 26.10458429883579,
+      "learning_rate": 9.62793755624649e-06,
+      "loss": 1.2768,
+      "step": 6285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 34.46040241880211,
+      "learning_rate": 9.626826521560387e-06,
+      "loss": 1.2453,
+      "step": 6290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 88.14491517650183,
+      "learning_rate": 9.625713894787696e-06,
+      "loss": 1.2459,
+      "step": 6295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 76.31252087250449,
+      "learning_rate": 9.624599676311273e-06,
+      "loss": 1.2018,
+      "step": 6300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 28.64535600337029,
+      "learning_rate": 9.623483866514517e-06,
+      "loss": 1.2601,
+      "step": 6305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.035308610589407,
+      "learning_rate": 9.622366465781378e-06,
+      "loss": 1.2424,
+      "step": 6310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.908505676676313,
+      "learning_rate": 9.621247474496357e-06,
+      "loss": 1.2472,
+      "step": 6315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.115144941515627,
+      "learning_rate": 9.620126893044491e-06,
+      "loss": 1.2438,
+      "step": 6320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 37.6967841689743,
+      "learning_rate": 9.619004721811372e-06,
+      "loss": 1.2521,
+      "step": 6325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 23.183543834786384,
+      "learning_rate": 9.617880961183143e-06,
+      "loss": 1.2517,
+      "step": 6330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 21.772660530997452,
+      "learning_rate": 9.616755611546484e-06,
+      "loss": 1.2042,
+      "step": 6335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.213710289509011,
+      "learning_rate": 9.615628673288629e-06,
+      "loss": 1.2467,
+      "step": 6340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.00506589316768,
+      "learning_rate": 9.614500146797356e-06,
+      "loss": 1.2252,
+      "step": 6345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.687034743993428,
+      "learning_rate": 9.61337003246099e-06,
+      "loss": 1.2222,
+      "step": 6350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.589184643064405,
+      "learning_rate": 9.612238330668401e-06,
+      "loss": 1.1862,
+      "step": 6355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.600735030357516,
+      "learning_rate": 9.61110504180901e-06,
+      "loss": 1.2781,
+      "step": 6360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.895062163861428,
+      "learning_rate": 9.609970166272777e-06,
+      "loss": 1.2641,
+      "step": 6365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.94101027761609,
+      "learning_rate": 9.608833704450213e-06,
+      "loss": 1.1981,
+      "step": 6370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.047485735007022,
+      "learning_rate": 9.607695656732375e-06,
+      "loss": 1.2508,
+      "step": 6375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.586597449322517,
+      "learning_rate": 9.60655602351086e-06,
+      "loss": 1.2631,
+      "step": 6380
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.42166296432965,
+      "learning_rate": 9.605414805177817e-06,
+      "loss": 1.2241,
+      "step": 6385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.152314852090806,
+      "learning_rate": 9.604272002125938e-06,
+      "loss": 1.2114,
+      "step": 6390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.068297972056287,
+      "learning_rate": 9.603127614748461e-06,
+      "loss": 1.2552,
+      "step": 6395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 19.653506664350946,
+      "learning_rate": 9.601981643439168e-06,
+      "loss": 1.259,
+      "step": 6400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 41.18724376017987,
+      "learning_rate": 9.600834088592388e-06,
+      "loss": 1.2159,
+      "step": 6405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 24.876109952152202,
+      "learning_rate": 9.599684950602991e-06,
+      "loss": 1.2188,
+      "step": 6410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 22.76694558163489,
+      "learning_rate": 9.598534229866398e-06,
+      "loss": 1.2662,
+      "step": 6415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 17.60411778830573,
+      "learning_rate": 9.597381926778567e-06,
+      "loss": 1.2528,
+      "step": 6420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 24.16780324595837,
+      "learning_rate": 9.596228041736007e-06,
+      "loss": 1.2151,
+      "step": 6425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.10966313520157,
+      "learning_rate": 9.595072575135767e-06,
+      "loss": 1.2488,
+      "step": 6430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.37016194782002,
+      "learning_rate": 9.593915527375443e-06,
+      "loss": 1.2451,
+      "step": 6435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 30.760442676404576,
+      "learning_rate": 9.592756898853173e-06,
+      "loss": 1.1768,
+      "step": 6440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.9327806692135825,
+      "learning_rate": 9.591596689967642e-06,
+      "loss": 1.239,
+      "step": 6445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.935892823030629,
+      "learning_rate": 9.590434901118073e-06,
+      "loss": 1.2067,
+      "step": 6450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.38240903059879,
+      "learning_rate": 9.58927153270424e-06,
+      "loss": 1.2602,
+      "step": 6455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.125354639455605,
+      "learning_rate": 9.588106585126457e-06,
+      "loss": 1.2293,
+      "step": 6460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.382238333633055,
+      "learning_rate": 9.58694005878558e-06,
+      "loss": 1.3058,
+      "step": 6465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 26.545333628441504,
+      "learning_rate": 9.58577195408301e-06,
+      "loss": 1.2527,
+      "step": 6470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.298505320684617,
+      "learning_rate": 9.584602271420688e-06,
+      "loss": 1.275,
+      "step": 6475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 26.275263338563576,
+      "learning_rate": 9.583431011201105e-06,
+      "loss": 1.251,
+      "step": 6480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.408209269007875,
+      "learning_rate": 9.58225817382729e-06,
+      "loss": 1.2217,
+      "step": 6485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 30.065555129882664,
+      "learning_rate": 9.581083759702813e-06,
+      "loss": 1.2567,
+      "step": 6490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 27.407127676629674,
+      "learning_rate": 9.579907769231789e-06,
+      "loss": 1.2801,
+      "step": 6495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 28.56496645801653,
+      "learning_rate": 9.578730202818875e-06,
+      "loss": 1.2509,
+      "step": 6500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.31986863958474,
+      "learning_rate": 9.577551060869274e-06,
+      "loss": 1.2163,
+      "step": 6505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.653806228046228,
+      "learning_rate": 9.576370343788723e-06,
+      "loss": 1.2312,
+      "step": 6510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 13.001146930549856,
+      "learning_rate": 9.57518805198351e-06,
+      "loss": 1.2487,
+      "step": 6515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.961593528674674,
+      "learning_rate": 9.574004185860456e-06,
+      "loss": 1.2569,
+      "step": 6520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 36.483473798179794,
+      "learning_rate": 9.57281874582693e-06,
+      "loss": 1.2644,
+      "step": 6525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 39.88278761239571,
+      "learning_rate": 9.571631732290842e-06,
+      "loss": 1.2123,
+      "step": 6530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 30.69636858528973,
+      "learning_rate": 9.570443145660643e-06,
+      "loss": 1.2888,
+      "step": 6535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.863130870943063,
+      "learning_rate": 9.56925298634532e-06,
+      "loss": 1.2528,
+      "step": 6540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 23.332506732937507,
+      "learning_rate": 9.568061254754411e-06,
+      "loss": 1.244,
+      "step": 6545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 21.000864386202174,
+      "learning_rate": 9.566867951297985e-06,
+      "loss": 1.1825,
+      "step": 6550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 24.943736587546084,
+      "learning_rate": 9.56567307638666e-06,
+      "loss": 1.2365,
+      "step": 6555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 26.197989638859376,
+      "learning_rate": 9.56447663043159e-06,
+      "loss": 1.2024,
+      "step": 6560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 27.33670350537627,
+      "learning_rate": 9.56327861384447e-06,
+      "loss": 1.2344,
+      "step": 6565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 54.36381056789312,
+      "learning_rate": 9.56207902703754e-06,
+      "loss": 1.2252,
+      "step": 6570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 24.11108123688655,
+      "learning_rate": 9.560877870423571e-06,
+      "loss": 1.255,
+      "step": 6575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 26.91323668325551,
+      "learning_rate": 9.559675144415884e-06,
+      "loss": 1.2607,
+      "step": 6580
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.21113053277181,
+      "learning_rate": 9.558470849428336e-06,
+      "loss": 1.2771,
+      "step": 6585
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 21.156681839865353,
+      "learning_rate": 9.557264985875322e-06,
+      "loss": 1.2141,
+      "step": 6590
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.857850606300715,
+      "learning_rate": 9.556057554171779e-06,
+      "loss": 1.2938,
+      "step": 6595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 36.77440472712633,
+      "learning_rate": 9.554848554733183e-06,
+      "loss": 1.2396,
+      "step": 6600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 32.7688105252039,
+      "learning_rate": 9.55363798797555e-06,
+      "loss": 1.2142,
+      "step": 6605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.13894292347681,
+      "learning_rate": 9.552425854315434e-06,
+      "loss": 1.2874,
+      "step": 6610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 26.93915633805332,
+      "learning_rate": 9.55121215416993e-06,
+      "loss": 1.2832,
+      "step": 6615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.665374233208155,
+      "learning_rate": 9.549996887956669e-06,
+      "loss": 1.2471,
+      "step": 6620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.125730711800594,
+      "learning_rate": 9.548780056093826e-06,
+      "loss": 1.2374,
+      "step": 6625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.90762837288638,
+      "learning_rate": 9.547561659000112e-06,
+      "loss": 1.2515,
+      "step": 6630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 22.7067280652705,
+      "learning_rate": 9.546341697094772e-06,
+      "loss": 1.2299,
+      "step": 6635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 24.480892629449738,
+      "learning_rate": 9.545120170797596e-06,
+      "loss": 1.2745,
+      "step": 6640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.729366330188203,
+      "learning_rate": 9.543897080528912e-06,
+      "loss": 1.2389,
+      "step": 6645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.27919718275337,
+      "learning_rate": 9.542672426709582e-06,
+      "loss": 1.2478,
+      "step": 6650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 29.536054001370225,
+      "learning_rate": 9.54144620976101e-06,
+      "loss": 1.208,
+      "step": 6655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.512285015383293,
+      "learning_rate": 9.540218430105133e-06,
+      "loss": 1.227,
+      "step": 6660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.868686602501446,
+      "learning_rate": 9.53898908816443e-06,
+      "loss": 1.2358,
+      "step": 6665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.812767303201554,
+      "learning_rate": 9.537758184361919e-06,
+      "loss": 1.2234,
+      "step": 6670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.52552803155819,
+      "learning_rate": 9.536525719121151e-06,
+      "loss": 1.256,
+      "step": 6675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.192657871379609,
+      "learning_rate": 9.535291692866214e-06,
+      "loss": 1.2279,
+      "step": 6680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.97303139318268,
+      "learning_rate": 9.534056106021739e-06,
+      "loss": 1.2419,
+      "step": 6685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.820777015333706,
+      "learning_rate": 9.532818959012885e-06,
+      "loss": 1.2114,
+      "step": 6690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.200670156862532,
+      "learning_rate": 9.53158025226536e-06,
+      "loss": 1.2167,
+      "step": 6695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 19.482322599930992,
+      "learning_rate": 9.530339986205398e-06,
+      "loss": 1.2416,
+      "step": 6700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 23.067243107377486,
+      "learning_rate": 9.529098161259774e-06,
+      "loss": 1.2141,
+      "step": 6705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.4198088592960145,
+      "learning_rate": 9.527854777855797e-06,
+      "loss": 1.2327,
+      "step": 6710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.785443206784343,
+      "learning_rate": 9.526609836421316e-06,
+      "loss": 1.226,
+      "step": 6715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.46431809622519,
+      "learning_rate": 9.525363337384715e-06,
+      "loss": 1.2453,
+      "step": 6720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 36.130670249795706,
+      "learning_rate": 9.52411528117491e-06,
+      "loss": 1.2554,
+      "step": 6725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 22.559710806107194,
+      "learning_rate": 9.522865668221357e-06,
+      "loss": 1.2525,
+      "step": 6730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.352798022234042,
+      "learning_rate": 9.52161449895405e-06,
+      "loss": 1.2245,
+      "step": 6735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.847839472599194,
+      "learning_rate": 9.52036177380351e-06,
+      "loss": 1.256,
+      "step": 6740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.453488936634951,
+      "learning_rate": 9.519107493200803e-06,
+      "loss": 1.2318,
+      "step": 6745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.633528345767568,
+      "learning_rate": 9.517851657577523e-06,
+      "loss": 1.2229,
+      "step": 6750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.626581824133057,
+      "learning_rate": 9.516594267365804e-06,
+      "loss": 1.2158,
+      "step": 6755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.393721825328813,
+      "learning_rate": 9.51533532299831e-06,
+      "loss": 1.2719,
+      "step": 6760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.769313015181671,
+      "learning_rate": 9.514074824908245e-06,
+      "loss": 1.3028,
+      "step": 6765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.33308617227934,
+      "learning_rate": 9.512812773529343e-06,
+      "loss": 1.2591,
+      "step": 6770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.166131697654846,
+      "learning_rate": 9.511549169295877e-06,
+      "loss": 1.1871,
+      "step": 6775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.136358746864397,
+      "learning_rate": 9.51028401264265e-06,
+      "loss": 1.2101,
+      "step": 6780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.390290412429662,
+      "learning_rate": 9.509017304005003e-06,
+      "loss": 1.195,
+      "step": 6785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.224827613504818,
+      "learning_rate": 9.507749043818806e-06,
+      "loss": 1.218,
+      "step": 6790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 47.04593173534263,
+      "learning_rate": 9.506479232520472e-06,
+      "loss": 1.2103,
+      "step": 6795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.859674831678463,
+      "learning_rate": 9.505207870546935e-06,
+      "loss": 1.1938,
+      "step": 6800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 31.86713786138509,
+      "learning_rate": 9.503934958335674e-06,
+      "loss": 1.1739,
+      "step": 6805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 45.09655897499265,
+      "learning_rate": 9.502660496324695e-06,
+      "loss": 1.2677,
+      "step": 6810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 22.36183306586514,
+      "learning_rate": 9.501384484952542e-06,
+      "loss": 1.2735,
+      "step": 6815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.55254057571248,
+      "learning_rate": 9.500106924658286e-06,
+      "loss": 1.2156,
+      "step": 6820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.712501636130115,
+      "learning_rate": 9.498827815881535e-06,
+      "loss": 1.2817,
+      "step": 6825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 18.22147105174755,
+      "learning_rate": 9.497547159062429e-06,
+      "loss": 1.1614,
+      "step": 6830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.77428145176956,
+      "learning_rate": 9.496264954641642e-06,
+      "loss": 1.184,
+      "step": 6835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 21.482036685862674,
+      "learning_rate": 9.494981203060377e-06,
+      "loss": 1.2329,
+      "step": 6840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 30.82452357899338,
+      "learning_rate": 9.493695904760374e-06,
+      "loss": 1.2574,
+      "step": 6845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 49.07622638487179,
+      "learning_rate": 9.492409060183902e-06,
+      "loss": 1.215,
+      "step": 6850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 27.852350571005495,
+      "learning_rate": 9.491120669773764e-06,
+      "loss": 1.253,
+      "step": 6855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 21.832099841552093,
+      "learning_rate": 9.489830733973294e-06,
+      "loss": 1.2532,
+      "step": 6860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.263626058875476,
+      "learning_rate": 9.488539253226355e-06,
+      "loss": 1.2408,
+      "step": 6865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.705639458146393,
+      "learning_rate": 9.487246227977344e-06,
+      "loss": 1.2532,
+      "step": 6870
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 20.862808964869487,
+      "learning_rate": 9.485951658671195e-06,
+      "loss": 1.2053,
+      "step": 6875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 25.52529554846823,
+      "learning_rate": 9.484655545753365e-06,
+      "loss": 1.2055,
+      "step": 6880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 27.135176135004095,
+      "learning_rate": 9.483357889669844e-06,
+      "loss": 1.2336,
+      "step": 6885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.919200196550955,
+      "learning_rate": 9.482058690867155e-06,
+      "loss": 1.222,
+      "step": 6890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.228121962617985,
+      "learning_rate": 9.480757949792352e-06,
+      "loss": 1.2375,
+      "step": 6895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 38.56471924185,
+      "learning_rate": 9.479455666893017e-06,
+      "loss": 1.2782,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.839326193627322,
+      "learning_rate": 9.478151842617266e-06,
+      "loss": 1.2418,
+      "step": 6905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 25.0169370675844,
+      "learning_rate": 9.476846477413744e-06,
+      "loss": 1.2698,
+      "step": 6910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.44364527322342,
+      "learning_rate": 9.475539571731623e-06,
+      "loss": 1.2483,
+      "step": 6915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 33.7266187137183,
+      "learning_rate": 9.474231126020611e-06,
+      "loss": 1.2221,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.670169197792678,
+      "learning_rate": 9.472921140730942e-06,
+      "loss": 1.2104,
+      "step": 6925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.691279573789632,
+      "learning_rate": 9.47160961631338e-06,
+      "loss": 1.245,
+      "step": 6930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.08452734229773,
+      "learning_rate": 9.470296553219221e-06,
+      "loss": 1.2224,
+      "step": 6935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.076040025569132,
+      "learning_rate": 9.468981951900288e-06,
+      "loss": 1.2418,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.83344240330592,
+      "learning_rate": 9.467665812808933e-06,
+      "loss": 1.232,
+      "step": 6945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.906164830629537,
+      "learning_rate": 9.466348136398038e-06,
+      "loss": 1.2528,
+      "step": 6950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.360240095907476,
+      "learning_rate": 9.465028923121016e-06,
+      "loss": 1.2909,
+      "step": 6955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.91868384199132,
+      "learning_rate": 9.463708173431808e-06,
+      "loss": 1.2611,
+      "step": 6960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 42.551322031638165,
+      "learning_rate": 9.462385887784878e-06,
+      "loss": 1.2037,
+      "step": 6965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.858468890492471,
+      "learning_rate": 9.461062066635227e-06,
+      "loss": 1.2246,
+      "step": 6970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 46.32101587841062,
+      "learning_rate": 9.45973671043838e-06,
+      "loss": 1.2431,
+      "step": 6975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 27.18892845859127,
+      "learning_rate": 9.45840981965039e-06,
+      "loss": 1.1757,
+      "step": 6980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 41.3387922376486,
+      "learning_rate": 9.457081394727839e-06,
+      "loss": 1.2043,
+      "step": 6985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 74.34153904649125,
+      "learning_rate": 9.455751436127838e-06,
+      "loss": 1.2271,
+      "step": 6990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 47.880845285006515,
+      "learning_rate": 9.454419944308023e-06,
+      "loss": 1.289,
+      "step": 6995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.31019170814755,
+      "learning_rate": 9.45308691972656e-06,
+      "loss": 1.2575,
+      "step": 7000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 35.09321220970617,
+      "learning_rate": 9.451752362842142e-06,
+      "loss": 1.2396,
+      "step": 7005
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 45.95642775325224,
+      "learning_rate": 9.450416274113984e-06,
+      "loss": 1.2274,
+      "step": 7010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 20.437604930753526,
+      "learning_rate": 9.44907865400184e-06,
+      "loss": 1.2589,
+      "step": 7015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 110.20869446046459,
+      "learning_rate": 9.447739502965981e-06,
+      "loss": 1.26,
+      "step": 7020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 55.010558102721724,
+      "learning_rate": 9.446398821467207e-06,
+      "loss": 1.27,
+      "step": 7025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 37.421245925011334,
+      "learning_rate": 9.445056609966843e-06,
+      "loss": 1.2753,
+      "step": 7030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 58.58172405386906,
+      "learning_rate": 9.443712868926747e-06,
+      "loss": 1.2242,
+      "step": 7035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 108.11313414129204,
+      "learning_rate": 9.442367598809296e-06,
+      "loss": 1.3262,
+      "step": 7040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 53.60255208092043,
+      "learning_rate": 9.441020800077398e-06,
+      "loss": 1.2836,
+      "step": 7045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 39.63944842158645,
+      "learning_rate": 9.439672473194484e-06,
+      "loss": 1.2709,
+      "step": 7050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 58.28398000888765,
+      "learning_rate": 9.438322618624514e-06,
+      "loss": 1.284,
+      "step": 7055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 47.30731552146769,
+      "learning_rate": 9.436971236831966e-06,
+      "loss": 1.2984,
+      "step": 7060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 35.50017341642059,
+      "learning_rate": 9.435618328281856e-06,
+      "loss": 1.2511,
+      "step": 7065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 19.540606098701666,
+      "learning_rate": 9.434263893439717e-06,
+      "loss": 1.2829,
+      "step": 7070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.46016735290377,
+      "learning_rate": 9.432907932771604e-06,
+      "loss": 1.2504,
+      "step": 7075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.716698511129454,
+      "learning_rate": 9.431550446744109e-06,
+      "loss": 1.2351,
+      "step": 7080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 22.396806981332013,
+      "learning_rate": 9.430191435824335e-06,
+      "loss": 1.2902,
+      "step": 7085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.093918197190177,
+      "learning_rate": 9.42883090047992e-06,
+      "loss": 1.2596,
+      "step": 7090
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.438683461999606,
+      "learning_rate": 9.427468841179025e-06,
+      "loss": 1.2259,
+      "step": 7095
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.475889676398804,
+      "learning_rate": 9.426105258390326e-06,
+      "loss": 1.1868,
+      "step": 7100
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.25807562795457,
+      "learning_rate": 9.424740152583037e-06,
+      "loss": 1.2288,
+      "step": 7105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.92935578434538,
+      "learning_rate": 9.423373524226888e-06,
+      "loss": 1.2415,
+      "step": 7110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.035920194295393,
+      "learning_rate": 9.422005373792134e-06,
+      "loss": 1.214,
+      "step": 7115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.952893664109181,
+      "learning_rate": 9.420635701749553e-06,
+      "loss": 1.2441,
+      "step": 7120
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.24312828879759,
+      "learning_rate": 9.41926450857045e-06,
+      "loss": 1.2545,
+      "step": 7125
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.895593284970484,
+      "learning_rate": 9.41789179472665e-06,
+      "loss": 1.2717,
+      "step": 7130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 25.915344582786382,
+      "learning_rate": 9.416517560690505e-06,
+      "loss": 1.2114,
+      "step": 7135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 28.063154100394737,
+      "learning_rate": 9.415141806934885e-06,
+      "loss": 1.1602,
+      "step": 7140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 60.96186172132934,
+      "learning_rate": 9.413764533933186e-06,
+      "loss": 1.2729,
+      "step": 7145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 34.2103281919088,
+      "learning_rate": 9.412385742159325e-06,
+      "loss": 1.1829,
+      "step": 7150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 31.55595638712618,
+      "learning_rate": 9.411005432087745e-06,
+      "loss": 1.2225,
+      "step": 7155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.44544536704509,
+      "learning_rate": 9.409623604193409e-06,
+      "loss": 1.1842,
+      "step": 7160
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 19.01930018142275,
+      "learning_rate": 9.408240258951803e-06,
+      "loss": 1.234,
+      "step": 7165
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.31729099988006,
+      "learning_rate": 9.406855396838934e-06,
+      "loss": 1.2533,
+      "step": 7170
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.758603793971947,
+      "learning_rate": 9.405469018331333e-06,
+      "loss": 1.2281,
+      "step": 7175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.729019158963341,
+      "learning_rate": 9.404081123906048e-06,
+      "loss": 1.2155,
+      "step": 7180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.591024723043679,
+      "learning_rate": 9.402691714040658e-06,
+      "loss": 1.2419,
+      "step": 7185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.408729873579752,
+      "learning_rate": 9.401300789213251e-06,
+      "loss": 1.2615,
+      "step": 7190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.981977669792066,
+      "learning_rate": 9.399908349902448e-06,
+      "loss": 1.2266,
+      "step": 7195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 20.35521316100864,
+      "learning_rate": 9.398514396587383e-06,
+      "loss": 1.2822,
+      "step": 7200
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.621038660870372,
+      "learning_rate": 9.397118929747716e-06,
+      "loss": 1.2243,
+      "step": 7205
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.1169999232885415,
+      "learning_rate": 9.395721949863626e-06,
+      "loss": 1.2136,
+      "step": 7210
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.05921741295788,
+      "learning_rate": 9.39432345741581e-06,
+      "loss": 1.2538,
+      "step": 7215
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.0163186370097,
+      "learning_rate": 9.39292345288549e-06,
+      "loss": 1.1999,
+      "step": 7220
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.509276797106333,
+      "learning_rate": 9.391521936754405e-06,
+      "loss": 1.1838,
+      "step": 7225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.682679767995246,
+      "learning_rate": 9.390118909504816e-06,
+      "loss": 1.2345,
+      "step": 7230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.298106706947552,
+      "learning_rate": 9.388714371619504e-06,
+      "loss": 1.2146,
+      "step": 7235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 32.68014420140746,
+      "learning_rate": 9.387308323581767e-06,
+      "loss": 1.2208,
+      "step": 7240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 32.94240281632185,
+      "learning_rate": 9.385900765875428e-06,
+      "loss": 1.2308,
+      "step": 7245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.20941662216896,
+      "learning_rate": 9.384491698984824e-06,
+      "loss": 1.2556,
+      "step": 7250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 44.34305604752469,
+      "learning_rate": 9.383081123394812e-06,
+      "loss": 1.2736,
+      "step": 7255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 25.712638034601298,
+      "learning_rate": 9.381669039590774e-06,
+      "loss": 1.2039,
+      "step": 7260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 48.74925265761191,
+      "learning_rate": 9.380255448058605e-06,
+      "loss": 1.2437,
+      "step": 7265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 57.63732263239281,
+      "learning_rate": 9.378840349284719e-06,
+      "loss": 1.2407,
+      "step": 7270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.103386894214017,
+      "learning_rate": 9.377423743756052e-06,
+      "loss": 1.2584,
+      "step": 7275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 23.587086106979573,
+      "learning_rate": 9.376005631960054e-06,
+      "loss": 1.1959,
+      "step": 7280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 47.89273622586558,
+      "learning_rate": 9.374586014384698e-06,
+      "loss": 1.2304,
+      "step": 7285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.207434380149833,
+      "learning_rate": 9.373164891518474e-06,
+      "loss": 1.2026,
+      "step": 7290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 25.565965293698454,
+      "learning_rate": 9.371742263850386e-06,
+      "loss": 1.2605,
+      "step": 7295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.987314783270103,
+      "learning_rate": 9.370318131869962e-06,
+      "loss": 1.2852,
+      "step": 7300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.170501768501227,
+      "learning_rate": 9.368892496067242e-06,
+      "loss": 1.2447,
+      "step": 7305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 24.983723833865042,
+      "learning_rate": 9.367465356932786e-06,
+      "loss": 1.2606,
+      "step": 7310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.967512176058351,
+      "learning_rate": 9.366036714957673e-06,
+      "loss": 1.2554,
+      "step": 7315
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.222476940885444,
+      "learning_rate": 9.364606570633496e-06,
+      "loss": 1.2277,
+      "step": 7320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.599376103937072,
+      "learning_rate": 9.363174924452368e-06,
+      "loss": 1.2546,
+      "step": 7325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.74402561354435,
+      "learning_rate": 9.361741776906914e-06,
+      "loss": 1.2431,
+      "step": 7330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.350739519335928,
+      "learning_rate": 9.360307128490282e-06,
+      "loss": 1.2834,
+      "step": 7335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 58.73070927096001,
+      "learning_rate": 9.358870979696132e-06,
+      "loss": 1.2686,
+      "step": 7340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.974912177329822,
+      "learning_rate": 9.35743333101864e-06,
+      "loss": 1.1898,
+      "step": 7345
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 17.78339694738643,
+      "learning_rate": 9.355994182952501e-06,
+      "loss": 1.2043,
+      "step": 7350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.31437657156583,
+      "learning_rate": 9.354553535992923e-06,
+      "loss": 1.2304,
+      "step": 7355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 20.665349554537844,
+      "learning_rate": 9.353111390635634e-06,
+      "loss": 1.189,
+      "step": 7360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.05751323569656,
+      "learning_rate": 9.351667747376874e-06,
+      "loss": 1.2734,
+      "step": 7365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.933435798473003,
+      "learning_rate": 9.350222606713396e-06,
+      "loss": 1.1759,
+      "step": 7370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.62867231923937,
+      "learning_rate": 9.348775969142475e-06,
+      "loss": 1.2595,
+      "step": 7375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.329043149847656,
+      "learning_rate": 9.347327835161897e-06,
+      "loss": 1.2506,
+      "step": 7380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.351497862234899,
+      "learning_rate": 9.345878205269962e-06,
+      "loss": 1.253,
+      "step": 7385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.719251752013907,
+      "learning_rate": 9.344427079965487e-06,
+      "loss": 1.2536,
+      "step": 7390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.942666744527387,
+      "learning_rate": 9.342974459747804e-06,
+      "loss": 1.26,
+      "step": 7395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.285466703612645,
+      "learning_rate": 9.341520345116759e-06,
+      "loss": 1.18,
+      "step": 7400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 30.12228972307043,
+      "learning_rate": 9.34006473657271e-06,
+      "loss": 1.2179,
+      "step": 7405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.21589668238483,
+      "learning_rate": 9.338607634616528e-06,
+      "loss": 1.2136,
+      "step": 7410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.842802840102594,
+      "learning_rate": 9.337149039749603e-06,
+      "loss": 1.1953,
+      "step": 7415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.58284469305697,
+      "learning_rate": 9.335688952473836e-06,
+      "loss": 1.2499,
+      "step": 7420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.622043712284013,
+      "learning_rate": 9.334227373291642e-06,
+      "loss": 1.2617,
+      "step": 7425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.883722967582399,
+      "learning_rate": 9.33276430270595e-06,
+      "loss": 1.2706,
+      "step": 7430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.60015415599963,
+      "learning_rate": 9.331299741220196e-06,
+      "loss": 1.2676,
+      "step": 7435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.111803702436317,
+      "learning_rate": 9.329833689338342e-06,
+      "loss": 1.2369,
+      "step": 7440
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.055944336562472,
+      "learning_rate": 9.32836614756485e-06,
+      "loss": 1.259,
+      "step": 7445
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 28.08428867001668,
+      "learning_rate": 9.326897116404698e-06,
+      "loss": 1.2645,
+      "step": 7450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.352690250363281,
+      "learning_rate": 9.325426596363382e-06,
+      "loss": 1.2373,
+      "step": 7455
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.900462142361196,
+      "learning_rate": 9.323954587946907e-06,
+      "loss": 1.186,
+      "step": 7460
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.566500090186759,
+      "learning_rate": 9.322481091661788e-06,
+      "loss": 1.2952,
+      "step": 7465
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.033891289696,
+      "learning_rate": 9.321006108015053e-06,
+      "loss": 1.272,
+      "step": 7470
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 14.899753735922808,
+      "learning_rate": 9.319529637514244e-06,
+      "loss": 1.2528,
+      "step": 7475
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.255310223264777,
+      "learning_rate": 9.318051680667412e-06,
+      "loss": 1.2085,
+      "step": 7480
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.864126836858343,
+      "learning_rate": 9.316572237983119e-06,
+      "loss": 1.2642,
+      "step": 7485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.551752495088672,
+      "learning_rate": 9.315091309970444e-06,
+      "loss": 1.2041,
+      "step": 7490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 22.80093972121296,
+      "learning_rate": 9.31360889713897e-06,
+      "loss": 1.2334,
+      "step": 7495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 28.90301544894272,
+      "learning_rate": 9.312124999998796e-06,
+      "loss": 1.2479,
+      "step": 7500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 23.575320478557227,
+      "learning_rate": 9.310639619060525e-06,
+      "loss": 1.1864,
+      "step": 7505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 27.62402399391801,
+      "learning_rate": 9.30915275483528e-06,
+      "loss": 1.2981,
+      "step": 7510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 29.52538342623024,
+      "learning_rate": 9.307664407834687e-06,
+      "loss": 1.2367,
+      "step": 7515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 27.250936315036952,
+      "learning_rate": 9.306174578570886e-06,
+      "loss": 1.2476,
+      "step": 7520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 29.535063998267166,
+      "learning_rate": 9.304683267556526e-06,
+      "loss": 1.253,
+      "step": 7525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 28.251195584812745,
+      "learning_rate": 9.303190475304765e-06,
+      "loss": 1.2189,
+      "step": 7530
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.902149710792362,
+      "learning_rate": 9.301696202329271e-06,
+      "loss": 1.2406,
+      "step": 7535
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 40.43104040021645,
+      "learning_rate": 9.300200449144222e-06,
+      "loss": 1.2063,
+      "step": 7540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.089939602148963,
+      "learning_rate": 9.298703216264306e-06,
+      "loss": 1.2279,
+      "step": 7545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 16.76947542440218,
+      "learning_rate": 9.29720450420472e-06,
+      "loss": 1.2208,
+      "step": 7550
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.613305081206319,
+      "learning_rate": 9.295704313481167e-06,
+      "loss": 1.2397,
+      "step": 7555
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 29.797823991883156,
+      "learning_rate": 9.294202644609863e-06,
+      "loss": 1.1968,
+      "step": 7560
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.489259290104659,
+      "learning_rate": 9.292699498107529e-06,
+      "loss": 1.2035,
+      "step": 7565
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.944090986437367,
+      "learning_rate": 9.291194874491401e-06,
+      "loss": 1.2632,
+      "step": 7570
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 21.959976388311894,
+      "learning_rate": 9.289688774279213e-06,
+      "loss": 1.2442,
+      "step": 7575
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.984896080723992,
+      "learning_rate": 9.288181197989215e-06,
+      "loss": 1.24,
+      "step": 7580
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.296258510485956,
+      "learning_rate": 9.286672146140162e-06,
+      "loss": 1.1646,
+      "step": 7585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.409430913426574,
+      "learning_rate": 9.28516161925132e-06,
+      "loss": 1.2424,
+      "step": 7590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.815724731539182,
+      "learning_rate": 9.283649617842455e-06,
+      "loss": 1.1497,
+      "step": 7595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.245525477882799,
+      "learning_rate": 9.282136142433849e-06,
+      "loss": 1.2134,
+      "step": 7600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.055682706845232,
+      "learning_rate": 9.280621193546286e-06,
+      "loss": 1.2681,
+      "step": 7605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.08566928740703,
+      "learning_rate": 9.279104771701059e-06,
+      "loss": 1.2022,
+      "step": 7610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.766664575908408,
+      "learning_rate": 9.277586877419967e-06,
+      "loss": 1.2152,
+      "step": 7615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.087774417016558,
+      "learning_rate": 9.276067511225318e-06,
+      "loss": 1.2306,
+      "step": 7620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 29.673522022975185,
+      "learning_rate": 9.274546673639919e-06,
+      "loss": 1.2166,
+      "step": 7625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 51.539591817311056,
+      "learning_rate": 9.273024365187093e-06,
+      "loss": 1.2284,
+      "step": 7630
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.140019429296594,
+      "learning_rate": 9.271500586390666e-06,
+      "loss": 1.2434,
+      "step": 7635
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 26.036718584654587,
+      "learning_rate": 9.269975337774967e-06,
+      "loss": 1.1555,
+      "step": 7640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.191433366044787,
+      "learning_rate": 9.268448619864832e-06,
+      "loss": 1.2142,
+      "step": 7645
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.33184538805689,
+      "learning_rate": 9.266920433185603e-06,
+      "loss": 1.2552,
+      "step": 7650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 25.21325091607401,
+      "learning_rate": 9.265390778263129e-06,
+      "loss": 1.2298,
+      "step": 7655
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.156688433644524,
+      "learning_rate": 9.263859655623761e-06,
+      "loss": 1.2132,
+      "step": 7660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.300367798956218,
+      "learning_rate": 9.262327065794358e-06,
+      "loss": 1.2519,
+      "step": 7665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.169919324760624,
+      "learning_rate": 9.260793009302284e-06,
+      "loss": 1.2031,
+      "step": 7670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 25.815209942981188,
+      "learning_rate": 9.259257486675404e-06,
+      "loss": 1.2059,
+      "step": 7675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 23.533425674532772,
+      "learning_rate": 9.25772049844209e-06,
+      "loss": 1.2389,
+      "step": 7680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.829696761260031,
+      "learning_rate": 9.256182045131222e-06,
+      "loss": 1.1899,
+      "step": 7685
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.084671135829076,
+      "learning_rate": 9.254642127272175e-06,
+      "loss": 1.2283,
+      "step": 7690
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.66814246775726,
+      "learning_rate": 9.253100745394836e-06,
+      "loss": 1.2656,
+      "step": 7695
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.733947935037307,
+      "learning_rate": 9.251557900029593e-06,
+      "loss": 1.2077,
+      "step": 7700
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.323735837842106,
+      "learning_rate": 9.250013591707339e-06,
+      "loss": 1.2519,
+      "step": 7705
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.768122637522222,
+      "learning_rate": 9.248467820959467e-06,
+      "loss": 1.2063,
+      "step": 7710
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 17.952374205143265,
+      "learning_rate": 9.246920588317873e-06,
+      "loss": 1.2455,
+      "step": 7715
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 22.129056751142848,
+      "learning_rate": 9.245371894314962e-06,
+      "loss": 1.2426,
+      "step": 7720
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.67588349421575,
+      "learning_rate": 9.243821739483638e-06,
+      "loss": 1.227,
+      "step": 7725
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 77.09409480945052,
+      "learning_rate": 9.242270124357306e-06,
+      "loss": 1.1896,
+      "step": 7730
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 17.986491603701044,
+      "learning_rate": 9.240717049469874e-06,
+      "loss": 1.2003,
+      "step": 7735
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 29.014583020401524,
+      "learning_rate": 9.239162515355759e-06,
+      "loss": 1.2147,
+      "step": 7740
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.372813185319659,
+      "learning_rate": 9.23760652254987e-06,
+      "loss": 1.2414,
+      "step": 7745
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.178809016800125,
+      "learning_rate": 9.236049071587623e-06,
+      "loss": 1.2296,
+      "step": 7750
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 20.288685383202317,
+      "learning_rate": 9.234490163004938e-06,
+      "loss": 1.2335,
+      "step": 7755
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 38.59795381962454,
+      "learning_rate": 9.232929797338231e-06,
+      "loss": 1.2161,
+      "step": 7760
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.725619510243467,
+      "learning_rate": 9.231367975124425e-06,
+      "loss": 1.2327,
+      "step": 7765
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 19.0956332935042,
+      "learning_rate": 9.229804696900938e-06,
+      "loss": 1.2172,
+      "step": 7770
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.363008188568402,
+      "learning_rate": 9.228239963205697e-06,
+      "loss": 1.2235,
+      "step": 7775
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.919405963264193,
+      "learning_rate": 9.226673774577123e-06,
+      "loss": 1.2012,
+      "step": 7780
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.370760427375297,
+      "learning_rate": 9.225106131554138e-06,
+      "loss": 1.2742,
+      "step": 7785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.574753334317899,
+      "learning_rate": 9.22353703467617e-06,
+      "loss": 1.2251,
+      "step": 7790
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.919204055521005,
+      "learning_rate": 9.221966484483143e-06,
+      "loss": 1.2137,
+      "step": 7795
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.117727775224042,
+      "learning_rate": 9.22039448151548e-06,
+      "loss": 1.2701,
+      "step": 7800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.821250378982779,
+      "learning_rate": 9.218821026314106e-06,
+      "loss": 1.2434,
+      "step": 7805
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.83726533956292,
+      "learning_rate": 9.217246119420449e-06,
+      "loss": 1.2364,
+      "step": 7810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 25.25069526526832,
+      "learning_rate": 9.215669761376428e-06,
+      "loss": 1.2415,
+      "step": 7815
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 27.832030681337162,
+      "learning_rate": 9.214091952724469e-06,
+      "loss": 1.2075,
+      "step": 7820
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 31.352547290909595,
+      "learning_rate": 9.212512694007494e-06,
+      "loss": 1.2429,
+      "step": 7825
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 59.52508502860254,
+      "learning_rate": 9.210931985768924e-06,
+      "loss": 1.2324,
+      "step": 7830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 42.29197901671513,
+      "learning_rate": 9.209349828552681e-06,
+      "loss": 1.2324,
+      "step": 7835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 58.09267680215089,
+      "learning_rate": 9.207766222903182e-06,
+      "loss": 1.281,
+      "step": 7840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.99424188929102,
+      "learning_rate": 9.206181169365345e-06,
+      "loss": 1.211,
+      "step": 7845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 37.79255731226824,
+      "learning_rate": 9.204594668484584e-06,
+      "loss": 1.2649,
+      "step": 7850
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 32.152434144982294,
+      "learning_rate": 9.203006720806813e-06,
+      "loss": 1.288,
+      "step": 7855
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.160381783938107,
+      "learning_rate": 9.201417326878444e-06,
+      "loss": 1.2589,
+      "step": 7860
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 24.820525431369962,
+      "learning_rate": 9.199826487246386e-06,
+      "loss": 1.2355,
+      "step": 7865
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.610850830909609,
+      "learning_rate": 9.198234202458045e-06,
+      "loss": 1.2876,
+      "step": 7870
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.256979389282636,
+      "learning_rate": 9.196640473061325e-06,
+      "loss": 1.2339,
+      "step": 7875
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.769905681918171,
+      "learning_rate": 9.195045299604626e-06,
+      "loss": 1.237,
+      "step": 7880
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.99103804897849,
+      "learning_rate": 9.193448682636846e-06,
+      "loss": 1.2418,
+      "step": 7885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.75457384662508,
+      "learning_rate": 9.19185062270738e-06,
+      "loss": 1.2511,
+      "step": 7890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.757096051723074,
+      "learning_rate": 9.190251120366118e-06,
+      "loss": 1.2335,
+      "step": 7895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 42.73464975981768,
+      "learning_rate": 9.188650176163448e-06,
+      "loss": 1.2636,
+      "step": 7900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.765238752923903,
+      "learning_rate": 9.187047790650252e-06,
+      "loss": 1.2848,
+      "step": 7905
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.11348257275226,
+      "learning_rate": 9.185443964377911e-06,
+      "loss": 1.2391,
+      "step": 7910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.159478032563705,
+      "learning_rate": 9.1838386978983e-06,
+      "loss": 1.2208,
+      "step": 7915
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.609278312997958,
+      "learning_rate": 9.18223199176379e-06,
+      "loss": 1.2904,
+      "step": 7920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 25.709645643557472,
+      "learning_rate": 9.180623846527244e-06,
+      "loss": 1.2852,
+      "step": 7925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.165928500885084,
+      "learning_rate": 9.179014262742027e-06,
+      "loss": 1.2282,
+      "step": 7930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.092276441061669,
+      "learning_rate": 9.177403240961993e-06,
+      "loss": 1.2677,
+      "step": 7935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.391862355681053,
+      "learning_rate": 9.175790781741493e-06,
+      "loss": 1.2481,
+      "step": 7940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.279028816900641,
+      "learning_rate": 9.174176885635373e-06,
+      "loss": 1.244,
+      "step": 7945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.923061253265221,
+      "learning_rate": 9.172561553198974e-06,
+      "loss": 1.184,
+      "step": 7950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.690954537377602,
+      "learning_rate": 9.17094478498813e-06,
+      "loss": 1.2314,
+      "step": 7955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.831051956873171,
+      "learning_rate": 9.16932658155917e-06,
+      "loss": 1.2684,
+      "step": 7960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.512000309343122,
+      "learning_rate": 9.167706943468916e-06,
+      "loss": 1.217,
+      "step": 7965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.104503401851176,
+      "learning_rate": 9.16608587127468e-06,
+      "loss": 1.1921,
+      "step": 7970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 31.74318596256183,
+      "learning_rate": 9.164463365534277e-06,
+      "loss": 1.2401,
+      "step": 7975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.413034406609874,
+      "learning_rate": 9.162839426806007e-06,
+      "loss": 1.2366,
+      "step": 7980
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.185415863724284,
+      "learning_rate": 9.161214055648667e-06,
+      "loss": 1.2459,
+      "step": 7985
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.188366129864246,
+      "learning_rate": 9.159587252621545e-06,
+      "loss": 1.2086,
+      "step": 7990
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.707374452769816,
+      "learning_rate": 9.157959018284421e-06,
+      "loss": 1.1813,
+      "step": 7995
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.354744724838543,
+      "learning_rate": 9.15632935319757e-06,
+      "loss": 1.2413,
+      "step": 8000
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 22.372889845701312,
+      "learning_rate": 9.15469825792176e-06,
+      "loss": 1.2657,
+      "step": 8005
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.73702140543741,
+      "learning_rate": 9.153065733018247e-06,
+      "loss": 1.2369,
+      "step": 8010
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.564815088861721,
+      "learning_rate": 9.15143177904878e-06,
+      "loss": 1.2148,
+      "step": 8015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.606622739126877,
+      "learning_rate": 9.149796396575606e-06,
+      "loss": 1.2042,
+      "step": 8020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.368792590493465,
+      "learning_rate": 9.148159586161454e-06,
+      "loss": 1.2278,
+      "step": 8025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.079034496857842,
+      "learning_rate": 9.146521348369549e-06,
+      "loss": 1.236,
+      "step": 8030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.571565172228567,
+      "learning_rate": 9.14488168376361e-06,
+      "loss": 1.2529,
+      "step": 8035
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.846307319390768,
+      "learning_rate": 9.143240592907842e-06,
+      "loss": 1.1753,
+      "step": 8040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.865132020941735,
+      "learning_rate": 9.141598076366942e-06,
+      "loss": 1.1693,
+      "step": 8045
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.827255680234257,
+      "learning_rate": 9.139954134706102e-06,
+      "loss": 1.2104,
+      "step": 8050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 30.57885664997976,
+      "learning_rate": 9.138308768490998e-06,
+      "loss": 1.3031,
+      "step": 8055
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.607095764384038,
+      "learning_rate": 9.136661978287799e-06,
+      "loss": 1.2254,
+      "step": 8060
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.89714796529194,
+      "learning_rate": 9.135013764663163e-06,
+      "loss": 1.2506,
+      "step": 8065
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.236539667441939,
+      "learning_rate": 9.133364128184242e-06,
+      "loss": 1.2887,
+      "step": 8070
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.598284133793374,
+      "learning_rate": 9.131713069418671e-06,
+      "loss": 1.218,
+      "step": 8075
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.89993729959541,
+      "learning_rate": 9.130060588934578e-06,
+      "loss": 1.2071,
+      "step": 8080
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.775313601119954,
+      "learning_rate": 9.128406687300582e-06,
+      "loss": 1.2395,
+      "step": 8085
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 27.60652757738097,
+      "learning_rate": 9.12675136508579e-06,
+      "loss": 1.1755,
+      "step": 8090
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 31.59229218037076,
+      "learning_rate": 9.125094622859791e-06,
+      "loss": 1.2039,
+      "step": 8095
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 38.65978472538587,
+      "learning_rate": 9.123436461192674e-06,
+      "loss": 1.2323,
+      "step": 8100
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 65.80980214245291,
+      "learning_rate": 9.12177688065501e-06,
+      "loss": 1.2492,
+      "step": 8105
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 59.28577568454184,
+      "learning_rate": 9.120115881817857e-06,
+      "loss": 1.2893,
+      "step": 8110
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 27.688683401321523,
+      "learning_rate": 9.118453465252764e-06,
+      "loss": 1.2641,
+      "step": 8115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.70332351800936,
+      "learning_rate": 9.116789631531769e-06,
+      "loss": 1.181,
+      "step": 8120
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 60.407514738702325,
+      "learning_rate": 9.115124381227392e-06,
+      "loss": 1.2316,
+      "step": 8125
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.859070082123754,
+      "learning_rate": 9.113457714912646e-06,
+      "loss": 1.2146,
+      "step": 8130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 42.63500580644595,
+      "learning_rate": 9.111789633161029e-06,
+      "loss": 1.2345,
+      "step": 8135
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 23.14253982357684,
+      "learning_rate": 9.110120136546528e-06,
+      "loss": 1.2374,
+      "step": 8140
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.973157486553424,
+      "learning_rate": 9.108449225643612e-06,
+      "loss": 1.2557,
+      "step": 8145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 21.157793547602015,
+      "learning_rate": 9.10677690102724e-06,
+      "loss": 1.2004,
+      "step": 8150
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.801783703870495,
+      "learning_rate": 9.105103163272862e-06,
+      "loss": 1.2679,
+      "step": 8155
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.444503620625873,
+      "learning_rate": 9.103428012956406e-06,
+      "loss": 1.228,
+      "step": 8160
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 26.61347370151708,
+      "learning_rate": 9.101751450654289e-06,
+      "loss": 1.2171,
+      "step": 8165
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 27.953381232035486,
+      "learning_rate": 9.100073476943415e-06,
+      "loss": 1.2415,
+      "step": 8170
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 20.37772713019263,
+      "learning_rate": 9.098394092401174e-06,
+      "loss": 1.2158,
+      "step": 8175
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 16.193271631825727,
+      "learning_rate": 9.096713297605439e-06,
+      "loss": 1.2632,
+      "step": 8180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.45725368852007,
+      "learning_rate": 9.095031093134574e-06,
+      "loss": 1.2415,
+      "step": 8185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 25.74841599073145,
+      "learning_rate": 9.093347479567419e-06,
+      "loss": 1.2874,
+      "step": 8190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.071323002953288,
+      "learning_rate": 9.091662457483305e-06,
+      "loss": 1.2536,
+      "step": 8195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 23.679119240128742,
+      "learning_rate": 9.08997602746205e-06,
+      "loss": 1.2502,
+      "step": 8200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.481301766638277,
+      "learning_rate": 9.088288190083949e-06,
+      "loss": 1.2335,
+      "step": 8205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.789345499913125,
+      "learning_rate": 9.086598945929787e-06,
+      "loss": 1.2299,
+      "step": 8210
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 31.442603486590052,
+      "learning_rate": 9.08490829558083e-06,
+      "loss": 1.2039,
+      "step": 8215
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 31.453182996449176,
+      "learning_rate": 9.083216239618831e-06,
+      "loss": 1.2333,
+      "step": 8220
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.40557681087287,
+      "learning_rate": 9.081522778626022e-06,
+      "loss": 1.2445,
+      "step": 8225
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 41.965222999552864,
+      "learning_rate": 9.079827913185126e-06,
+      "loss": 1.2503,
+      "step": 8230
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 33.22209321149795,
+      "learning_rate": 9.07813164387934e-06,
+      "loss": 1.2313,
+      "step": 8235
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 81.53473188534082,
+      "learning_rate": 9.07643397129235e-06,
+      "loss": 1.2446,
+      "step": 8240
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 120.40449446440331,
+      "learning_rate": 9.074734896008326e-06,
+      "loss": 1.2962,
+      "step": 8245
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 65.36493591941586,
+      "learning_rate": 9.073034418611915e-06,
+      "loss": 1.2235,
+      "step": 8250
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 40.4010764969845,
+      "learning_rate": 9.071332539688248e-06,
+      "loss": 1.2348,
+      "step": 8255
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.64350332332248,
+      "learning_rate": 9.069629259822947e-06,
+      "loss": 1.2535,
+      "step": 8260
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 46.141321205756675,
+      "learning_rate": 9.067924579602102e-06,
+      "loss": 1.2254,
+      "step": 8265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 22.800453773851473,
+      "learning_rate": 9.066218499612296e-06,
+      "loss": 1.2383,
+      "step": 8270
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 32.825239418145735,
+      "learning_rate": 9.064511020440587e-06,
+      "loss": 1.1774,
+      "step": 8275
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.849725814097734,
+      "learning_rate": 9.062802142674519e-06,
+      "loss": 1.2256,
+      "step": 8280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 36.209761027309284,
+      "learning_rate": 9.061091866902112e-06,
+      "loss": 1.171,
+      "step": 8285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 25.1646410223101,
+      "learning_rate": 9.059380193711873e-06,
+      "loss": 1.277,
+      "step": 8290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.127930781671427,
+      "learning_rate": 9.057667123692788e-06,
+      "loss": 1.2592,
+      "step": 8295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.664642683604065,
+      "learning_rate": 9.05595265743432e-06,
+      "loss": 1.2532,
+      "step": 8300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.241837976622543,
+      "learning_rate": 9.054236795526416e-06,
+      "loss": 1.2686,
+      "step": 8305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.854922348067305,
+      "learning_rate": 9.052519538559505e-06,
+      "loss": 1.247,
+      "step": 8310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.39256460774842,
+      "learning_rate": 9.050800887124492e-06,
+      "loss": 1.2109,
+      "step": 8315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.011338239652282,
+      "learning_rate": 9.04908084181276e-06,
+      "loss": 1.242,
+      "step": 8320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.290562496694669,
+      "learning_rate": 9.047359403216177e-06,
+      "loss": 1.2043,
+      "step": 8325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 20.502784890291515,
+      "learning_rate": 9.04563657192709e-06,
+      "loss": 1.2437,
+      "step": 8330
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.85465910133973,
+      "learning_rate": 9.043912348538324e-06,
+      "loss": 1.2447,
+      "step": 8335
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 21.084864859510795,
+      "learning_rate": 9.04218673364318e-06,
+      "loss": 1.2518,
+      "step": 8340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.3528319487047,
+      "learning_rate": 9.040459727835442e-06,
+      "loss": 1.1677,
+      "step": 8345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.499465506508274,
+      "learning_rate": 9.038731331709371e-06,
+      "loss": 1.2745,
+      "step": 8350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 22.8566905996232,
+      "learning_rate": 9.037001545859706e-06,
+      "loss": 1.226,
+      "step": 8355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 20.300693565803126,
+      "learning_rate": 9.035270370881666e-06,
+      "loss": 1.1771,
+      "step": 8360
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 22.637425395129974,
+      "learning_rate": 9.033537807370943e-06,
+      "loss": 1.1876,
+      "step": 8365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 91.96974249962282,
+      "learning_rate": 9.031803855923715e-06,
+      "loss": 1.2614,
+      "step": 8370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 69.88547886649809,
+      "learning_rate": 9.03006851713663e-06,
+      "loss": 1.2672,
+      "step": 8375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.499419152260945,
+      "learning_rate": 9.028331791606819e-06,
+      "loss": 1.1971,
+      "step": 8380
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 18.395960923241695,
+      "learning_rate": 9.026593679931885e-06,
+      "loss": 1.2513,
+      "step": 8385
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.08015527528984,
+      "learning_rate": 9.02485418270991e-06,
+      "loss": 1.2314,
+      "step": 8390
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 23.025894946322406,
+      "learning_rate": 9.023113300539457e-06,
+      "loss": 1.2254,
+      "step": 8395
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 93.79636474761914,
+      "learning_rate": 9.021371034019559e-06,
+      "loss": 1.2282,
+      "step": 8400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 28.079331258559648,
+      "learning_rate": 9.019627383749728e-06,
+      "loss": 1.2048,
+      "step": 8405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 60.348869043870465,
+      "learning_rate": 9.017882350329955e-06,
+      "loss": 1.3066,
+      "step": 8410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 33.47748076849988,
+      "learning_rate": 9.016135934360703e-06,
+      "loss": 1.2841,
+      "step": 8415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 32.28164414314165,
+      "learning_rate": 9.014388136442912e-06,
+      "loss": 1.2336,
+      "step": 8420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.344102265445358,
+      "learning_rate": 9.012638957177994e-06,
+      "loss": 1.2027,
+      "step": 8425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.961658100350237,
+      "learning_rate": 9.010888397167848e-06,
+      "loss": 1.227,
+      "step": 8430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 30.993547565692637,
+      "learning_rate": 9.009136457014833e-06,
+      "loss": 1.1944,
+      "step": 8435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.533667423823333,
+      "learning_rate": 9.007383137321793e-06,
+      "loss": 1.1939,
+      "step": 8440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 20.641505087389255,
+      "learning_rate": 9.005628438692042e-06,
+      "loss": 1.1807,
+      "step": 8445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 27.084097419574157,
+      "learning_rate": 9.003872361729371e-06,
+      "loss": 1.2487,
+      "step": 8450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 55.487939119016765,
+      "learning_rate": 9.002114907038045e-06,
+      "loss": 1.2511,
+      "step": 8455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 38.647380685218614,
+      "learning_rate": 9.000356075222803e-06,
+      "loss": 1.2469,
+      "step": 8460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 18.655068951279034,
+      "learning_rate": 8.998595866888855e-06,
+      "loss": 1.2072,
+      "step": 8465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.370284045382212,
+      "learning_rate": 8.996834282641889e-06,
+      "loss": 1.2035,
+      "step": 8470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.700022848670333,
+      "learning_rate": 8.995071323088063e-06,
+      "loss": 1.1962,
+      "step": 8475
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 19.48844904276957,
+      "learning_rate": 8.99330698883401e-06,
+      "loss": 1.2056,
+      "step": 8480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 20.600346419580166,
+      "learning_rate": 8.991541280486838e-06,
+      "loss": 1.2016,
+      "step": 8485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.50302243666553,
+      "learning_rate": 8.989774198654123e-06,
+      "loss": 1.2685,
+      "step": 8490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.320021532955366,
+      "learning_rate": 8.988005743943916e-06,
+      "loss": 1.228,
+      "step": 8495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.707022281835492,
+      "learning_rate": 8.986235916964742e-06,
+      "loss": 1.2113,
+      "step": 8500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.868031314799666,
+      "learning_rate": 8.984464718325596e-06,
+      "loss": 1.2128,
+      "step": 8505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.321884298163736,
+      "learning_rate": 8.982692148635947e-06,
+      "loss": 1.1707,
+      "step": 8510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.031382634589063,
+      "learning_rate": 8.980918208505734e-06,
+      "loss": 1.1996,
+      "step": 8515
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.593011499567609,
+      "learning_rate": 8.979142898545366e-06,
+      "loss": 1.2297,
+      "step": 8520
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 17.85682476258015,
+      "learning_rate": 8.977366219365728e-06,
+      "loss": 1.2096,
+      "step": 8525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 17.337151667033524,
+      "learning_rate": 8.975588171578172e-06,
+      "loss": 1.2184,
+      "step": 8530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 19.574082092030107,
+      "learning_rate": 8.973808755794525e-06,
+      "loss": 1.2287,
+      "step": 8535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.340481711809604,
+      "learning_rate": 8.972027972627078e-06,
+      "loss": 1.1706,
+      "step": 8540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.560002968647215,
+      "learning_rate": 8.970245822688601e-06,
+      "loss": 1.2048,
+      "step": 8545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.707635187199964,
+      "learning_rate": 8.968462306592328e-06,
+      "loss": 1.195,
+      "step": 8550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.796462917778843,
+      "learning_rate": 8.966677424951966e-06,
+      "loss": 1.168,
+      "step": 8555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.260137171097879,
+      "learning_rate": 8.964891178381691e-06,
+      "loss": 1.2253,
+      "step": 8560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 25.65853338757763,
+      "learning_rate": 8.963103567496148e-06,
+      "loss": 1.344,
+      "step": 8565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.153776235558424,
+      "learning_rate": 8.961314592910452e-06,
+      "loss": 1.1815,
+      "step": 8570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 21.907940778880246,
+      "learning_rate": 8.959524255240189e-06,
+      "loss": 1.2651,
+      "step": 8575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 22.521581700435902,
+      "learning_rate": 8.95773255510141e-06,
+      "loss": 1.1989,
+      "step": 8580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.53693385733085,
+      "learning_rate": 8.95593949311064e-06,
+      "loss": 1.1989,
+      "step": 8585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 34.65238445292161,
+      "learning_rate": 8.95414506988487e-06,
+      "loss": 1.2189,
+      "step": 8590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.09106999665172,
+      "learning_rate": 8.952349286041556e-06,
+      "loss": 1.2075,
+      "step": 8595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 46.468132754859965,
+      "learning_rate": 8.950552142198628e-06,
+      "loss": 1.1656,
+      "step": 8600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 26.328771913319503,
+      "learning_rate": 8.948753638974482e-06,
+      "loss": 1.2187,
+      "step": 8605
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.830332243757532,
+      "learning_rate": 8.94695377698798e-06,
+      "loss": 1.2091,
+      "step": 8610
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 53.041771160912965,
+      "learning_rate": 8.945152556858452e-06,
+      "loss": 1.2352,
+      "step": 8615
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 49.80951360537134,
+      "learning_rate": 8.9433499792057e-06,
+      "loss": 1.262,
+      "step": 8620
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 95.03977633541976,
+      "learning_rate": 8.941546044649985e-06,
+      "loss": 1.2122,
+      "step": 8625
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.152155839898636,
+      "learning_rate": 8.93974075381204e-06,
+      "loss": 1.1913,
+      "step": 8630
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.884864637091642,
+      "learning_rate": 8.937934107313065e-06,
+      "loss": 1.1907,
+      "step": 8635
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 21.758968553236418,
+      "learning_rate": 8.936126105774725e-06,
+      "loss": 1.2776,
+      "step": 8640
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 65.77897117650656,
+      "learning_rate": 8.93431674981915e-06,
+      "loss": 1.2564,
+      "step": 8645
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 37.48469408393138,
+      "learning_rate": 8.93250604006894e-06,
+      "loss": 1.2701,
+      "step": 8650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 20.5819566041243,
+      "learning_rate": 8.930693977147157e-06,
+      "loss": 1.2082,
+      "step": 8655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.838012542117974,
+      "learning_rate": 8.928880561677333e-06,
+      "loss": 1.2151,
+      "step": 8660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 28.934816659757903,
+      "learning_rate": 8.927065794283458e-06,
+      "loss": 1.2108,
+      "step": 8665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 30.015851158231182,
+      "learning_rate": 8.925249675589995e-06,
+      "loss": 1.207,
+      "step": 8670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.05708344697593,
+      "learning_rate": 8.923432206221868e-06,
+      "loss": 1.2508,
+      "step": 8675
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.071404346946036,
+      "learning_rate": 8.921613386804465e-06,
+      "loss": 1.25,
+      "step": 8680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.1873966208491,
+      "learning_rate": 8.919793217963642e-06,
+      "loss": 1.1895,
+      "step": 8685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 86.5409264169609,
+      "learning_rate": 8.917971700325717e-06,
+      "loss": 1.2177,
+      "step": 8690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 48.8292154136072,
+      "learning_rate": 8.91614883451747e-06,
+      "loss": 1.2553,
+      "step": 8695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.72765388759314,
+      "learning_rate": 8.914324621166151e-06,
+      "loss": 1.2569,
+      "step": 8700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.99381671321517,
+      "learning_rate": 8.912499060899469e-06,
+      "loss": 1.2478,
+      "step": 8705
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 23.038501634929837,
+      "learning_rate": 8.910672154345596e-06,
+      "loss": 1.2592,
+      "step": 8710
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 86.91259519907479,
+      "learning_rate": 8.90884390213317e-06,
+      "loss": 1.2319,
+      "step": 8715
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 72.00559372046911,
+      "learning_rate": 8.907014304891291e-06,
+      "loss": 1.1971,
+      "step": 8720
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 52.695079478540215,
+      "learning_rate": 8.905183363249521e-06,
+      "loss": 1.2671,
+      "step": 8725
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.294026987914467,
+      "learning_rate": 8.903351077837885e-06,
+      "loss": 1.2787,
+      "step": 8730
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 26.77369152335356,
+      "learning_rate": 8.901517449286871e-06,
+      "loss": 1.241,
+      "step": 8735
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 50.896197869777964,
+      "learning_rate": 8.89968247822743e-06,
+      "loss": 1.2331,
+      "step": 8740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 31.150932876356467,
+      "learning_rate": 8.897846165290974e-06,
+      "loss": 1.2359,
+      "step": 8745
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.45636798929398,
+      "learning_rate": 8.896008511109373e-06,
+      "loss": 1.2213,
+      "step": 8750
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 34.80863089297944,
+      "learning_rate": 8.894169516314966e-06,
+      "loss": 1.2084,
+      "step": 8755
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.824283701844934,
+      "learning_rate": 8.892329181540547e-06,
+      "loss": 1.2012,
+      "step": 8760
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 20.875788115200002,
+      "learning_rate": 8.890487507419375e-06,
+      "loss": 1.1967,
+      "step": 8765
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.419811862747704,
+      "learning_rate": 8.888644494585167e-06,
+      "loss": 1.2338,
+      "step": 8770
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.537355999648662,
+      "learning_rate": 8.886800143672103e-06,
+      "loss": 1.2357,
+      "step": 8775
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.75738643127998,
+      "learning_rate": 8.884954455314822e-06,
+      "loss": 1.2745,
+      "step": 8780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.229991642182542,
+      "learning_rate": 8.883107430148422e-06,
+      "loss": 1.2192,
+      "step": 8785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.311212308408923,
+      "learning_rate": 8.881259068808466e-06,
+      "loss": 1.239,
+      "step": 8790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.739353701546275,
+      "learning_rate": 8.879409371930969e-06,
+      "loss": 1.1639,
+      "step": 8795
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 20.381728011219263,
+      "learning_rate": 8.877558340152414e-06,
+      "loss": 1.2314,
+      "step": 8800
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.386860710130472,
+      "learning_rate": 8.875705974109738e-06,
+      "loss": 1.2133,
+      "step": 8805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 19.316580467726308,
+      "learning_rate": 8.873852274440337e-06,
+      "loss": 1.234,
+      "step": 8810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.727198524509611,
+      "learning_rate": 8.87199724178207e-06,
+      "loss": 1.2077,
+      "step": 8815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.958332741842913,
+      "learning_rate": 8.87014087677325e-06,
+      "loss": 1.2547,
+      "step": 8820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.067840849747313,
+      "learning_rate": 8.868283180052648e-06,
+      "loss": 1.227,
+      "step": 8825
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 23.86742366801994,
+      "learning_rate": 8.866424152259501e-06,
+      "loss": 1.2002,
+      "step": 8830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.579342887032047,
+      "learning_rate": 8.864563794033496e-06,
+      "loss": 1.226,
+      "step": 8835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 32.68986457966961,
+      "learning_rate": 8.86270210601478e-06,
+      "loss": 1.2414,
+      "step": 8840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.087093156037234,
+      "learning_rate": 8.860839088843958e-06,
+      "loss": 1.2746,
+      "step": 8845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 22.212447048092486,
+      "learning_rate": 8.858974743162095e-06,
+      "loss": 1.2399,
+      "step": 8850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.496678463979974,
+      "learning_rate": 8.857109069610706e-06,
+      "loss": 1.1913,
+      "step": 8855
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.26558139753938,
+      "learning_rate": 8.85524206883177e-06,
+      "loss": 1.2259,
+      "step": 8860
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.567739935203726,
+      "learning_rate": 8.853373741467724e-06,
+      "loss": 1.2014,
+      "step": 8865
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.243019396023581,
+      "learning_rate": 8.85150408816145e-06,
+      "loss": 1.1706,
+      "step": 8870
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.90760732651893,
+      "learning_rate": 8.8496331095563e-06,
+      "loss": 1.231,
+      "step": 8875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.272240575737805,
+      "learning_rate": 8.847760806296072e-06,
+      "loss": 1.1923,
+      "step": 8880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.876889907445083,
+      "learning_rate": 8.845887179025029e-06,
+      "loss": 1.2276,
+      "step": 8885
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.256098634241363,
+      "learning_rate": 8.844012228387879e-06,
+      "loss": 1.2183,
+      "step": 8890
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.472058437265845,
+      "learning_rate": 8.84213595502979e-06,
+      "loss": 1.2006,
+      "step": 8895
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.871714887277207,
+      "learning_rate": 8.840258359596392e-06,
+      "loss": 1.2144,
+      "step": 8900
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.80383694900924,
+      "learning_rate": 8.83837944273376e-06,
+      "loss": 1.2478,
+      "step": 8905
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.231795176525273,
+      "learning_rate": 8.836499205088426e-06,
+      "loss": 1.1868,
+      "step": 8910
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.804885651323739,
+      "learning_rate": 8.834617647307383e-06,
+      "loss": 1.24,
+      "step": 8915
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.386285962788818,
+      "learning_rate": 8.832734770038068e-06,
+      "loss": 1.2363,
+      "step": 8920
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.478128391048558,
+      "learning_rate": 8.830850573928378e-06,
+      "loss": 1.2555,
+      "step": 8925
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.276309193456074,
+      "learning_rate": 8.828965059626666e-06,
+      "loss": 1.2382,
+      "step": 8930
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 20.538374321553064,
+      "learning_rate": 8.827078227781734e-06,
+      "loss": 1.2285,
+      "step": 8935
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.22973284606407,
+      "learning_rate": 8.825190079042839e-06,
+      "loss": 1.1935,
+      "step": 8940
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.121712987282244,
+      "learning_rate": 8.823300614059692e-06,
+      "loss": 1.2082,
+      "step": 8945
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 18.410891172450526,
+      "learning_rate": 8.821409833482454e-06,
+      "loss": 1.1824,
+      "step": 8950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 23.697063264981985,
+      "learning_rate": 8.819517737961742e-06,
+      "loss": 1.2121,
+      "step": 8955
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 22.61499589372635,
+      "learning_rate": 8.817624328148627e-06,
+      "loss": 1.2562,
+      "step": 8960
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 62.12866389191985,
+      "learning_rate": 8.815729604694624e-06,
+      "loss": 1.24,
+      "step": 8965
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 15.910338303371752,
+      "learning_rate": 8.813833568251708e-06,
+      "loss": 1.2244,
+      "step": 8970
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 40.97912228130415,
+      "learning_rate": 8.811936219472306e-06,
+      "loss": 1.2409,
+      "step": 8975
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 185.10887018584373,
+      "learning_rate": 8.810037559009292e-06,
+      "loss": 1.2355,
+      "step": 8980
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 124.19054332408055,
+      "learning_rate": 8.808137587515992e-06,
+      "loss": 1.2855,
+      "step": 8985
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 68.21812846024108,
+      "learning_rate": 8.806236305646185e-06,
+      "loss": 1.3029,
+      "step": 8990
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 26.761045112155376,
+      "learning_rate": 8.8043337140541e-06,
+      "loss": 1.3163,
+      "step": 8995
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 49.033990563804736,
+      "learning_rate": 8.80242981339442e-06,
+      "loss": 1.2956,
+      "step": 9000
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 21.560566087023556,
+      "learning_rate": 8.800524604322272e-06,
+      "loss": 1.2232,
+      "step": 9005
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.567735923502728,
+      "learning_rate": 8.798618087493237e-06,
+      "loss": 1.2081,
+      "step": 9010
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.993297853077632,
+      "learning_rate": 8.796710263563345e-06,
+      "loss": 1.2857,
+      "step": 9015
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.15594183914007,
+      "learning_rate": 8.794801133189079e-06,
+      "loss": 1.2217,
+      "step": 9020
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.100897474881501,
+      "learning_rate": 8.792890697027367e-06,
+      "loss": 1.1967,
+      "step": 9025
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.349787053381577,
+      "learning_rate": 8.79097895573559e-06,
+      "loss": 1.2492,
+      "step": 9030
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 54.46800846028436,
+      "learning_rate": 8.789065909971574e-06,
+      "loss": 1.2935,
+      "step": 9035
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 58.03382374458775,
+      "learning_rate": 8.787151560393597e-06,
+      "loss": 1.2109,
+      "step": 9040
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.560042500814026,
+      "learning_rate": 8.785235907660385e-06,
+      "loss": 1.2293,
+      "step": 9045
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.94549857333933,
+      "learning_rate": 8.783318952431114e-06,
+      "loss": 1.2127,
+      "step": 9050
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 19.386889362913447,
+      "learning_rate": 8.781400695365405e-06,
+      "loss": 1.235,
+      "step": 9055
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.863732384837983,
+      "learning_rate": 8.779481137123327e-06,
+      "loss": 1.2258,
+      "step": 9060
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 34.821485508785,
+      "learning_rate": 8.7775602783654e-06,
+      "loss": 1.2098,
+      "step": 9065
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 21.310302561173884,
+      "learning_rate": 8.77563811975259e-06,
+      "loss": 1.2393,
+      "step": 9070
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.004806784226558,
+      "learning_rate": 8.77371466194631e-06,
+      "loss": 1.2415,
+      "step": 9075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 23.93245509241067,
+      "learning_rate": 8.771789905608416e-06,
+      "loss": 1.2573,
+      "step": 9080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 25.861977789777615,
+      "learning_rate": 8.769863851401219e-06,
+      "loss": 1.2743,
+      "step": 9085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.755736619167987,
+      "learning_rate": 8.767936499987473e-06,
+      "loss": 1.2376,
+      "step": 9090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.211424345701793,
+      "learning_rate": 8.766007852030373e-06,
+      "loss": 1.2578,
+      "step": 9095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.148920549885833,
+      "learning_rate": 8.764077908193571e-06,
+      "loss": 1.1645,
+      "step": 9100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.930321153879788,
+      "learning_rate": 8.762146669141156e-06,
+      "loss": 1.1894,
+      "step": 9105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.029191550269419,
+      "learning_rate": 8.760214135537663e-06,
+      "loss": 1.1963,
+      "step": 9110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.684012600037251,
+      "learning_rate": 8.758280308048079e-06,
+      "loss": 1.2488,
+      "step": 9115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 25.043411822630535,
+      "learning_rate": 8.75634518733783e-06,
+      "loss": 1.2259,
+      "step": 9120
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 20.978717883249782,
+      "learning_rate": 8.754408774072791e-06,
+      "loss": 1.2128,
+      "step": 9125
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.364554753385073,
+      "learning_rate": 8.752471068919277e-06,
+      "loss": 1.274,
+      "step": 9130
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.852807191673204,
+      "learning_rate": 8.750532072544053e-06,
+      "loss": 1.269,
+      "step": 9135
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.0768275552660755,
+      "learning_rate": 8.748591785614327e-06,
+      "loss": 1.2448,
+      "step": 9140
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.895062035573548,
+      "learning_rate": 8.746650208797745e-06,
+      "loss": 1.237,
+      "step": 9145
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 18.688327529636332,
+      "learning_rate": 8.744707342762406e-06,
+      "loss": 1.2442,
+      "step": 9150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.773096278719505,
+      "learning_rate": 8.742763188176845e-06,
+      "loss": 1.2289,
+      "step": 9155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.539377156107374,
+      "learning_rate": 8.740817745710049e-06,
+      "loss": 1.2787,
+      "step": 9160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.669054910974296,
+      "learning_rate": 8.738871016031438e-06,
+      "loss": 1.1643,
+      "step": 9165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.535314129564306,
+      "learning_rate": 8.73692299981088e-06,
+      "loss": 1.2147,
+      "step": 9170
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.407433647498465,
+      "learning_rate": 8.734973697718689e-06,
+      "loss": 1.241,
+      "step": 9175
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.41579698367922,
+      "learning_rate": 8.733023110425616e-06,
+      "loss": 1.2248,
+      "step": 9180
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.82990665062719,
+      "learning_rate": 8.731071238602855e-06,
+      "loss": 1.2382,
+      "step": 9185
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.384785763412171,
+      "learning_rate": 8.729118082922044e-06,
+      "loss": 1.2981,
+      "step": 9190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.22399926884024,
+      "learning_rate": 8.727163644055263e-06,
+      "loss": 1.2077,
+      "step": 9195
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.248519938538731,
+      "learning_rate": 8.725207922675032e-06,
+      "loss": 1.1737,
+      "step": 9200
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 20.2802850223925,
+      "learning_rate": 8.723250919454313e-06,
+      "loss": 1.2105,
+      "step": 9205
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.164284527748045,
+      "learning_rate": 8.721292635066509e-06,
+      "loss": 1.1661,
+      "step": 9210
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.118906619701703,
+      "learning_rate": 8.719333070185463e-06,
+      "loss": 1.1894,
+      "step": 9215
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.555578503418722,
+      "learning_rate": 8.71737222548546e-06,
+      "loss": 1.2029,
+      "step": 9220
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.287993256735366,
+      "learning_rate": 8.715410101641225e-06,
+      "loss": 1.239,
+      "step": 9225
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.616509952185027,
+      "learning_rate": 8.71344669932792e-06,
+      "loss": 1.2277,
+      "step": 9230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.266653583169361,
+      "learning_rate": 8.711482019221157e-06,
+      "loss": 1.2479,
+      "step": 9235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.831512190806212,
+      "learning_rate": 8.709516061996973e-06,
+      "loss": 1.2383,
+      "step": 9240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.612810498848723,
+      "learning_rate": 8.707548828331856e-06,
+      "loss": 1.1593,
+      "step": 9245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.042277080028542,
+      "learning_rate": 8.705580318902728e-06,
+      "loss": 1.2517,
+      "step": 9250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.223576413834543,
+      "learning_rate": 8.703610534386952e-06,
+      "loss": 1.2395,
+      "step": 9255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.846067819244304,
+      "learning_rate": 8.70163947546233e-06,
+      "loss": 1.1662,
+      "step": 9260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 19.750262940528504,
+      "learning_rate": 8.699667142807096e-06,
+      "loss": 1.2339,
+      "step": 9265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.201453699126599,
+      "learning_rate": 8.697693537099935e-06,
+      "loss": 1.218,
+      "step": 9270
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.547911380471197,
+      "learning_rate": 8.695718659019957e-06,
+      "loss": 1.2112,
+      "step": 9275
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.642260834594786,
+      "learning_rate": 8.69374250924672e-06,
+      "loss": 1.2476,
+      "step": 9280
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.92848949067963,
+      "learning_rate": 8.691765088460214e-06,
+      "loss": 1.193,
+      "step": 9285
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 27.8880184994152,
+      "learning_rate": 8.689786397340866e-06,
+      "loss": 1.2467,
+      "step": 9290
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 33.71047112323762,
+      "learning_rate": 8.687806436569544e-06,
+      "loss": 1.2282,
+      "step": 9295
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 30.28245393486485,
+      "learning_rate": 8.685825206827549e-06,
+      "loss": 1.1944,
+      "step": 9300
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.01931733205294,
+      "learning_rate": 8.68384270879662e-06,
+      "loss": 1.1915,
+      "step": 9305
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.301290633165818,
+      "learning_rate": 8.681858943158934e-06,
+      "loss": 1.218,
+      "step": 9310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 59.894527194127896,
+      "learning_rate": 8.6798739105971e-06,
+      "loss": 1.2276,
+      "step": 9315
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 60.59704754802048,
+      "learning_rate": 8.677887611794171e-06,
+      "loss": 1.2136,
+      "step": 9320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 31.229948653849934,
+      "learning_rate": 8.675900047433628e-06,
+      "loss": 1.2148,
+      "step": 9325
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.751502635285522,
+      "learning_rate": 8.673911218199387e-06,
+      "loss": 1.2416,
+      "step": 9330
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 27.156635876069526,
+      "learning_rate": 8.671921124775807e-06,
+      "loss": 1.2311,
+      "step": 9335
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 28.13243527055898,
+      "learning_rate": 8.669929767847673e-06,
+      "loss": 1.2309,
+      "step": 9340
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 21.92555759390989,
+      "learning_rate": 8.667937148100211e-06,
+      "loss": 1.2145,
+      "step": 9345
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.51310034385813,
+      "learning_rate": 8.665943266219081e-06,
+      "loss": 1.273,
+      "step": 9350
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 60.47307916368799,
+      "learning_rate": 8.663948122890376e-06,
+      "loss": 1.2216,
+      "step": 9355
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 27.728669094488204,
+      "learning_rate": 8.661951718800618e-06,
+      "loss": 1.2355,
+      "step": 9360
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.484033499295984,
+      "learning_rate": 8.659954054636774e-06,
+      "loss": 1.2516,
+      "step": 9365
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.625017951128116,
+      "learning_rate": 8.657955131086234e-06,
+      "loss": 1.1943,
+      "step": 9370
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.072922518170913,
+      "learning_rate": 8.655954948836826e-06,
+      "loss": 1.2521,
+      "step": 9375
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 20.147834103743097,
+      "learning_rate": 8.653953508576813e-06,
+      "loss": 1.2457,
+      "step": 9380
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.187688426230986,
+      "learning_rate": 8.65195081099489e-06,
+      "loss": 1.2233,
+      "step": 9385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.52875596393606,
+      "learning_rate": 8.649946856780178e-06,
+      "loss": 1.1896,
+      "step": 9390
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.634859897158842,
+      "learning_rate": 8.64794164662224e-06,
+      "loss": 1.2319,
+      "step": 9395
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.724500236711942,
+      "learning_rate": 8.645935181211065e-06,
+      "loss": 1.185,
+      "step": 9400
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.001379200664274,
+      "learning_rate": 8.643927461237076e-06,
+      "loss": 1.1959,
+      "step": 9405
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 21.132559590260563,
+      "learning_rate": 8.641918487391129e-06,
+      "loss": 1.2037,
+      "step": 9410
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.29694004153204,
+      "learning_rate": 8.639908260364506e-06,
+      "loss": 1.2135,
+      "step": 9415
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.94645524463928,
+      "learning_rate": 8.637896780848932e-06,
+      "loss": 1.1843,
+      "step": 9420
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 33.47190874015566,
+      "learning_rate": 8.635884049536548e-06,
+      "loss": 1.2311,
+      "step": 9425
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 23.42170464750877,
+      "learning_rate": 8.633870067119934e-06,
+      "loss": 1.2217,
+      "step": 9430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.017418447170066,
+      "learning_rate": 8.631854834292102e-06,
+      "loss": 1.2402,
+      "step": 9435
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.972512670568745,
+      "learning_rate": 8.629838351746488e-06,
+      "loss": 1.1981,
+      "step": 9440
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.461769001581415,
+      "learning_rate": 8.627820620176967e-06,
+      "loss": 1.2776,
+      "step": 9445
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.786493970019889,
+      "learning_rate": 8.625801640277835e-06,
+      "loss": 1.1868,
+      "step": 9450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 29.018649268371522,
+      "learning_rate": 8.62378141274382e-06,
+      "loss": 1.1648,
+      "step": 9455
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 45.83096642377824,
+      "learning_rate": 8.621759938270085e-06,
+      "loss": 1.204,
+      "step": 9460
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 37.56036797387887,
+      "learning_rate": 8.619737217552213e-06,
+      "loss": 1.2293,
+      "step": 9465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.85963464956369,
+      "learning_rate": 8.617713251286221e-06,
+      "loss": 1.2041,
+      "step": 9470
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 28.08967115530905,
+      "learning_rate": 8.615688040168554e-06,
+      "loss": 1.2017,
+      "step": 9475
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.488225184800497,
+      "learning_rate": 8.613661584896088e-06,
+      "loss": 1.1837,
+      "step": 9480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.232954988608356,
+      "learning_rate": 8.61163388616612e-06,
+      "loss": 1.2455,
+      "step": 9485
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.295162473894017,
+      "learning_rate": 8.609604944676382e-06,
+      "loss": 1.1843,
+      "step": 9490
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.202453090548019,
+      "learning_rate": 8.607574761125029e-06,
+      "loss": 1.1921,
+      "step": 9495
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.589082214759323,
+      "learning_rate": 8.605543336210648e-06,
+      "loss": 1.2447,
+      "step": 9500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 21.868788541707016,
+      "learning_rate": 8.603510670632248e-06,
+      "loss": 1.1866,
+      "step": 9505
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 15.43176697786205,
+      "learning_rate": 8.601476765089267e-06,
+      "loss": 1.2228,
+      "step": 9510
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 25.64496682281605,
+      "learning_rate": 8.599441620281573e-06,
+      "loss": 1.188,
+      "step": 9515
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 29.764310160845312,
+      "learning_rate": 8.597405236909454e-06,
+      "loss": 1.2362,
+      "step": 9520
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 32.23508806917358,
+      "learning_rate": 8.59536761567363e-06,
+      "loss": 1.1883,
+      "step": 9525
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.162439336224594,
+      "learning_rate": 8.593328757275244e-06,
+      "loss": 1.2109,
+      "step": 9530
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.8497243633031735,
+      "learning_rate": 8.591288662415863e-06,
+      "loss": 1.1849,
+      "step": 9535
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.325698207035451,
+      "learning_rate": 8.589247331797482e-06,
+      "loss": 1.2294,
+      "step": 9540
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.684046114236223,
+      "learning_rate": 8.587204766122524e-06,
+      "loss": 1.2249,
+      "step": 9545
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.780997127520526,
+      "learning_rate": 8.585160966093832e-06,
+      "loss": 1.1899,
+      "step": 9550
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.076217961018143,
+      "learning_rate": 8.583115932414677e-06,
+      "loss": 1.1961,
+      "step": 9555
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.558273083924005,
+      "learning_rate": 8.58106966578875e-06,
+      "loss": 1.1607,
+      "step": 9560
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.86979617003217,
+      "learning_rate": 8.579022166920172e-06,
+      "loss": 1.2633,
+      "step": 9565
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 25.659561988607937,
+      "learning_rate": 8.576973436513485e-06,
+      "loss": 1.1857,
+      "step": 9570
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 46.048303619796755,
+      "learning_rate": 8.574923475273656e-06,
+      "loss": 1.2248,
+      "step": 9575
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.727155163521066,
+      "learning_rate": 8.572872283906071e-06,
+      "loss": 1.2353,
+      "step": 9580
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.94732184496867,
+      "learning_rate": 8.570819863116546e-06,
+      "loss": 1.1864,
+      "step": 9585
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 20.209059290959434,
+      "learning_rate": 8.56876621361132e-06,
+      "loss": 1.2085,
+      "step": 9590
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.880908707190672,
+      "learning_rate": 8.566711336097045e-06,
+      "loss": 1.2717,
+      "step": 9595
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.070944510809447,
+      "learning_rate": 8.564655231280808e-06,
+      "loss": 1.2236,
+      "step": 9600
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 34.890685751008064,
+      "learning_rate": 8.562597899870107e-06,
+      "loss": 1.1806,
+      "step": 9605
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.757711801324973,
+      "learning_rate": 8.560539342572874e-06,
+      "loss": 1.1937,
+      "step": 9610
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 26.934656558915368,
+      "learning_rate": 8.558479560097455e-06,
+      "loss": 1.2282,
+      "step": 9615
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.051031400680982,
+      "learning_rate": 8.556418553152617e-06,
+      "loss": 1.199,
+      "step": 9620
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.064718410605512,
+      "learning_rate": 8.55435632244755e-06,
+      "loss": 1.1695,
+      "step": 9625
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.585082014770853,
+      "learning_rate": 8.552292868691869e-06,
+      "loss": 1.184,
+      "step": 9630
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.378246934152191,
+      "learning_rate": 8.550228192595604e-06,
+      "loss": 1.2185,
+      "step": 9635
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.97285764971995,
+      "learning_rate": 8.54816229486921e-06,
+      "loss": 1.195,
+      "step": 9640
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.008911383877557,
+      "learning_rate": 8.546095176223559e-06,
+      "loss": 1.1942,
+      "step": 9645
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.92183947353392,
+      "learning_rate": 8.544026837369944e-06,
+      "loss": 1.2008,
+      "step": 9650
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.115687819170043,
+      "learning_rate": 8.54195727902008e-06,
+      "loss": 1.2601,
+      "step": 9655
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.63085545181519,
+      "learning_rate": 8.539886501886096e-06,
+      "loss": 1.245,
+      "step": 9660
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.0638585143837895,
+      "learning_rate": 8.537814506680552e-06,
+      "loss": 1.2002,
+      "step": 9665
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.479375704581546,
+      "learning_rate": 8.535741294116413e-06,
+      "loss": 1.2232,
+      "step": 9670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.557551719070158,
+      "learning_rate": 8.533666864907072e-06,
+      "loss": 1.1973,
+      "step": 9675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.673967352600394,
+      "learning_rate": 8.531591219766337e-06,
+      "loss": 1.2523,
+      "step": 9680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.669799825253532,
+      "learning_rate": 8.52951435940844e-06,
+      "loss": 1.2175,
+      "step": 9685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.04964319057395,
+      "learning_rate": 8.527436284548019e-06,
+      "loss": 1.1941,
+      "step": 9690
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.990033812811166,
+      "learning_rate": 8.525356995900142e-06,
+      "loss": 1.1702,
+      "step": 9695
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.20945799802136,
+      "learning_rate": 8.523276494180293e-06,
+      "loss": 1.1931,
+      "step": 9700
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.84362195949145,
+      "learning_rate": 8.521194780104364e-06,
+      "loss": 1.2362,
+      "step": 9705
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.616349187919386,
+      "learning_rate": 8.519111854388676e-06,
+      "loss": 1.1801,
+      "step": 9710
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.920360038100625,
+      "learning_rate": 8.51702771774996e-06,
+      "loss": 1.2,
+      "step": 9715
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 31.170767115548916,
+      "learning_rate": 8.514942370905364e-06,
+      "loss": 1.2017,
+      "step": 9720
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 24.598565989478793,
+      "learning_rate": 8.512855814572456e-06,
+      "loss": 1.1829,
+      "step": 9725
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.275465737808293,
+      "learning_rate": 8.510768049469218e-06,
+      "loss": 1.2021,
+      "step": 9730
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 72.26265395071671,
+      "learning_rate": 8.508679076314048e-06,
+      "loss": 1.2837,
+      "step": 9735
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 31.72169190093813,
+      "learning_rate": 8.506588895825758e-06,
+      "loss": 1.2686,
+      "step": 9740
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 176.50463584566126,
+      "learning_rate": 8.504497508723578e-06,
+      "loss": 1.2294,
+      "step": 9745
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 58.2525129534981,
+      "learning_rate": 8.502404915727154e-06,
+      "loss": 1.2327,
+      "step": 9750
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 44.345100118213004,
+      "learning_rate": 8.50031111755654e-06,
+      "loss": 1.2719,
+      "step": 9755
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 58.80090588929816,
+      "learning_rate": 8.498216114932217e-06,
+      "loss": 1.225,
+      "step": 9760
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.63017726386444,
+      "learning_rate": 8.49611990857507e-06,
+      "loss": 1.2465,
+      "step": 9765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 20.442324359928453,
+      "learning_rate": 8.4940224992064e-06,
+      "loss": 1.264,
+      "step": 9770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.434910677344071,
+      "learning_rate": 8.491923887547926e-06,
+      "loss": 1.1754,
+      "step": 9775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 17.22452900332348,
+      "learning_rate": 8.489824074321778e-06,
+      "loss": 1.2591,
+      "step": 9780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.660771484878797,
+      "learning_rate": 8.487723060250498e-06,
+      "loss": 1.1941,
+      "step": 9785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.793455215195363,
+      "learning_rate": 8.485620846057044e-06,
+      "loss": 1.2331,
+      "step": 9790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 20.189243088111986,
+      "learning_rate": 8.483517432464788e-06,
+      "loss": 1.1827,
+      "step": 9795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 31.80617544301878,
+      "learning_rate": 8.481412820197508e-06,
+      "loss": 1.209,
+      "step": 9800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 51.16240905871924,
+      "learning_rate": 8.479307009979403e-06,
+      "loss": 1.1993,
+      "step": 9805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 37.645660326349336,
+      "learning_rate": 8.47720000253508e-06,
+      "loss": 1.2541,
+      "step": 9810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.537646193886891,
+      "learning_rate": 8.475091798589556e-06,
+      "loss": 1.2196,
+      "step": 9815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 74.70213748980201,
+      "learning_rate": 8.472982398868263e-06,
+      "loss": 1.2352,
+      "step": 9820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 135.7574431462861,
+      "learning_rate": 8.470871804097045e-06,
+      "loss": 1.2056,
+      "step": 9825
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 57.281387936043224,
+      "learning_rate": 8.468760015002153e-06,
+      "loss": 1.1798,
+      "step": 9830
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 49.41630425177251,
+      "learning_rate": 8.466647032310253e-06,
+      "loss": 1.2587,
+      "step": 9835
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 24.117672854707138,
+      "learning_rate": 8.46453285674842e-06,
+      "loss": 1.2449,
+      "step": 9840
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.016753129993532,
+      "learning_rate": 8.46241748904414e-06,
+      "loss": 1.2152,
+      "step": 9845
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.079832042525334,
+      "learning_rate": 8.46030092992531e-06,
+      "loss": 1.2344,
+      "step": 9850
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.341759458322716,
+      "learning_rate": 8.458183180120235e-06,
+      "loss": 1.2856,
+      "step": 9855
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.787416825397322,
+      "learning_rate": 8.456064240357628e-06,
+      "loss": 1.2745,
+      "step": 9860
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 45.30087761027042,
+      "learning_rate": 8.45394411136662e-06,
+      "loss": 1.1844,
+      "step": 9865
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 35.027289288363285,
+      "learning_rate": 8.451822793876742e-06,
+      "loss": 1.2217,
+      "step": 9870
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 19.68667017662464,
+      "learning_rate": 8.449700288617935e-06,
+      "loss": 1.2125,
+      "step": 9875
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.34371606158347,
+      "learning_rate": 8.447576596320558e-06,
+      "loss": 1.2053,
+      "step": 9880
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.141594774196527,
+      "learning_rate": 8.445451717715363e-06,
+      "loss": 1.164,
+      "step": 9885
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 25.351385895760103,
+      "learning_rate": 8.443325653533525e-06,
+      "loss": 1.2194,
+      "step": 9890
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 16.307025203392016,
+      "learning_rate": 8.441198404506616e-06,
+      "loss": 1.1969,
+      "step": 9895
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.175196142826138,
+      "learning_rate": 8.439069971366625e-06,
+      "loss": 1.1987,
+      "step": 9900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.931195059799597,
+      "learning_rate": 8.436940354845939e-06,
+      "loss": 1.2042,
+      "step": 9905
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.002472470006175,
+      "learning_rate": 8.434809555677361e-06,
+      "loss": 1.2303,
+      "step": 9910
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.290291630079329,
+      "learning_rate": 8.432677574594095e-06,
+      "loss": 1.1862,
+      "step": 9915
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.2109025716781616,
+      "eval_runtime": 25.7471,
+      "eval_samples_per_second": 31.305,
+      "eval_steps_per_second": 3.923,
+      "step": 9918
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.829768346435534,
+      "learning_rate": 8.430544412329756e-06,
+      "loss": 1.1478,
+      "step": 9920
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.970145587847068,
+      "learning_rate": 8.42841006961836e-06,
+      "loss": 0.9985,
+      "step": 9925
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.882498532354235,
+      "learning_rate": 8.426274547194332e-06,
+      "loss": 1.0612,
+      "step": 9930
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.97482649042033,
+      "learning_rate": 8.424137845792509e-06,
+      "loss": 0.9763,
+      "step": 9935
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.76466163919948,
+      "learning_rate": 8.42199996614812e-06,
+      "loss": 1.0238,
+      "step": 9940
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.112399675573785,
+      "learning_rate": 8.41986090899681e-06,
+      "loss": 1.0398,
+      "step": 9945
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.213687218919592,
+      "learning_rate": 8.417720675074633e-06,
+      "loss": 1.051,
+      "step": 9950
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.663983135535817,
+      "learning_rate": 8.415579265118032e-06,
+      "loss": 1.0585,
+      "step": 9955
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.359268922806681,
+      "learning_rate": 8.413436679863868e-06,
+      "loss": 0.9919,
+      "step": 9960
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.245242498527361,
+      "learning_rate": 8.411292920049403e-06,
+      "loss": 1.0313,
+      "step": 9965
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.768501766101803,
+      "learning_rate": 8.409147986412302e-06,
+      "loss": 0.9857,
+      "step": 9970
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 20.84868605870357,
+      "learning_rate": 8.407001879690637e-06,
+      "loss": 1.0668,
+      "step": 9975
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.832758318711273,
+      "learning_rate": 8.404854600622878e-06,
+      "loss": 1.061,
+      "step": 9980
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 28.323129537818932,
+      "learning_rate": 8.402706149947903e-06,
+      "loss": 1.0443,
+      "step": 9985
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 32.08304774339511,
+      "learning_rate": 8.40055652840499e-06,
+      "loss": 1.102,
+      "step": 9990
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.982485513234868,
+      "learning_rate": 8.398405736733827e-06,
+      "loss": 1.0058,
+      "step": 9995
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.166678030612518,
+      "learning_rate": 8.396253775674495e-06,
+      "loss": 0.9781,
+      "step": 10000
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 24.744865722386095,
+      "learning_rate": 8.39410064596748e-06,
+      "loss": 1.054,
+      "step": 10005
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 33.39987986980466,
+      "learning_rate": 8.391946348353676e-06,
+      "loss": 1.0492,
+      "step": 10010
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 16.197408664885675,
+      "learning_rate": 8.389790883574374e-06,
+      "loss": 1.0285,
+      "step": 10015
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.478049205668278,
+      "learning_rate": 8.387634252371267e-06,
+      "loss": 1.0178,
+      "step": 10020
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.22929308446771,
+      "learning_rate": 8.385476455486447e-06,
+      "loss": 1.0622,
+      "step": 10025
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.37257185988015,
+      "learning_rate": 8.383317493662412e-06,
+      "loss": 1.0177,
+      "step": 10030
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 34.315595057653006,
+      "learning_rate": 8.381157367642062e-06,
+      "loss": 1.0417,
+      "step": 10035
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 18.402346346049598,
+      "learning_rate": 8.37899607816869e-06,
+      "loss": 1.052,
+      "step": 10040
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.337683054348759,
+      "learning_rate": 8.376833625985994e-06,
+      "loss": 1.062,
+      "step": 10045
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 20.548054154408476,
+      "learning_rate": 8.374670011838072e-06,
+      "loss": 1.031,
+      "step": 10050
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.210150969500315,
+      "learning_rate": 8.372505236469424e-06,
+      "loss": 1.0524,
+      "step": 10055
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.613329324259729,
+      "learning_rate": 8.370339300624943e-06,
+      "loss": 1.0022,
+      "step": 10060
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.485345502725573,
+      "learning_rate": 8.36817220504993e-06,
+      "loss": 0.9975,
+      "step": 10065
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.671073845035285,
+      "learning_rate": 8.366003950490076e-06,
+      "loss": 1.0465,
+      "step": 10070
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.982752291924099,
+      "learning_rate": 8.36383453769148e-06,
+      "loss": 1.0219,
+      "step": 10075
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 14.116848311765736,
+      "learning_rate": 8.36166396740063e-06,
+      "loss": 1.0371,
+      "step": 10080
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.768301556799226,
+      "learning_rate": 8.35949224036442e-06,
+      "loss": 1.0401,
+      "step": 10085
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 13.02668779107601,
+      "learning_rate": 8.35731935733014e-06,
+      "loss": 0.9951,
+      "step": 10090
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.426296707679093,
+      "learning_rate": 8.355145319045475e-06,
+      "loss": 1.0167,
+      "step": 10095
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.00541707616357,
+      "learning_rate": 8.35297012625851e-06,
+      "loss": 0.9962,
+      "step": 10100
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.588001957684907,
+      "learning_rate": 8.350793779717727e-06,
+      "loss": 1.0284,
+      "step": 10105
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.013644599649537,
+      "learning_rate": 8.348616280172006e-06,
+      "loss": 1.0332,
+      "step": 10110
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 19.298108799226043,
+      "learning_rate": 8.346437628370621e-06,
+      "loss": 1.0447,
+      "step": 10115
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.074142078355483,
+      "learning_rate": 8.344257825063242e-06,
+      "loss": 1.0473,
+      "step": 10120
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.3778894218672,
+      "learning_rate": 8.342076870999943e-06,
+      "loss": 0.9901,
+      "step": 10125
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.1894231376209685,
+      "learning_rate": 8.339894766931183e-06,
+      "loss": 1.0352,
+      "step": 10130
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 15.009292576306457,
+      "learning_rate": 8.337711513607823e-06,
+      "loss": 0.9798,
+      "step": 10135
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.916583311141933,
+      "learning_rate": 8.335527111781121e-06,
+      "loss": 1.0658,
+      "step": 10140
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.335359800171835,
+      "learning_rate": 8.333341562202724e-06,
+      "loss": 1.0424,
+      "step": 10145
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.405141871430068,
+      "learning_rate": 8.33115486562468e-06,
+      "loss": 1.0568,
+      "step": 10150
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 15.878882635298778,
+      "learning_rate": 8.328967022799427e-06,
+      "loss": 1.0359,
+      "step": 10155
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.011786564450279,
+      "learning_rate": 8.326778034479802e-06,
+      "loss": 1.0394,
+      "step": 10160
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.231043280092075,
+      "learning_rate": 8.324587901419033e-06,
+      "loss": 1.0051,
+      "step": 10165
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.168059044544621,
+      "learning_rate": 8.322396624370741e-06,
+      "loss": 1.0036,
+      "step": 10170
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.939236046593278,
+      "learning_rate": 8.320204204088946e-06,
+      "loss": 1.0453,
+      "step": 10175
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.345591223655666,
+      "learning_rate": 8.318010641328053e-06,
+      "loss": 1.0282,
+      "step": 10180
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 18.90314250251431,
+      "learning_rate": 8.31581593684287e-06,
+      "loss": 0.9976,
+      "step": 10185
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 18.084551690785258,
+      "learning_rate": 8.313620091388588e-06,
+      "loss": 1.0772,
+      "step": 10190
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.684884049853789,
+      "learning_rate": 8.311423105720799e-06,
+      "loss": 1.0162,
+      "step": 10195
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.549694639447544,
+      "learning_rate": 8.30922498059548e-06,
+      "loss": 0.987,
+      "step": 10200
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.143518433232426,
+      "learning_rate": 8.307025716769008e-06,
+      "loss": 1.0323,
+      "step": 10205
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.884891662639422,
+      "learning_rate": 8.304825314998147e-06,
+      "loss": 0.9754,
+      "step": 10210
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.655469405362567,
+      "learning_rate": 8.302623776040047e-06,
+      "loss": 1.0069,
+      "step": 10215
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 14.075675562314656,
+      "learning_rate": 8.300421100652263e-06,
+      "loss": 1.0739,
+      "step": 10220
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 18.065814927344455,
+      "learning_rate": 8.298217289592729e-06,
+      "loss": 1.0376,
+      "step": 10225
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 54.61351609242025,
+      "learning_rate": 8.296012343619777e-06,
+      "loss": 1.1097,
+      "step": 10230
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.53814558868182,
+      "learning_rate": 8.293806263492126e-06,
+      "loss": 1.0131,
+      "step": 10235
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.438272279146418,
+      "learning_rate": 8.291599049968885e-06,
+      "loss": 1.0001,
+      "step": 10240
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.00508276425213,
+      "learning_rate": 8.289390703809554e-06,
+      "loss": 1.0835,
+      "step": 10245
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.395690192570587,
+      "learning_rate": 8.287181225774025e-06,
+      "loss": 1.051,
+      "step": 10250
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.050017934085144,
+      "learning_rate": 8.284970616622575e-06,
+      "loss": 0.9877,
+      "step": 10255
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.029724756383363,
+      "learning_rate": 8.282758877115873e-06,
+      "loss": 1.0171,
+      "step": 10260
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.208701180126377,
+      "learning_rate": 8.280546008014979e-06,
+      "loss": 1.0231,
+      "step": 10265
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.966732406844317,
+      "learning_rate": 8.278332010081333e-06,
+      "loss": 1.0141,
+      "step": 10270
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.461619182024871,
+      "learning_rate": 8.276116884076777e-06,
+      "loss": 1.0603,
+      "step": 10275
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.178804487830194,
+      "learning_rate": 8.273900630763529e-06,
+      "loss": 1.0469,
+      "step": 10280
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 15.349058379616233,
+      "learning_rate": 8.2716832509042e-06,
+      "loss": 0.9975,
+      "step": 10285
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 19.535910730995127,
+      "learning_rate": 8.26946474526179e-06,
+      "loss": 1.0539,
+      "step": 10290
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.929726128486617,
+      "learning_rate": 8.267245114599684e-06,
+      "loss": 1.0546,
+      "step": 10295
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.196993975441478,
+      "learning_rate": 8.265024359681655e-06,
+      "loss": 0.9936,
+      "step": 10300
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.850173272592965,
+      "learning_rate": 8.262802481271861e-06,
+      "loss": 1.041,
+      "step": 10305
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.691774593198793,
+      "learning_rate": 8.26057948013485e-06,
+      "loss": 1.0955,
+      "step": 10310
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.1729861230834295,
+      "learning_rate": 8.258355357035555e-06,
+      "loss": 1.0714,
+      "step": 10315
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.238572105953923,
+      "learning_rate": 8.256130112739293e-06,
+      "loss": 1.0458,
+      "step": 10320
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.624595274100203,
+      "learning_rate": 8.253903748011769e-06,
+      "loss": 1.0115,
+      "step": 10325
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.442304927152296,
+      "learning_rate": 8.251676263619074e-06,
+      "loss": 1.0122,
+      "step": 10330
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.0748278749495785,
+      "learning_rate": 8.249447660327681e-06,
+      "loss": 1.0554,
+      "step": 10335
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.41122547786913,
+      "learning_rate": 8.247217938904453e-06,
+      "loss": 1.0389,
+      "step": 10340
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 14.149361527527443,
+      "learning_rate": 8.244987100116632e-06,
+      "loss": 1.066,
+      "step": 10345
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 20.382488116202655,
+      "learning_rate": 8.242755144731851e-06,
+      "loss": 1.0616,
+      "step": 10350
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.805046554209081,
+      "learning_rate": 8.240522073518122e-06,
+      "loss": 1.0178,
+      "step": 10355
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.771420372480122,
+      "learning_rate": 8.23828788724384e-06,
+      "loss": 1.0467,
+      "step": 10360
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 39.87695906960376,
+      "learning_rate": 8.236052586677792e-06,
+      "loss": 1.0675,
+      "step": 10365
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 26.86833571951981,
+      "learning_rate": 8.233816172589139e-06,
+      "loss": 1.0649,
+      "step": 10370
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.884515060507346,
+      "learning_rate": 8.231578645747428e-06,
+      "loss": 1.0472,
+      "step": 10375
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 23.93658008919047,
+      "learning_rate": 8.229340006922592e-06,
+      "loss": 1.0236,
+      "step": 10380
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.962471938855685,
+      "learning_rate": 8.227100256884945e-06,
+      "loss": 1.0748,
+      "step": 10385
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.589553264943177,
+      "learning_rate": 8.224859396405183e-06,
+      "loss": 1.0212,
+      "step": 10390
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 27.241871994524665,
+      "learning_rate": 8.22261742625438e-06,
+      "loss": 1.0397,
+      "step": 10395
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 35.438534496728494,
+      "learning_rate": 8.220374347204001e-06,
+      "loss": 1.0519,
+      "step": 10400
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 63.37721092054966,
+      "learning_rate": 8.218130160025883e-06,
+      "loss": 1.0447,
+      "step": 10405
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 46.362651376223255,
+      "learning_rate": 8.215884865492252e-06,
+      "loss": 1.0396,
+      "step": 10410
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 86.81240093881469,
+      "learning_rate": 8.21363846437571e-06,
+      "loss": 1.0687,
+      "step": 10415
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 38.108316598515685,
+      "learning_rate": 8.211390957449242e-06,
+      "loss": 1.0728,
+      "step": 10420
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 33.206193958643894,
+      "learning_rate": 8.20914234548621e-06,
+      "loss": 1.0477,
+      "step": 10425
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.020883745985015,
+      "learning_rate": 8.206892629260366e-06,
+      "loss": 1.0067,
+      "step": 10430
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.45834233869216,
+      "learning_rate": 8.204641809545829e-06,
+      "loss": 1.0321,
+      "step": 10435
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 22.878622903566985,
+      "learning_rate": 8.202389887117106e-06,
+      "loss": 1.0424,
+      "step": 10440
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.914121611087722,
+      "learning_rate": 8.200136862749083e-06,
+      "loss": 1.0094,
+      "step": 10445
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.991708402704003,
+      "learning_rate": 8.19788273721702e-06,
+      "loss": 1.08,
+      "step": 10450
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 43.71924056121477,
+      "learning_rate": 8.195627511296562e-06,
+      "loss": 1.0456,
+      "step": 10455
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 74.34005505033582,
+      "learning_rate": 8.19337118576373e-06,
+      "loss": 1.0431,
+      "step": 10460
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 26.077817865542375,
+      "learning_rate": 8.191113761394924e-06,
+      "loss": 1.0635,
+      "step": 10465
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.513061881158686,
+      "learning_rate": 8.188855238966918e-06,
+      "loss": 1.0094,
+      "step": 10470
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 27.962173814297504,
+      "learning_rate": 8.186595619256872e-06,
+      "loss": 1.0165,
+      "step": 10475
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.287680421146506,
+      "learning_rate": 8.184334903042317e-06,
+      "loss": 1.0481,
+      "step": 10480
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 18.770639134900765,
+      "learning_rate": 8.182073091101162e-06,
+      "loss": 1.0454,
+      "step": 10485
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.44461465762257,
+      "learning_rate": 8.179810184211697e-06,
+      "loss": 1.0993,
+      "step": 10490
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.95781697961938,
+      "learning_rate": 8.177546183152584e-06,
+      "loss": 1.0544,
+      "step": 10495
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.94690016611478,
+      "learning_rate": 8.175281088702865e-06,
+      "loss": 1.0863,
+      "step": 10500
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.348800475665943,
+      "learning_rate": 8.173014901641955e-06,
+      "loss": 1.079,
+      "step": 10505
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.869679015605106,
+      "learning_rate": 8.170747622749648e-06,
+      "loss": 1.029,
+      "step": 10510
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.568519455147601,
+      "learning_rate": 8.168479252806111e-06,
+      "loss": 1.069,
+      "step": 10515
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.191276539726678,
+      "learning_rate": 8.166209792591893e-06,
+      "loss": 1.0137,
+      "step": 10520
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 17.364086529976788,
+      "learning_rate": 8.163939242887909e-06,
+      "loss": 1.0729,
+      "step": 10525
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.752862921532925,
+      "learning_rate": 8.161667604475452e-06,
+      "loss": 1.0175,
+      "step": 10530
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.257352736574848,
+      "learning_rate": 8.15939487813619e-06,
+      "loss": 1.0339,
+      "step": 10535
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 21.954806474603796,
+      "learning_rate": 8.157121064652171e-06,
+      "loss": 1.0565,
+      "step": 10540
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.781500694131484,
+      "learning_rate": 8.154846164805807e-06,
+      "loss": 1.0432,
+      "step": 10545
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 22.378219693704544,
+      "learning_rate": 8.152570179379893e-06,
+      "loss": 1.0174,
+      "step": 10550
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.7171653504915385,
+      "learning_rate": 8.15029310915759e-06,
+      "loss": 1.0416,
+      "step": 10555
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.155624751237,
+      "learning_rate": 8.148014954922438e-06,
+      "loss": 1.0438,
+      "step": 10560
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.151260055253163,
+      "learning_rate": 8.145735717458344e-06,
+      "loss": 0.9877,
+      "step": 10565
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.068599316429127,
+      "learning_rate": 8.143455397549597e-06,
+      "loss": 1.0721,
+      "step": 10570
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.183688049917007,
+      "learning_rate": 8.14117399598085e-06,
+      "loss": 1.0617,
+      "step": 10575
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 16.8120516390695,
+      "learning_rate": 8.13889151353713e-06,
+      "loss": 1.0236,
+      "step": 10580
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.849510964885902,
+      "learning_rate": 8.13660795100384e-06,
+      "loss": 1.0305,
+      "step": 10585
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.130947304232826,
+      "learning_rate": 8.134323309166747e-06,
+      "loss": 1.0564,
+      "step": 10590
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.548870019213742,
+      "learning_rate": 8.132037588811998e-06,
+      "loss": 1.0183,
+      "step": 10595
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 35.41958649859262,
+      "learning_rate": 8.129750790726108e-06,
+      "loss": 1.0407,
+      "step": 10600
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 20.220901648050884,
+      "learning_rate": 8.127462915695958e-06,
+      "loss": 1.0309,
+      "step": 10605
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.431443131862906,
+      "learning_rate": 8.125173964508806e-06,
+      "loss": 0.9904,
+      "step": 10610
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.018049652826944,
+      "learning_rate": 8.12288393795228e-06,
+      "loss": 1.05,
+      "step": 10615
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.331762857681298,
+      "learning_rate": 8.120592836814372e-06,
+      "loss": 0.9944,
+      "step": 10620
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.694322359008826,
+      "learning_rate": 8.118300661883447e-06,
+      "loss": 1.0141,
+      "step": 10625
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.139637852801844,
+      "learning_rate": 8.116007413948246e-06,
+      "loss": 1.0605,
+      "step": 10630
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 27.75660179197848,
+      "learning_rate": 8.113713093797868e-06,
+      "loss": 1.0805,
+      "step": 10635
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 23.475321565301055,
+      "learning_rate": 8.111417702221789e-06,
+      "loss": 1.0264,
+      "step": 10640
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 21.672369017405423,
+      "learning_rate": 8.109121240009851e-06,
+      "loss": 1.0534,
+      "step": 10645
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.602144775695785,
+      "learning_rate": 8.106823707952263e-06,
+      "loss": 1.0419,
+      "step": 10650
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.930638541831872,
+      "learning_rate": 8.104525106839606e-06,
+      "loss": 1.0507,
+      "step": 10655
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 21.190413953730527,
+      "learning_rate": 8.102225437462824e-06,
+      "loss": 1.0411,
+      "step": 10660
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 20.32191528592284,
+      "learning_rate": 8.09992470061323e-06,
+      "loss": 1.0641,
+      "step": 10665
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.930762358782742,
+      "learning_rate": 8.097622897082512e-06,
+      "loss": 1.0278,
+      "step": 10670
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 34.26451847010071,
+      "learning_rate": 8.09532002766271e-06,
+      "loss": 1.0161,
+      "step": 10675
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 54.58971847600707,
+      "learning_rate": 8.093016093146242e-06,
+      "loss": 1.0074,
+      "step": 10680
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 21.47990163607273,
+      "learning_rate": 8.090711094325891e-06,
+      "loss": 1.0681,
+      "step": 10685
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.386616762935336,
+      "learning_rate": 8.088405031994807e-06,
+      "loss": 1.0295,
+      "step": 10690
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 27.232856368628745,
+      "learning_rate": 8.0860979069465e-06,
+      "loss": 1.0289,
+      "step": 10695
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.453651242681508,
+      "learning_rate": 8.083789719974851e-06,
+      "loss": 1.0618,
+      "step": 10700
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.668346991769979,
+      "learning_rate": 8.081480471874104e-06,
+      "loss": 1.0612,
+      "step": 10705
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.643278409339298,
+      "learning_rate": 8.079170163438868e-06,
+      "loss": 1.0549,
+      "step": 10710
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.331052850098343,
+      "learning_rate": 8.076858795464122e-06,
+      "loss": 1.0276,
+      "step": 10715
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.186490783994433,
+      "learning_rate": 8.074546368745203e-06,
+      "loss": 1.0888,
+      "step": 10720
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.851520851508756,
+      "learning_rate": 8.072232884077816e-06,
+      "loss": 1.0049,
+      "step": 10725
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 43.982567273199294,
+      "learning_rate": 8.069918342258026e-06,
+      "loss": 1.0366,
+      "step": 10730
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 15.929804138237563,
+      "learning_rate": 8.067602744082268e-06,
+      "loss": 1.0421,
+      "step": 10735
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.961094668401225,
+      "learning_rate": 8.065286090347335e-06,
+      "loss": 1.0774,
+      "step": 10740
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 24.28551240410909,
+      "learning_rate": 8.062968381850386e-06,
+      "loss": 1.0076,
+      "step": 10745
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.853929888653179,
+      "learning_rate": 8.060649619388942e-06,
+      "loss": 1.0128,
+      "step": 10750
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.068621266149146,
+      "learning_rate": 8.058329803760887e-06,
+      "loss": 1.0167,
+      "step": 10755
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.775858823799506,
+      "learning_rate": 8.056008935764469e-06,
+      "loss": 1.0708,
+      "step": 10760
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 28.512827810024064,
+      "learning_rate": 8.053687016198292e-06,
+      "loss": 1.0837,
+      "step": 10765
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 23.63028805052106,
+      "learning_rate": 8.051364045861332e-06,
+      "loss": 1.0027,
+      "step": 10770
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.396814053507557,
+      "learning_rate": 8.049040025552919e-06,
+      "loss": 1.0051,
+      "step": 10775
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 33.01296800751487,
+      "learning_rate": 8.046714956072744e-06,
+      "loss": 1.0547,
+      "step": 10780
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 38.4056182742984,
+      "learning_rate": 8.044388838220863e-06,
+      "loss": 1.0533,
+      "step": 10785
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.4678980940781985,
+      "learning_rate": 8.04206167279769e-06,
+      "loss": 1.0223,
+      "step": 10790
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 16.31958176301968,
+      "learning_rate": 8.039733460604002e-06,
+      "loss": 1.0467,
+      "step": 10795
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 29.78757054516638,
+      "learning_rate": 8.037404202440932e-06,
+      "loss": 1.1068,
+      "step": 10800
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.351514786042655,
+      "learning_rate": 8.035073899109978e-06,
+      "loss": 1.055,
+      "step": 10805
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 16.13405242140683,
+      "learning_rate": 8.032742551412994e-06,
+      "loss": 1.0259,
+      "step": 10810
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 23.518528190560342,
+      "learning_rate": 8.030410160152193e-06,
+      "loss": 1.0932,
+      "step": 10815
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 39.96884044404456,
+      "learning_rate": 8.028076726130152e-06,
+      "loss": 1.037,
+      "step": 10820
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 27.79746347759069,
+      "learning_rate": 8.0257422501498e-06,
+      "loss": 1.0445,
+      "step": 10825
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 41.14827167125988,
+      "learning_rate": 8.023406733014432e-06,
+      "loss": 1.12,
+      "step": 10830
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 41.22766937081543,
+      "learning_rate": 8.021070175527693e-06,
+      "loss": 1.0478,
+      "step": 10835
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.38343686690903,
+      "learning_rate": 8.018732578493593e-06,
+      "loss": 1.0177,
+      "step": 10840
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.7279491440924115,
+      "learning_rate": 8.016393942716495e-06,
+      "loss": 1.0698,
+      "step": 10845
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.603661113936377,
+      "learning_rate": 8.014054269001124e-06,
+      "loss": 1.0172,
+      "step": 10850
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.832556522153212,
+      "learning_rate": 8.011713558152556e-06,
+      "loss": 1.0496,
+      "step": 10855
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.44330576857254,
+      "learning_rate": 8.00937181097623e-06,
+      "loss": 1.0259,
+      "step": 10860
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 19.961190197259683,
+      "learning_rate": 8.00702902827794e-06,
+      "loss": 1.0294,
+      "step": 10865
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.772105224394243,
+      "learning_rate": 8.004685210863831e-06,
+      "loss": 1.0449,
+      "step": 10870
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 15.284396898913977,
+      "learning_rate": 8.00234035954041e-06,
+      "loss": 1.0185,
+      "step": 10875
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 29.785290427174154,
+      "learning_rate": 7.999994475114541e-06,
+      "loss": 0.9921,
+      "step": 10880
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.357731611223983,
+      "learning_rate": 7.997647558393438e-06,
+      "loss": 1.0381,
+      "step": 10885
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.406314279732902,
+      "learning_rate": 7.995299610184673e-06,
+      "loss": 1.0861,
+      "step": 10890
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.625404656509023,
+      "learning_rate": 7.992950631296172e-06,
+      "loss": 1.0584,
+      "step": 10895
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 34.69791299273681,
+      "learning_rate": 7.990600622536217e-06,
+      "loss": 1.0001,
+      "step": 10900
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.127613634453,
+      "learning_rate": 7.988249584713447e-06,
+      "loss": 1.1006,
+      "step": 10905
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.244903524023991,
+      "learning_rate": 7.985897518636847e-06,
+      "loss": 1.1066,
+      "step": 10910
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.772660505535132,
+      "learning_rate": 7.983544425115763e-06,
+      "loss": 1.0398,
+      "step": 10915
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.461138582086308,
+      "learning_rate": 7.981190304959891e-06,
+      "loss": 1.066,
+      "step": 10920
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 26.736545490527853,
+      "learning_rate": 7.978835158979285e-06,
+      "loss": 1.0661,
+      "step": 10925
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.0084969505494,
+      "learning_rate": 7.976478987984345e-06,
+      "loss": 1.02,
+      "step": 10930
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 25.307357719882972,
+      "learning_rate": 7.974121792785826e-06,
+      "loss": 1.0413,
+      "step": 10935
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.887929931927594,
+      "learning_rate": 7.97176357419484e-06,
+      "loss": 1.0631,
+      "step": 10940
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.280734504239222,
+      "learning_rate": 7.969404333022846e-06,
+      "loss": 1.0711,
+      "step": 10945
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 22.243380183936026,
+      "learning_rate": 7.967044070081658e-06,
+      "loss": 1.0771,
+      "step": 10950
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 32.605761545212125,
+      "learning_rate": 7.96468278618344e-06,
+      "loss": 1.0772,
+      "step": 10955
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.284422751807178,
+      "learning_rate": 7.962320482140704e-06,
+      "loss": 1.0444,
+      "step": 10960
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.770567621626649,
+      "learning_rate": 7.959957158766323e-06,
+      "loss": 1.0371,
+      "step": 10965
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.864176917653522,
+      "learning_rate": 7.957592816873509e-06,
+      "loss": 1.0122,
+      "step": 10970
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.70443716107041,
+      "learning_rate": 7.95522745727583e-06,
+      "loss": 1.0528,
+      "step": 10975
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.24545368926349,
+      "learning_rate": 7.952861080787209e-06,
+      "loss": 1.0377,
+      "step": 10980
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.757228862705906,
+      "learning_rate": 7.950493688221907e-06,
+      "loss": 1.0669,
+      "step": 10985
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.116188153519241,
+      "learning_rate": 7.948125280394548e-06,
+      "loss": 1.0545,
+      "step": 10990
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.717967247580755,
+      "learning_rate": 7.945755858120095e-06,
+      "loss": 1.0256,
+      "step": 10995
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.836371646522697,
+      "learning_rate": 7.943385422213862e-06,
+      "loss": 1.0441,
+      "step": 11000
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 30.217546600081967,
+      "learning_rate": 7.941013973491518e-06,
+      "loss": 1.0654,
+      "step": 11005
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 40.932544472644175,
+      "learning_rate": 7.938641512769077e-06,
+      "loss": 1.0572,
+      "step": 11010
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 22.72804791336364,
+      "learning_rate": 7.936268040862895e-06,
+      "loss": 1.0351,
+      "step": 11015
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 15.044052151674721,
+      "learning_rate": 7.933893558589687e-06,
+      "loss": 1.0369,
+      "step": 11020
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.747254479015107,
+      "learning_rate": 7.931518066766506e-06,
+      "loss": 1.0475,
+      "step": 11025
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.632051553686237,
+      "learning_rate": 7.92914156621076e-06,
+      "loss": 1.035,
+      "step": 11030
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.1829450330708955,
+      "learning_rate": 7.926764057740198e-06,
+      "loss": 1.0905,
+      "step": 11035
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 26.81836768913999,
+      "learning_rate": 7.924385542172919e-06,
+      "loss": 1.0201,
+      "step": 11040
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 17.98999389891583,
+      "learning_rate": 7.922006020327365e-06,
+      "loss": 1.0039,
+      "step": 11045
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 21.47333072000413,
+      "learning_rate": 7.919625493022335e-06,
+      "loss": 1.0052,
+      "step": 11050
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.565135655710572,
+      "learning_rate": 7.917243961076959e-06,
+      "loss": 1.0425,
+      "step": 11055
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.493952061391708,
+      "learning_rate": 7.914861425310723e-06,
+      "loss": 1.0527,
+      "step": 11060
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.28996338406018,
+      "learning_rate": 7.912477886543454e-06,
+      "loss": 1.1012,
+      "step": 11065
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.786521467272293,
+      "learning_rate": 7.910093345595326e-06,
+      "loss": 1.064,
+      "step": 11070
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 17.638360793188944,
+      "learning_rate": 7.907707803286856e-06,
+      "loss": 1.0817,
+      "step": 11075
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.579350487528567,
+      "learning_rate": 7.90532126043891e-06,
+      "loss": 1.0424,
+      "step": 11080
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.179693273682382,
+      "learning_rate": 7.902933717872693e-06,
+      "loss": 1.0705,
+      "step": 11085
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 18.410102516344413,
+      "learning_rate": 7.900545176409756e-06,
+      "loss": 1.0398,
+      "step": 11090
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.49004821436449,
+      "learning_rate": 7.898155636871995e-06,
+      "loss": 0.9996,
+      "step": 11095
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.33850196492309,
+      "learning_rate": 7.895765100081646e-06,
+      "loss": 1.0006,
+      "step": 11100
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.626847004913168,
+      "learning_rate": 7.893373566861292e-06,
+      "loss": 1.0345,
+      "step": 11105
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.446076591813217,
+      "learning_rate": 7.890981038033859e-06,
+      "loss": 1.0387,
+      "step": 11110
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.060957710750417,
+      "learning_rate": 7.88858751442261e-06,
+      "loss": 1.026,
+      "step": 11115
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.228268268058988,
+      "learning_rate": 7.886192996851157e-06,
+      "loss": 1.0508,
+      "step": 11120
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.8075659425069,
+      "learning_rate": 7.88379748614345e-06,
+      "loss": 1.077,
+      "step": 11125
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.819613988104902,
+      "learning_rate": 7.881400983123781e-06,
+      "loss": 1.0307,
+      "step": 11130
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.468776605044408,
+      "learning_rate": 7.879003488616788e-06,
+      "loss": 1.0522,
+      "step": 11135
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.1731817046492745,
+      "learning_rate": 7.876605003447443e-06,
+      "loss": 1.0664,
+      "step": 11140
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.531644225900602,
+      "learning_rate": 7.874205528441066e-06,
+      "loss": 1.0354,
+      "step": 11145
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.930059962381911,
+      "learning_rate": 7.871805064423308e-06,
+      "loss": 1.0302,
+      "step": 11150
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 23.65862826045937,
+      "learning_rate": 7.869403612220174e-06,
+      "loss": 1.0645,
+      "step": 11155
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.90337054778474,
+      "learning_rate": 7.867001172657995e-06,
+      "loss": 1.0685,
+      "step": 11160
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 17.389451509530925,
+      "learning_rate": 7.864597746563453e-06,
+      "loss": 1.0228,
+      "step": 11165
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 19.688390048504406,
+      "learning_rate": 7.862193334763562e-06,
+      "loss": 1.0236,
+      "step": 11170
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.867171483942066,
+      "learning_rate": 7.859787938085677e-06,
+      "loss": 1.098,
+      "step": 11175
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.273633495798792,
+      "learning_rate": 7.857381557357495e-06,
+      "loss": 1.0321,
+      "step": 11180
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 30.61533719738187,
+      "learning_rate": 7.854974193407047e-06,
+      "loss": 1.0454,
+      "step": 11185
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 18.968282250852365,
+      "learning_rate": 7.852565847062706e-06,
+      "loss": 1.0331,
+      "step": 11190
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 28.827140133606875,
+      "learning_rate": 7.850156519153183e-06,
+      "loss": 1.024,
+      "step": 11195
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 36.740794689535356,
+      "learning_rate": 7.847746210507522e-06,
+      "loss": 1.099,
+      "step": 11200
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 44.494900973755506,
+      "learning_rate": 7.845334921955111e-06,
+      "loss": 1.0322,
+      "step": 11205
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 42.45845432983005,
+      "learning_rate": 7.842922654325672e-06,
+      "loss": 1.05,
+      "step": 11210
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.523642147761985,
+      "learning_rate": 7.84050940844926e-06,
+      "loss": 1.0771,
+      "step": 11215
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 21.394059843462124,
+      "learning_rate": 7.838095185156276e-06,
+      "loss": 1.0405,
+      "step": 11220
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.283677153672595,
+      "learning_rate": 7.83567998527745e-06,
+      "loss": 1.0834,
+      "step": 11225
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 28.72774136482472,
+      "learning_rate": 7.833263809643848e-06,
+      "loss": 1.0269,
+      "step": 11230
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 19.66660259770272,
+      "learning_rate": 7.830846659086876e-06,
+      "loss": 1.0759,
+      "step": 11235
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.813658613230237,
+      "learning_rate": 7.82842853443827e-06,
+      "loss": 1.0506,
+      "step": 11240
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.133952082068999,
+      "learning_rate": 7.826009436530109e-06,
+      "loss": 1.0687,
+      "step": 11245
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.930521893875791,
+      "learning_rate": 7.823589366194799e-06,
+      "loss": 1.043,
+      "step": 11250
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.752854297463534,
+      "learning_rate": 7.821168324265088e-06,
+      "loss": 1.0623,
+      "step": 11255
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.145263152954865,
+      "learning_rate": 7.818746311574047e-06,
+      "loss": 1.0475,
+      "step": 11260
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.308207081197901,
+      "learning_rate": 7.816323328955094e-06,
+      "loss": 1.0341,
+      "step": 11265
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 5.357752176077241,
+      "learning_rate": 7.813899377241973e-06,
+      "loss": 1.0696,
+      "step": 11270
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 35.70239091708709,
+      "learning_rate": 7.811474457268765e-06,
+      "loss": 1.0979,
+      "step": 11275
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.081859081902147,
+      "learning_rate": 7.809048569869881e-06,
+      "loss": 1.018,
+      "step": 11280
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.144620324513742,
+      "learning_rate": 7.806621715880066e-06,
+      "loss": 1.0702,
+      "step": 11285
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.676098661137126,
+      "learning_rate": 7.804193896134402e-06,
+      "loss": 1.0519,
+      "step": 11290
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 13.957609131873946,
+      "learning_rate": 7.801765111468295e-06,
+      "loss": 1.065,
+      "step": 11295
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 30.804669147183866,
+      "learning_rate": 7.799335362717488e-06,
+      "loss": 1.0803,
+      "step": 11300
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.972976576259954,
+      "learning_rate": 7.796904650718058e-06,
+      "loss": 1.0595,
+      "step": 11305
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.42614691099274,
+      "learning_rate": 7.794472976306409e-06,
+      "loss": 1.1013,
+      "step": 11310
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.3877898822812,
+      "learning_rate": 7.792040340319275e-06,
+      "loss": 1.0484,
+      "step": 11315
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.5635085577917245,
+      "learning_rate": 7.78960674359373e-06,
+      "loss": 1.0075,
+      "step": 11320
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.7544240823692085,
+      "learning_rate": 7.787172186967169e-06,
+      "loss": 1.0101,
+      "step": 11325
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.521578068619176,
+      "learning_rate": 7.784736671277318e-06,
+      "loss": 1.0461,
+      "step": 11330
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.301669315226013,
+      "learning_rate": 7.782300197362241e-06,
+      "loss": 0.9991,
+      "step": 11335
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.220694186657157,
+      "learning_rate": 7.77986276606032e-06,
+      "loss": 1.0877,
+      "step": 11340
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.972744230181513,
+      "learning_rate": 7.777424378210279e-06,
+      "loss": 1.0245,
+      "step": 11345
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 15.665389124617688,
+      "learning_rate": 7.77498503465116e-06,
+      "loss": 1.0289,
+      "step": 11350
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 22.08979678059866,
+      "learning_rate": 7.772544736222345e-06,
+      "loss": 1.0317,
+      "step": 11355
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 26.602186669253843,
+      "learning_rate": 7.77010348376353e-06,
+      "loss": 1.024,
+      "step": 11360
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 27.038319449125172,
+      "learning_rate": 7.767661278114754e-06,
+      "loss": 1.015,
+      "step": 11365
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 32.94987066921108,
+      "learning_rate": 7.765218120116376e-06,
+      "loss": 1.0769,
+      "step": 11370
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 23.66418718772051,
+      "learning_rate": 7.762774010609083e-06,
+      "loss": 1.0451,
+      "step": 11375
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 32.45744837740459,
+      "learning_rate": 7.760328950433892e-06,
+      "loss": 1.0282,
+      "step": 11380
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 29.766784519041305,
+      "learning_rate": 7.757882940432145e-06,
+      "loss": 1.0397,
+      "step": 11385
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.7197884991075405,
+      "learning_rate": 7.755435981445513e-06,
+      "loss": 1.0311,
+      "step": 11390
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 23.359189631838863,
+      "learning_rate": 7.752988074315991e-06,
+      "loss": 1.0214,
+      "step": 11395
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 23.444606363008706,
+      "learning_rate": 7.750539219885902e-06,
+      "loss": 1.0268,
+      "step": 11400
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.199249449692479,
+      "learning_rate": 7.748089418997895e-06,
+      "loss": 1.018,
+      "step": 11405
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.347617826238135,
+      "learning_rate": 7.745638672494944e-06,
+      "loss": 1.0613,
+      "step": 11410
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.417151079608356,
+      "learning_rate": 7.743186981220348e-06,
+      "loss": 1.061,
+      "step": 11415
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.959055237876981,
+      "learning_rate": 7.740734346017733e-06,
+      "loss": 1.056,
+      "step": 11420
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.176059141528643,
+      "learning_rate": 7.738280767731045e-06,
+      "loss": 0.9817,
+      "step": 11425
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.2690785308256105,
+      "learning_rate": 7.73582624720456e-06,
+      "loss": 1.0171,
+      "step": 11430
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.486811724837144,
+      "learning_rate": 7.733370785282879e-06,
+      "loss": 1.0282,
+      "step": 11435
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.823763052543553,
+      "learning_rate": 7.730914382810919e-06,
+      "loss": 1.0827,
+      "step": 11440
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.257250829565304,
+      "learning_rate": 7.72845704063393e-06,
+      "loss": 1.0779,
+      "step": 11445
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.893179851041156,
+      "learning_rate": 7.725998759597478e-06,
+      "loss": 1.0631,
+      "step": 11450
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.05310239761301,
+      "learning_rate": 7.723539540547455e-06,
+      "loss": 1.087,
+      "step": 11455
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.568081314755192,
+      "learning_rate": 7.72107938433008e-06,
+      "loss": 1.0407,
+      "step": 11460
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.30451232867941,
+      "learning_rate": 7.718618291791887e-06,
+      "loss": 1.0147,
+      "step": 11465
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.581416562349123,
+      "learning_rate": 7.716156263779736e-06,
+      "loss": 1.0496,
+      "step": 11470
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.265076701434122,
+      "learning_rate": 7.71369330114081e-06,
+      "loss": 1.0413,
+      "step": 11475
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.449712305653418,
+      "learning_rate": 7.71122940472261e-06,
+      "loss": 1.0169,
+      "step": 11480
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.470154640544608,
+      "learning_rate": 7.708764575372962e-06,
+      "loss": 1.0046,
+      "step": 11485
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.209267270155115,
+      "learning_rate": 7.70629881394001e-06,
+      "loss": 1.0392,
+      "step": 11490
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.664536173796234,
+      "learning_rate": 7.703832121272221e-06,
+      "loss": 1.0161,
+      "step": 11495
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.939112700472921,
+      "learning_rate": 7.701364498218381e-06,
+      "loss": 1.0694,
+      "step": 11500
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 17.043229302231182,
+      "learning_rate": 7.698895945627597e-06,
+      "loss": 0.9974,
+      "step": 11505
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.990478258309913,
+      "learning_rate": 7.696426464349299e-06,
+      "loss": 1.0676,
+      "step": 11510
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 25.65013648668628,
+      "learning_rate": 7.693956055233227e-06,
+      "loss": 1.068,
+      "step": 11515
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.336450698290124,
+      "learning_rate": 7.69148471912945e-06,
+      "loss": 1.0261,
+      "step": 11520
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.209808655252218,
+      "learning_rate": 7.689012456888352e-06,
+      "loss": 1.0366,
+      "step": 11525
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.653020039541587,
+      "learning_rate": 7.686539269360636e-06,
+      "loss": 1.0248,
+      "step": 11530
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 21.89442345749155,
+      "learning_rate": 7.684065157397323e-06,
+      "loss": 1.0303,
+      "step": 11535
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.183307252692156,
+      "learning_rate": 7.681590121849754e-06,
+      "loss": 1.0576,
+      "step": 11540
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.55522118711657,
+      "learning_rate": 7.679114163569584e-06,
+      "loss": 0.9881,
+      "step": 11545
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.645009341875197,
+      "learning_rate": 7.67663728340879e-06,
+      "loss": 1.0538,
+      "step": 11550
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.288234650084943,
+      "learning_rate": 7.674159482219663e-06,
+      "loss": 1.0916,
+      "step": 11555
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.323944478111165,
+      "learning_rate": 7.671680760854812e-06,
+      "loss": 1.0389,
+      "step": 11560
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.28682974878779,
+      "learning_rate": 7.669201120167164e-06,
+      "loss": 1.0077,
+      "step": 11565
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.965117672513008,
+      "learning_rate": 7.666720561009959e-06,
+      "loss": 1.0556,
+      "step": 11570
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.879913896859524,
+      "learning_rate": 7.664239084236756e-06,
+      "loss": 1.0431,
+      "step": 11575
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.09897871614661,
+      "learning_rate": 7.661756690701429e-06,
+      "loss": 1.0235,
+      "step": 11580
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.193103071161919,
+      "learning_rate": 7.659273381258165e-06,
+      "loss": 1.0652,
+      "step": 11585
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 36.002717652272395,
+      "learning_rate": 7.656789156761473e-06,
+      "loss": 1.0214,
+      "step": 11590
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 27.389027634831674,
+      "learning_rate": 7.654304018066169e-06,
+      "loss": 1.0712,
+      "step": 11595
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 29.295790763413898,
+      "learning_rate": 7.651817966027387e-06,
+      "loss": 1.0563,
+      "step": 11600
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 78.5986973673433,
+      "learning_rate": 7.649331001500576e-06,
+      "loss": 1.0787,
+      "step": 11605
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.747085214926916,
+      "learning_rate": 7.646843125341498e-06,
+      "loss": 1.0394,
+      "step": 11610
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 42.82699540343177,
+      "learning_rate": 7.644354338406229e-06,
+      "loss": 1.0461,
+      "step": 11615
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 38.36576292944948,
+      "learning_rate": 7.641864641551157e-06,
+      "loss": 1.0601,
+      "step": 11620
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 42.49809837717228,
+      "learning_rate": 7.639374035632984e-06,
+      "loss": 1.0101,
+      "step": 11625
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 21.982321503938692,
+      "learning_rate": 7.636882521508728e-06,
+      "loss": 1.0013,
+      "step": 11630
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 31.605933081763478,
+      "learning_rate": 7.634390100035716e-06,
+      "loss": 1.0223,
+      "step": 11635
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.527420466489241,
+      "learning_rate": 7.631896772071585e-06,
+      "loss": 1.0067,
+      "step": 11640
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 57.497324855910215,
+      "learning_rate": 7.629402538474288e-06,
+      "loss": 1.0984,
+      "step": 11645
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 70.73824629974315,
+      "learning_rate": 7.626907400102089e-06,
+      "loss": 1.0807,
+      "step": 11650
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 38.39492524899637,
+      "learning_rate": 7.624411357813564e-06,
+      "loss": 1.021,
+      "step": 11655
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 23.23834490509645,
+      "learning_rate": 7.621914412467597e-06,
+      "loss": 1.0724,
+      "step": 11660
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.626825151703164,
+      "learning_rate": 7.6194165649233834e-06,
+      "loss": 1.0526,
+      "step": 11665
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.050704659349623,
+      "learning_rate": 7.616917816040433e-06,
+      "loss": 1.058,
+      "step": 11670
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 25.018079860661217,
+      "learning_rate": 7.614418166678562e-06,
+      "loss": 1.0797,
+      "step": 11675
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 19.988606047033,
+      "learning_rate": 7.611917617697896e-06,
+      "loss": 1.0386,
+      "step": 11680
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.389834101700366,
+      "learning_rate": 7.609416169958872e-06,
+      "loss": 1.0451,
+      "step": 11685
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.369819552441536,
+      "learning_rate": 7.606913824322238e-06,
+      "loss": 1.013,
+      "step": 11690
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.873428629131478,
+      "learning_rate": 7.604410581649045e-06,
+      "loss": 1.0511,
+      "step": 11695
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.226147562203659,
+      "learning_rate": 7.601906442800658e-06,
+      "loss": 1.0711,
+      "step": 11700
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.70917967690263,
+      "learning_rate": 7.599401408638751e-06,
+      "loss": 1.0148,
+      "step": 11705
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.286083740875113,
+      "learning_rate": 7.5968954800253015e-06,
+      "loss": 1.0611,
+      "step": 11710
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.715994147646205,
+      "learning_rate": 7.594388657822596e-06,
+      "loss": 1.0474,
+      "step": 11715
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.514305533684258,
+      "learning_rate": 7.591880942893234e-06,
+      "loss": 1.0829,
+      "step": 11720
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.13120395760116,
+      "learning_rate": 7.5893723361001135e-06,
+      "loss": 1.0444,
+      "step": 11725
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.178578396505047,
+      "learning_rate": 7.586862838306445e-06,
+      "loss": 1.0816,
+      "step": 11730
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.746854313765677,
+      "learning_rate": 7.5843524503757445e-06,
+      "loss": 0.9941,
+      "step": 11735
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.078364992056647,
+      "learning_rate": 7.581841173171835e-06,
+      "loss": 1.0006,
+      "step": 11740
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.740739798927343,
+      "learning_rate": 7.579329007558843e-06,
+      "loss": 1.0418,
+      "step": 11745
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.539722370722517,
+      "learning_rate": 7.576815954401204e-06,
+      "loss": 1.0112,
+      "step": 11750
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.299343963705468,
+      "learning_rate": 7.574302014563656e-06,
+      "loss": 1.0479,
+      "step": 11755
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 29.875561566756147,
+      "learning_rate": 7.571787188911243e-06,
+      "loss": 1.0701,
+      "step": 11760
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 25.05908382103154,
+      "learning_rate": 7.569271478309315e-06,
+      "loss": 1.0341,
+      "step": 11765
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.601761644761815,
+      "learning_rate": 7.566754883623525e-06,
+      "loss": 1.0087,
+      "step": 11770
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 17.813865904730815,
+      "learning_rate": 7.564237405719831e-06,
+      "loss": 1.004,
+      "step": 11775
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 17.25236748256092,
+      "learning_rate": 7.561719045464495e-06,
+      "loss": 1.081,
+      "step": 11780
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 27.333124687407285,
+      "learning_rate": 7.5591998037240825e-06,
+      "loss": 1.0157,
+      "step": 11785
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 22.089282057926013,
+      "learning_rate": 7.556679681365462e-06,
+      "loss": 1.0206,
+      "step": 11790
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 39.214568937531084,
+      "learning_rate": 7.554158679255805e-06,
+      "loss": 1.0419,
+      "step": 11795
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 26.746819599838243,
+      "learning_rate": 7.551636798262585e-06,
+      "loss": 1.0797,
+      "step": 11800
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 25.734360047448632,
+      "learning_rate": 7.549114039253581e-06,
+      "loss": 1.0184,
+      "step": 11805
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 45.72851511737219,
+      "learning_rate": 7.546590403096871e-06,
+      "loss": 1.0573,
+      "step": 11810
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.937427370671701,
+      "learning_rate": 7.544065890660834e-06,
+      "loss": 1.0428,
+      "step": 11815
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 56.33537705264059,
+      "learning_rate": 7.541540502814154e-06,
+      "loss": 1.0469,
+      "step": 11820
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 33.71330192354937,
+      "learning_rate": 7.539014240425816e-06,
+      "loss": 1.0631,
+      "step": 11825
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.407018181834788,
+      "learning_rate": 7.536487104365102e-06,
+      "loss": 1.0612,
+      "step": 11830
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 29.552370346206306,
+      "learning_rate": 7.533959095501597e-06,
+      "loss": 1.0555,
+      "step": 11835
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.965135513648303,
+      "learning_rate": 7.531430214705189e-06,
+      "loss": 1.0246,
+      "step": 11840
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 37.11431532290853,
+      "learning_rate": 7.528900462846062e-06,
+      "loss": 1.0686,
+      "step": 11845
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.840035222559349,
+      "learning_rate": 7.526369840794699e-06,
+      "loss": 1.0106,
+      "step": 11850
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.281527677159895,
+      "learning_rate": 7.523838349421889e-06,
+      "loss": 1.028,
+      "step": 11855
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.020245250646116,
+      "learning_rate": 7.5213059895987124e-06,
+      "loss": 1.0179,
+      "step": 11860
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.354434431670976,
+      "learning_rate": 7.518772762196553e-06,
+      "loss": 1.0332,
+      "step": 11865
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 25.30003234851166,
+      "learning_rate": 7.516238668087092e-06,
+      "loss": 1.0493,
+      "step": 11870
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.778100727598297,
+      "learning_rate": 7.513703708142309e-06,
+      "loss": 1.0585,
+      "step": 11875
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.594776106824048,
+      "learning_rate": 7.5111678832344806e-06,
+      "loss": 1.0452,
+      "step": 11880
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 19.96266800576846,
+      "learning_rate": 7.508631194236182e-06,
+      "loss": 1.0303,
+      "step": 11885
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.193457442652853,
+      "learning_rate": 7.506093642020286e-06,
+      "loss": 1.0045,
+      "step": 11890
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.817856979567341,
+      "learning_rate": 7.50355522745996e-06,
+      "loss": 1.02,
+      "step": 11895
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.320404713505287,
+      "learning_rate": 7.501015951428673e-06,
+      "loss": 1.0394,
+      "step": 11900
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.896707491370909,
+      "learning_rate": 7.498475814800184e-06,
+      "loss": 1.0197,
+      "step": 11905
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 23.15339531944063,
+      "learning_rate": 7.495934818448555e-06,
+      "loss": 1.0546,
+      "step": 11910
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 21.27469954026058,
+      "learning_rate": 7.493392963248138e-06,
+      "loss": 1.0375,
+      "step": 11915
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.41665952623711,
+      "learning_rate": 7.490850250073585e-06,
+      "loss": 1.0324,
+      "step": 11920
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 20.31198194365891,
+      "learning_rate": 7.488306679799838e-06,
+      "loss": 1.0249,
+      "step": 11925
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.6231025150440415,
+      "learning_rate": 7.4857622533021415e-06,
+      "loss": 1.014,
+      "step": 11930
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.329948465900326,
+      "learning_rate": 7.483216971456027e-06,
+      "loss": 1.0423,
+      "step": 11935
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.185718359298546,
+      "learning_rate": 7.480670835137324e-06,
+      "loss": 1.0605,
+      "step": 11940
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.293068193481355,
+      "learning_rate": 7.478123845222155e-06,
+      "loss": 0.9933,
+      "step": 11945
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 21.82002169605213,
+      "learning_rate": 7.47557600258694e-06,
+      "loss": 1.0965,
+      "step": 11950
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 21.896520825672138,
+      "learning_rate": 7.4730273081083845e-06,
+      "loss": 1.0227,
+      "step": 11955
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.134215363044147,
+      "learning_rate": 7.470477762663495e-06,
+      "loss": 0.9876,
+      "step": 11960
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.438427314214314,
+      "learning_rate": 7.467927367129568e-06,
+      "loss": 1.0533,
+      "step": 11965
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.797106966731714,
+      "learning_rate": 7.4653761223841906e-06,
+      "loss": 1.0289,
+      "step": 11970
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.665798452998185,
+      "learning_rate": 7.462824029305243e-06,
+      "loss": 1.0613,
+      "step": 11975
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.152409531921233,
+      "learning_rate": 7.460271088770902e-06,
+      "loss": 1.0811,
+      "step": 11980
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.889424672209163,
+      "learning_rate": 7.457717301659626e-06,
+      "loss": 1.0123,
+      "step": 11985
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.443733425576188,
+      "learning_rate": 7.455162668850175e-06,
+      "loss": 1.0179,
+      "step": 11990
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.578724563179945,
+      "learning_rate": 7.452607191221596e-06,
+      "loss": 1.043,
+      "step": 11995
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 20.593865569307333,
+      "learning_rate": 7.450050869653224e-06,
+      "loss": 0.9908,
+      "step": 12000
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 42.410456039081524,
+      "learning_rate": 7.4474937050246865e-06,
+      "loss": 1.0648,
+      "step": 12005
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 32.589676758330064,
+      "learning_rate": 7.444935698215905e-06,
+      "loss": 1.0305,
+      "step": 12010
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.916616819955006,
+      "learning_rate": 7.442376850107083e-06,
+      "loss": 1.0317,
+      "step": 12015
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.563970452745615,
+      "learning_rate": 7.439817161578721e-06,
+      "loss": 1.0074,
+      "step": 12020
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.369544319819505,
+      "learning_rate": 7.437256633511603e-06,
+      "loss": 1.0282,
+      "step": 12025
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 81.02214877177478,
+      "learning_rate": 7.434695266786807e-06,
+      "loss": 1.0539,
+      "step": 12030
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 63.362129758959256,
+      "learning_rate": 7.432133062285693e-06,
+      "loss": 1.0339,
+      "step": 12035
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.466818473382034,
+      "learning_rate": 7.429570020889916e-06,
+      "loss": 1.123,
+      "step": 12040
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 117.99242146591413,
+      "learning_rate": 7.4270061434814156e-06,
+      "loss": 1.1451,
+      "step": 12045
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 26.428118691401227,
+      "learning_rate": 7.424441430942418e-06,
+      "loss": 1.062,
+      "step": 12050
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 61.68523063344123,
+      "learning_rate": 7.421875884155442e-06,
+      "loss": 1.0528,
+      "step": 12055
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.310893704321423,
+      "learning_rate": 7.419309504003286e-06,
+      "loss": 1.009,
+      "step": 12060
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.134186371943398,
+      "learning_rate": 7.416742291369041e-06,
+      "loss": 1.092,
+      "step": 12065
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.07378676429573,
+      "learning_rate": 7.414174247136081e-06,
+      "loss": 1.0399,
+      "step": 12070
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.895453691924525,
+      "learning_rate": 7.411605372188068e-06,
+      "loss": 1.057,
+      "step": 12075
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.816493232257297,
+      "learning_rate": 7.409035667408951e-06,
+      "loss": 1.0485,
+      "step": 12080
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.004634418174367,
+      "learning_rate": 7.406465133682961e-06,
+      "loss": 1.0245,
+      "step": 12085
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.262160758749364,
+      "learning_rate": 7.403893771894618e-06,
+      "loss": 1.0168,
+      "step": 12090
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 83.46486853809938,
+      "learning_rate": 7.401321582928722e-06,
+      "loss": 1.0558,
+      "step": 12095
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 45.59415566925905,
+      "learning_rate": 7.398748567670364e-06,
+      "loss": 1.0833,
+      "step": 12100
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.413525481182806,
+      "learning_rate": 7.396174727004915e-06,
+      "loss": 1.0297,
+      "step": 12105
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 36.060354029203495,
+      "learning_rate": 7.393600061818031e-06,
+      "loss": 0.9854,
+      "step": 12110
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.742621477803867,
+      "learning_rate": 7.391024572995652e-06,
+      "loss": 1.0082,
+      "step": 12115
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.899628942463798,
+      "learning_rate": 7.388448261424002e-06,
+      "loss": 1.026,
+      "step": 12120
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.47051525142764,
+      "learning_rate": 7.385871127989584e-06,
+      "loss": 1.0435,
+      "step": 12125
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.032898832474729,
+      "learning_rate": 7.383293173579192e-06,
+      "loss": 1.0847,
+      "step": 12130
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.214473601243663,
+      "learning_rate": 7.380714399079894e-06,
+      "loss": 1.0917,
+      "step": 12135
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.741330271398156,
+      "learning_rate": 7.378134805379046e-06,
+      "loss": 1.0164,
+      "step": 12140
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.721926747859045,
+      "learning_rate": 7.375554393364281e-06,
+      "loss": 1.0236,
+      "step": 12145
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 18.241684309958316,
+      "learning_rate": 7.372973163923521e-06,
+      "loss": 1.0521,
+      "step": 12150
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.768640195269594,
+      "learning_rate": 7.37039111794496e-06,
+      "loss": 1.0152,
+      "step": 12155
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.11180816202643,
+      "learning_rate": 7.3678082563170795e-06,
+      "loss": 1.0445,
+      "step": 12160
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.321856476735322,
+      "learning_rate": 7.36522457992864e-06,
+      "loss": 1.0724,
+      "step": 12165
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.679558986447608,
+      "learning_rate": 7.36264008966868e-06,
+      "loss": 1.0118,
+      "step": 12170
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.763045702610519,
+      "learning_rate": 7.360054786426523e-06,
+      "loss": 1.0341,
+      "step": 12175
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.664202828772289,
+      "learning_rate": 7.357468671091769e-06,
+      "loss": 1.0163,
+      "step": 12180
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.474044749820074,
+      "learning_rate": 7.354881744554295e-06,
+      "loss": 1.0537,
+      "step": 12185
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.806941862617317,
+      "learning_rate": 7.352294007704264e-06,
+      "loss": 1.0285,
+      "step": 12190
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.439091677523939,
+      "learning_rate": 7.349705461432111e-06,
+      "loss": 1.0277,
+      "step": 12195
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.9110356966133075,
+      "learning_rate": 7.347116106628552e-06,
+      "loss": 1.0517,
+      "step": 12200
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.628650555715147,
+      "learning_rate": 7.344525944184583e-06,
+      "loss": 1.0052,
+      "step": 12205
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.59051396975955,
+      "learning_rate": 7.341934974991475e-06,
+      "loss": 1.0597,
+      "step": 12210
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.506905837854545,
+      "learning_rate": 7.339343199940781e-06,
+      "loss": 1.0487,
+      "step": 12215
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 29.591091767578742,
+      "learning_rate": 7.336750619924324e-06,
+      "loss": 1.0879,
+      "step": 12220
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.500775694028768,
+      "learning_rate": 7.3341572358342106e-06,
+      "loss": 0.9942,
+      "step": 12225
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.698100785554281,
+      "learning_rate": 7.331563048562823e-06,
+      "loss": 0.995,
+      "step": 12230
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.678619222102471,
+      "learning_rate": 7.328968059002816e-06,
+      "loss": 1.0036,
+      "step": 12235
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.717058220261013,
+      "learning_rate": 7.326372268047126e-06,
+      "loss": 1.0141,
+      "step": 12240
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 32.58218285532491,
+      "learning_rate": 7.32377567658896e-06,
+      "loss": 1.0866,
+      "step": 12245
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 22.211976172679357,
+      "learning_rate": 7.321178285521803e-06,
+      "loss": 1.048,
+      "step": 12250
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.174005536221587,
+      "learning_rate": 7.318580095739414e-06,
+      "loss": 1.0712,
+      "step": 12255
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.402187876841412,
+      "learning_rate": 7.31598110813583e-06,
+      "loss": 1.0803,
+      "step": 12260
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.961387598255712,
+      "learning_rate": 7.313381323605358e-06,
+      "loss": 1.0511,
+      "step": 12265
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.117317501652218,
+      "learning_rate": 7.310780743042582e-06,
+      "loss": 1.0348,
+      "step": 12270
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.799743957011334,
+      "learning_rate": 7.30817936734236e-06,
+      "loss": 1.0166,
+      "step": 12275
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.5633985549972955,
+      "learning_rate": 7.30557719739982e-06,
+      "loss": 1.0362,
+      "step": 12280
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.630547954930222,
+      "learning_rate": 7.302974234110371e-06,
+      "loss": 1.047,
+      "step": 12285
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.977244441347235,
+      "learning_rate": 7.300370478369687e-06,
+      "loss": 1.0569,
+      "step": 12290
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.767745576845352,
+      "learning_rate": 7.297765931073718e-06,
+      "loss": 1.0362,
+      "step": 12295
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.83769986316688,
+      "learning_rate": 7.295160593118687e-06,
+      "loss": 1.0446,
+      "step": 12300
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.351824272242793,
+      "learning_rate": 7.2925544654010885e-06,
+      "loss": 1.0334,
+      "step": 12305
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.969839115109984,
+      "learning_rate": 7.289947548817687e-06,
+      "loss": 1.0491,
+      "step": 12310
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 17.136217622195936,
+      "learning_rate": 7.287339844265522e-06,
+      "loss": 1.0099,
+      "step": 12315
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.247329863483515,
+      "learning_rate": 7.284731352641903e-06,
+      "loss": 1.0323,
+      "step": 12320
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.78098142765102,
+      "learning_rate": 7.282122074844404e-06,
+      "loss": 1.0202,
+      "step": 12325
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.18913385981386,
+      "learning_rate": 7.2795120117708804e-06,
+      "loss": 1.0699,
+      "step": 12330
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.796562129763549,
+      "learning_rate": 7.276901164319452e-06,
+      "loss": 1.0205,
+      "step": 12335
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.081581545913867,
+      "learning_rate": 7.274289533388505e-06,
+      "loss": 0.9852,
+      "step": 12340
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 5.599993939832297,
+      "learning_rate": 7.2716771198767035e-06,
+      "loss": 1.0266,
+      "step": 12345
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.122890189257696,
+      "learning_rate": 7.269063924682974e-06,
+      "loss": 1.001,
+      "step": 12350
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.483029398431478,
+      "learning_rate": 7.266449948706517e-06,
+      "loss": 1.0569,
+      "step": 12355
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.532734085560354,
+      "learning_rate": 7.263835192846795e-06,
+      "loss": 1.0619,
+      "step": 12360
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.057663962553224,
+      "learning_rate": 7.2612196580035465e-06,
+      "loss": 1.0185,
+      "step": 12365
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.063995399514143,
+      "learning_rate": 7.258603345076773e-06,
+      "loss": 1.0223,
+      "step": 12370
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.492031391508416,
+      "learning_rate": 7.255986254966747e-06,
+      "loss": 1.0152,
+      "step": 12375
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.980889415213772,
+      "learning_rate": 7.253368388574004e-06,
+      "loss": 1.0853,
+      "step": 12380
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.932420119536721,
+      "learning_rate": 7.2507497467993515e-06,
+      "loss": 1.0281,
+      "step": 12385
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.442263891628073,
+      "learning_rate": 7.24813033054386e-06,
+      "loss": 1.0653,
+      "step": 12390
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 19.04890079244909,
+      "learning_rate": 7.2455101407088694e-06,
+      "loss": 1.0582,
+      "step": 12395
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.76606447908117,
+      "learning_rate": 7.242889178195984e-06,
+      "loss": 1.0594,
+      "step": 12400
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.407517494631122,
+      "learning_rate": 7.240267443907074e-06,
+      "loss": 1.0406,
+      "step": 12405
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.844117091500154,
+      "learning_rate": 7.237644938744277e-06,
+      "loss": 1.028,
+      "step": 12410
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.081519520518418,
+      "learning_rate": 7.235021663609995e-06,
+      "loss": 1.04,
+      "step": 12415
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.90657997718108,
+      "learning_rate": 7.232397619406891e-06,
+      "loss": 1.0339,
+      "step": 12420
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.055211671272745,
+      "learning_rate": 7.2297728070379e-06,
+      "loss": 1.0916,
+      "step": 12425
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.292463026715902,
+      "learning_rate": 7.227147227406215e-06,
+      "loss": 1.119,
+      "step": 12430
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.40782131860608,
+      "learning_rate": 7.224520881415296e-06,
+      "loss": 1.0729,
+      "step": 12435
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.861234186251867,
+      "learning_rate": 7.221893769968866e-06,
+      "loss": 1.0636,
+      "step": 12440
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.435769423839574,
+      "learning_rate": 7.219265893970913e-06,
+      "loss": 1.0333,
+      "step": 12445
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.968760917967472,
+      "learning_rate": 7.216637254325685e-06,
+      "loss": 1.0419,
+      "step": 12450
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.53215039683486,
+      "learning_rate": 7.214007851937696e-06,
+      "loss": 1.0314,
+      "step": 12455
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.877044861260606,
+      "learning_rate": 7.21137768771172e-06,
+      "loss": 1.019,
+      "step": 12460
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.287028573377304,
+      "learning_rate": 7.208746762552792e-06,
+      "loss": 0.9964,
+      "step": 12465
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.675474516635187,
+      "learning_rate": 7.206115077366211e-06,
+      "loss": 0.9874,
+      "step": 12470
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.109227137571904,
+      "learning_rate": 7.203482633057542e-06,
+      "loss": 1.0754,
+      "step": 12475
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.617850345781858,
+      "learning_rate": 7.200849430532603e-06,
+      "loss": 1.0351,
+      "step": 12480
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.918311738342016,
+      "learning_rate": 7.198215470697476e-06,
+      "loss": 1.0259,
+      "step": 12485
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.622793750849116,
+      "learning_rate": 7.195580754458506e-06,
+      "loss": 1.0338,
+      "step": 12490
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.284078241578221,
+      "learning_rate": 7.192945282722295e-06,
+      "loss": 0.9951,
+      "step": 12495
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.934648099905459,
+      "learning_rate": 7.190309056395707e-06,
+      "loss": 1.0494,
+      "step": 12500
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 27.260267090753327,
+      "learning_rate": 7.187672076385866e-06,
+      "loss": 1.0635,
+      "step": 12505
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 31.740758025053612,
+      "learning_rate": 7.18503434360015e-06,
+      "loss": 1.0887,
+      "step": 12510
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.88891262912335,
+      "learning_rate": 7.182395858946205e-06,
+      "loss": 1.0174,
+      "step": 12515
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 25.142187362474495,
+      "learning_rate": 7.1797566233319295e-06,
+      "loss": 1.0363,
+      "step": 12520
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.329384696868719,
+      "learning_rate": 7.177116637665481e-06,
+      "loss": 1.0686,
+      "step": 12525
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 25.886128600810267,
+      "learning_rate": 7.174475902855277e-06,
+      "loss": 1.0288,
+      "step": 12530
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 28.974868937037098,
+      "learning_rate": 7.171834419809993e-06,
+      "loss": 1.0078,
+      "step": 12535
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.79249425342462,
+      "learning_rate": 7.169192189438558e-06,
+      "loss": 1.0652,
+      "step": 12540
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.237221329610547,
+      "learning_rate": 7.166549212650162e-06,
+      "loss": 1.018,
+      "step": 12545
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.120941011725753,
+      "learning_rate": 7.163905490354254e-06,
+      "loss": 1.0207,
+      "step": 12550
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 27.040197948817184,
+      "learning_rate": 7.16126102346053e-06,
+      "loss": 1.0338,
+      "step": 12555
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.480407746930139,
+      "learning_rate": 7.158615812878954e-06,
+      "loss": 1.0465,
+      "step": 12560
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.480070729256095,
+      "learning_rate": 7.155969859519739e-06,
+      "loss": 1.0365,
+      "step": 12565
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.25539248289707,
+      "learning_rate": 7.153323164293353e-06,
+      "loss": 1.071,
+      "step": 12570
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.285603597985358,
+      "learning_rate": 7.150675728110525e-06,
+      "loss": 1.024,
+      "step": 12575
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 22.656511072501885,
+      "learning_rate": 7.148027551882232e-06,
+      "loss": 1.0349,
+      "step": 12580
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 25.51686516079161,
+      "learning_rate": 7.14537863651971e-06,
+      "loss": 1.0305,
+      "step": 12585
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 36.296584926767956,
+      "learning_rate": 7.142728982934448e-06,
+      "loss": 1.0582,
+      "step": 12590
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.6989012937694,
+      "learning_rate": 7.140078592038188e-06,
+      "loss": 1.0583,
+      "step": 12595
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.384304551882453,
+      "learning_rate": 7.137427464742931e-06,
+      "loss": 1.0389,
+      "step": 12600
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.185076794077693,
+      "learning_rate": 7.134775601960924e-06,
+      "loss": 0.9996,
+      "step": 12605
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.546886645297697,
+      "learning_rate": 7.1321230046046704e-06,
+      "loss": 1.0152,
+      "step": 12610
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.036397979494588,
+      "learning_rate": 7.129469673586928e-06,
+      "loss": 1.0489,
+      "step": 12615
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.682921144015696,
+      "learning_rate": 7.126815609820705e-06,
+      "loss": 1.0305,
+      "step": 12620
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.658399193206822,
+      "learning_rate": 7.124160814219262e-06,
+      "loss": 1.0618,
+      "step": 12625
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.945032393377005,
+      "learning_rate": 7.1215052876961126e-06,
+      "loss": 1.0678,
+      "step": 12630
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.638107952258533,
+      "learning_rate": 7.118849031165018e-06,
+      "loss": 1.0623,
+      "step": 12635
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.605684454562965,
+      "learning_rate": 7.116192045539996e-06,
+      "loss": 1.0466,
+      "step": 12640
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.750111366604184,
+      "learning_rate": 7.113534331735313e-06,
+      "loss": 1.0526,
+      "step": 12645
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.176878081409408,
+      "learning_rate": 7.110875890665485e-06,
+      "loss": 1.0297,
+      "step": 12650
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.296951539236836,
+      "learning_rate": 7.1082167232452785e-06,
+      "loss": 1.0467,
+      "step": 12655
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.97042566401193,
+      "learning_rate": 7.105556830389713e-06,
+      "loss": 1.0344,
+      "step": 12660
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.3407772774995985,
+      "learning_rate": 7.102896213014051e-06,
+      "loss": 1.0328,
+      "step": 12665
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.607682097535177,
+      "learning_rate": 7.100234872033811e-06,
+      "loss": 1.0044,
+      "step": 12670
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 40.043840619621726,
+      "learning_rate": 7.097572808364759e-06,
+      "loss": 1.0728,
+      "step": 12675
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 27.527787667760947,
+      "learning_rate": 7.094910022922905e-06,
+      "loss": 1.0092,
+      "step": 12680
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.531847109483222,
+      "learning_rate": 7.092246516624513e-06,
+      "loss": 1.0342,
+      "step": 12685
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 75.7379895208296,
+      "learning_rate": 7.089582290386095e-06,
+      "loss": 1.0137,
+      "step": 12690
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 64.10051819695862,
+      "learning_rate": 7.086917345124406e-06,
+      "loss": 1.0882,
+      "step": 12695
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 41.26126004999705,
+      "learning_rate": 7.084251681756451e-06,
+      "loss": 1.0554,
+      "step": 12700
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.337770024854704,
+      "learning_rate": 7.081585301199483e-06,
+      "loss": 1.0163,
+      "step": 12705
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.561604971280826,
+      "learning_rate": 7.078918204371003e-06,
+      "loss": 1.0499,
+      "step": 12710
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.69201828658827,
+      "learning_rate": 7.076250392188752e-06,
+      "loss": 1.0125,
+      "step": 12715
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.476788419263224,
+      "learning_rate": 7.073581865570724e-06,
+      "loss": 1.0923,
+      "step": 12720
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.125474784180392,
+      "learning_rate": 7.070912625435158e-06,
+      "loss": 1.0808,
+      "step": 12725
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.270760693192891,
+      "learning_rate": 7.0682426727005345e-06,
+      "loss": 0.9814,
+      "step": 12730
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.556102509114638,
+      "learning_rate": 7.065572008285584e-06,
+      "loss": 1.0297,
+      "step": 12735
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.606222908410741,
+      "learning_rate": 7.062900633109277e-06,
+      "loss": 1.0731,
+      "step": 12740
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.49691656987773,
+      "learning_rate": 7.060228548090833e-06,
+      "loss": 1.0223,
+      "step": 12745
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.925933039779347,
+      "learning_rate": 7.057555754149713e-06,
+      "loss": 0.9925,
+      "step": 12750
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.128472275615595,
+      "learning_rate": 7.054882252205624e-06,
+      "loss": 1.0409,
+      "step": 12755
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.126108692814256,
+      "learning_rate": 7.052208043178514e-06,
+      "loss": 1.0924,
+      "step": 12760
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.21221054468124,
+      "learning_rate": 7.049533127988576e-06,
+      "loss": 1.0358,
+      "step": 12765
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.1484499062030125,
+      "learning_rate": 7.046857507556247e-06,
+      "loss": 1.0567,
+      "step": 12770
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 38.45334141445022,
+      "learning_rate": 7.0441811828022045e-06,
+      "loss": 1.0411,
+      "step": 12775
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 25.874324500701785,
+      "learning_rate": 7.041504154647369e-06,
+      "loss": 1.0286,
+      "step": 12780
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.270163623871788,
+      "learning_rate": 7.038826424012904e-06,
+      "loss": 1.0301,
+      "step": 12785
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.999079089209222,
+      "learning_rate": 7.036147991820215e-06,
+      "loss": 1.0585,
+      "step": 12790
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 24.206478065357793,
+      "learning_rate": 7.033468858990944e-06,
+      "loss": 1.0118,
+      "step": 12795
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.89523978208532,
+      "learning_rate": 7.030789026446984e-06,
+      "loss": 0.9707,
+      "step": 12800
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.747962327773623,
+      "learning_rate": 7.028108495110457e-06,
+      "loss": 1.0732,
+      "step": 12805
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 23.8698334255745,
+      "learning_rate": 7.025427265903735e-06,
+      "loss": 1.0465,
+      "step": 12810
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.05917059719966,
+      "learning_rate": 7.022745339749426e-06,
+      "loss": 1.0668,
+      "step": 12815
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 18.915197740297916,
+      "learning_rate": 7.020062717570376e-06,
+      "loss": 1.0105,
+      "step": 12820
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.876299249569453,
+      "learning_rate": 7.017379400289675e-06,
+      "loss": 1.0386,
+      "step": 12825
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.659062012775592,
+      "learning_rate": 7.01469538883065e-06,
+      "loss": 1.0328,
+      "step": 12830
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.95405327987556,
+      "learning_rate": 7.012010684116865e-06,
+      "loss": 1.025,
+      "step": 12835
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.551544747776951,
+      "learning_rate": 7.009325287072125e-06,
+      "loss": 1.0814,
+      "step": 12840
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 26.97255684330524,
+      "learning_rate": 7.006639198620473e-06,
+      "loss": 0.9959,
+      "step": 12845
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 37.41006235833254,
+      "learning_rate": 7.00395241968619e-06,
+      "loss": 1.0512,
+      "step": 12850
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 20.749955587173165,
+      "learning_rate": 7.001264951193793e-06,
+      "loss": 1.0193,
+      "step": 12855
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.119022299789663,
+      "learning_rate": 6.998576794068037e-06,
+      "loss": 0.9929,
+      "step": 12860
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.41580671020243,
+      "learning_rate": 6.995887949233917e-06,
+      "loss": 1.0261,
+      "step": 12865
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.30687944722217,
+      "learning_rate": 6.993198417616658e-06,
+      "loss": 1.0262,
+      "step": 12870
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.579780848368085,
+      "learning_rate": 6.990508200141728e-06,
+      "loss": 0.9938,
+      "step": 12875
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.586664085842704,
+      "learning_rate": 6.987817297734828e-06,
+      "loss": 1.0025,
+      "step": 12880
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.894681746357387,
+      "learning_rate": 6.985125711321894e-06,
+      "loss": 1.0483,
+      "step": 12885
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.365915987102852,
+      "learning_rate": 6.982433441829097e-06,
+      "loss": 1.0512,
+      "step": 12890
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.951731138083796,
+      "learning_rate": 6.979740490182849e-06,
+      "loss": 1.0282,
+      "step": 12895
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.217412059486604,
+      "learning_rate": 6.977046857309788e-06,
+      "loss": 1.0322,
+      "step": 12900
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.806719211080662,
+      "learning_rate": 6.974352544136791e-06,
+      "loss": 1.045,
+      "step": 12905
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 5.92757104630876,
+      "learning_rate": 6.971657551590971e-06,
+      "loss": 1.0264,
+      "step": 12910
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.25657133532226,
+      "learning_rate": 6.968961880599668e-06,
+      "loss": 1.0021,
+      "step": 12915
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.516066347033293,
+      "learning_rate": 6.9662655320904636e-06,
+      "loss": 1.0564,
+      "step": 12920
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.063912907532362,
+      "learning_rate": 6.96356850699117e-06,
+      "loss": 1.0061,
+      "step": 12925
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.391986135466741,
+      "learning_rate": 6.960870806229826e-06,
+      "loss": 1.0156,
+      "step": 12930
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.738347589270799,
+      "learning_rate": 6.958172430734711e-06,
+      "loss": 1.0248,
+      "step": 12935
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.753620709514878,
+      "learning_rate": 6.955473381434332e-06,
+      "loss": 1.03,
+      "step": 12940
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 21.557169202777263,
+      "learning_rate": 6.952773659257431e-06,
+      "loss": 1.0125,
+      "step": 12945
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.965167026749218,
+      "learning_rate": 6.95007326513298e-06,
+      "loss": 1.0619,
+      "step": 12950
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.265830292021931,
+      "learning_rate": 6.94737219999018e-06,
+      "loss": 1.0392,
+      "step": 12955
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 23.52081806715469,
+      "learning_rate": 6.9446704647584665e-06,
+      "loss": 1.0245,
+      "step": 12960
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.871420177875759,
+      "learning_rate": 6.9419680603675026e-06,
+      "loss": 0.9975,
+      "step": 12965
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.222317702243371,
+      "learning_rate": 6.9392649877471855e-06,
+      "loss": 1.0536,
+      "step": 12970
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 42.25272051987114,
+      "learning_rate": 6.936561247827635e-06,
+      "loss": 1.0668,
+      "step": 12975
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.740236086629558,
+      "learning_rate": 6.93385684153921e-06,
+      "loss": 1.0125,
+      "step": 12980
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 18.645595937400724,
+      "learning_rate": 6.931151769812496e-06,
+      "loss": 1.0658,
+      "step": 12985
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.141950619913802,
+      "learning_rate": 6.928446033578299e-06,
+      "loss": 1.0382,
+      "step": 12990
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.023119671967667,
+      "learning_rate": 6.925739633767664e-06,
+      "loss": 1.0282,
+      "step": 12995
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 24.068304301841803,
+      "learning_rate": 6.923032571311863e-06,
+      "loss": 1.0067,
+      "step": 13000
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 27.996209165869637,
+      "learning_rate": 6.920324847142388e-06,
+      "loss": 1.0103,
+      "step": 13005
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.417572565792332,
+      "learning_rate": 6.917616462190968e-06,
+      "loss": 1.0201,
+      "step": 13010
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.657853453808597,
+      "learning_rate": 6.914907417389556e-06,
+      "loss": 1.0766,
+      "step": 13015
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.65758461824999,
+      "learning_rate": 6.9121977136703285e-06,
+      "loss": 1.0146,
+      "step": 13020
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.007045939186899,
+      "learning_rate": 6.9094873519656955e-06,
+      "loss": 0.9793,
+      "step": 13025
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.578988873364581,
+      "learning_rate": 6.90677633320829e-06,
+      "loss": 1.0458,
+      "step": 13030
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.951781002259686,
+      "learning_rate": 6.904064658330967e-06,
+      "loss": 1.0555,
+      "step": 13035
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.35412417238803,
+      "learning_rate": 6.901352328266814e-06,
+      "loss": 1.0519,
+      "step": 13040
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.959471511350942,
+      "learning_rate": 6.898639343949141e-06,
+      "loss": 1.0015,
+      "step": 13045
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.821702563866582,
+      "learning_rate": 6.895925706311484e-06,
+      "loss": 1.0324,
+      "step": 13050
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.07143200791322,
+      "learning_rate": 6.893211416287601e-06,
+      "loss": 1.0619,
+      "step": 13055
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.163118330139236,
+      "learning_rate": 6.890496474811478e-06,
+      "loss": 1.0037,
+      "step": 13060
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 22.921157404139223,
+      "learning_rate": 6.887780882817325e-06,
+      "loss": 1.0314,
+      "step": 13065
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 21.488001146763192,
+      "learning_rate": 6.885064641239572e-06,
+      "loss": 1.0426,
+      "step": 13070
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.64297873012137,
+      "learning_rate": 6.882347751012877e-06,
+      "loss": 1.0084,
+      "step": 13075
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.848258797644815,
+      "learning_rate": 6.879630213072119e-06,
+      "loss": 1.0571,
+      "step": 13080
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.16142770751216,
+      "learning_rate": 6.8769120283524e-06,
+      "loss": 0.9836,
+      "step": 13085
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.09570378515828,
+      "learning_rate": 6.874193197789044e-06,
+      "loss": 1.009,
+      "step": 13090
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 22.433742219274283,
+      "learning_rate": 6.8714737223176e-06,
+      "loss": 1.024,
+      "step": 13095
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.383163564586508,
+      "learning_rate": 6.868753602873834e-06,
+      "loss": 1.0429,
+      "step": 13100
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.936132776012842,
+      "learning_rate": 6.866032840393738e-06,
+      "loss": 1.0228,
+      "step": 13105
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.030396329550592,
+      "learning_rate": 6.863311435813525e-06,
+      "loss": 1.0008,
+      "step": 13110
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.397392977750727,
+      "learning_rate": 6.860589390069626e-06,
+      "loss": 1.0177,
+      "step": 13115
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 18.55638265726068,
+      "learning_rate": 6.857866704098695e-06,
+      "loss": 1.0323,
+      "step": 13120
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.559656051710714,
+      "learning_rate": 6.8551433788376066e-06,
+      "loss": 1.0315,
+      "step": 13125
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.641928541047324,
+      "learning_rate": 6.852419415223451e-06,
+      "loss": 1.0941,
+      "step": 13130
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.212436486827265,
+      "learning_rate": 6.849694814193546e-06,
+      "loss": 1.0541,
+      "step": 13135
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.8214796520446495,
+      "learning_rate": 6.846969576685422e-06,
+      "loss": 1.0366,
+      "step": 13140
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.157033991262095,
+      "learning_rate": 6.84424370363683e-06,
+      "loss": 0.9949,
+      "step": 13145
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.45530704947713,
+      "learning_rate": 6.841517195985741e-06,
+      "loss": 1.0393,
+      "step": 13150
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.143450026136613,
+      "learning_rate": 6.838790054670345e-06,
+      "loss": 1.0291,
+      "step": 13155
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.624898335910864,
+      "learning_rate": 6.836062280629046e-06,
+      "loss": 1.0026,
+      "step": 13160
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.84600348155945,
+      "learning_rate": 6.83333387480047e-06,
+      "loss": 1.018,
+      "step": 13165
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 43.80741925650718,
+      "learning_rate": 6.830604838123459e-06,
+      "loss": 1.0479,
+      "step": 13170
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 41.92711636866129,
+      "learning_rate": 6.827875171537071e-06,
+      "loss": 1.0578,
+      "step": 13175
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 19.51673300351281,
+      "learning_rate": 6.8251448759805824e-06,
+      "loss": 1.0711,
+      "step": 13180
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.556515933677582,
+      "learning_rate": 6.8224139523934865e-06,
+      "loss": 1.0446,
+      "step": 13185
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 37.85927373587871,
+      "learning_rate": 6.81968240171549e-06,
+      "loss": 1.0247,
+      "step": 13190
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.662322953416247,
+      "learning_rate": 6.816950224886515e-06,
+      "loss": 1.0237,
+      "step": 13195
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.907269676994511,
+      "learning_rate": 6.814217422846705e-06,
+      "loss": 1.0179,
+      "step": 13200
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 26.99504945572047,
+      "learning_rate": 6.811483996536412e-06,
+      "loss": 1.006,
+      "step": 13205
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.444229002874957,
+      "learning_rate": 6.808749946896206e-06,
+      "loss": 1.0098,
+      "step": 13210
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.473057195996981,
+      "learning_rate": 6.80601527486687e-06,
+      "loss": 1.0325,
+      "step": 13215
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.804280215536732,
+      "learning_rate": 6.8032799813894055e-06,
+      "loss": 1.0363,
+      "step": 13220
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.313675131537839,
+      "learning_rate": 6.800544067405019e-06,
+      "loss": 1.0658,
+      "step": 13225
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.706735278471638,
+      "learning_rate": 6.7978075338551395e-06,
+      "loss": 1.057,
+      "step": 13230
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.756023104740889,
+      "learning_rate": 6.795070381681405e-06,
+      "loss": 1.0309,
+      "step": 13235
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 22.676044196697127,
+      "learning_rate": 6.792332611825667e-06,
+      "loss": 1.0114,
+      "step": 13240
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 26.87964038992477,
+      "learning_rate": 6.789594225229987e-06,
+      "loss": 1.0055,
+      "step": 13245
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.512523423043234,
+      "learning_rate": 6.7868552228366455e-06,
+      "loss": 1.0165,
+      "step": 13250
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 23.665933617476504,
+      "learning_rate": 6.784115605588129e-06,
+      "loss": 1.0666,
+      "step": 13255
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 47.00172811271067,
+      "learning_rate": 6.781375374427135e-06,
+      "loss": 1.0533,
+      "step": 13260
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 50.63177323507287,
+      "learning_rate": 6.778634530296577e-06,
+      "loss": 1.0132,
+      "step": 13265
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.536518733773919,
+      "learning_rate": 6.775893074139575e-06,
+      "loss": 1.0746,
+      "step": 13270
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 29.53133532836771,
+      "learning_rate": 6.773151006899462e-06,
+      "loss": 1.0575,
+      "step": 13275
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 77.40998959421712,
+      "learning_rate": 6.770408329519783e-06,
+      "loss": 1.0598,
+      "step": 13280
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 30.24829150706806,
+      "learning_rate": 6.767665042944287e-06,
+      "loss": 1.0107,
+      "step": 13285
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.192949203694534,
+      "learning_rate": 6.764921148116938e-06,
+      "loss": 1.0108,
+      "step": 13290
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.9501053589897985,
+      "learning_rate": 6.7621766459819095e-06,
+      "loss": 1.0401,
+      "step": 13295
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.111409883932251,
+      "learning_rate": 6.759431537483578e-06,
+      "loss": 1.0443,
+      "step": 13300
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.679607540016216,
+      "learning_rate": 6.756685823566537e-06,
+      "loss": 1.0365,
+      "step": 13305
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.376957592481851,
+      "learning_rate": 6.753939505175581e-06,
+      "loss": 1.0523,
+      "step": 13310
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.391994052623102,
+      "learning_rate": 6.751192583255716e-06,
+      "loss": 1.0234,
+      "step": 13315
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.930521794582923,
+      "learning_rate": 6.748445058752155e-06,
+      "loss": 1.0303,
+      "step": 13320
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.669913265990586,
+      "learning_rate": 6.745696932610322e-06,
+      "loss": 1.0876,
+      "step": 13325
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.146589085411706,
+      "learning_rate": 6.742948205775839e-06,
+      "loss": 1.0389,
+      "step": 13330
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.971442314488811,
+      "learning_rate": 6.740198879194544e-06,
+      "loss": 1.0264,
+      "step": 13335
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.669283894212775,
+      "learning_rate": 6.7374489538124775e-06,
+      "loss": 1.0593,
+      "step": 13340
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.953616910642543,
+      "learning_rate": 6.734698430575885e-06,
+      "loss": 1.0129,
+      "step": 13345
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.523231959616394,
+      "learning_rate": 6.731947310431219e-06,
+      "loss": 1.0135,
+      "step": 13350
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.804972747545044,
+      "learning_rate": 6.7291955943251385e-06,
+      "loss": 1.0272,
+      "step": 13355
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.953005179019751,
+      "learning_rate": 6.726443283204506e-06,
+      "loss": 1.0357,
+      "step": 13360
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.35695685522286,
+      "learning_rate": 6.723690378016387e-06,
+      "loss": 0.9987,
+      "step": 13365
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.64247175046237,
+      "learning_rate": 6.720936879708055e-06,
+      "loss": 0.9996,
+      "step": 13370
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.212790234072378,
+      "learning_rate": 6.718182789226988e-06,
+      "loss": 1.0104,
+      "step": 13375
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.377406455852372,
+      "learning_rate": 6.715428107520864e-06,
+      "loss": 1.0282,
+      "step": 13380
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 16.39394240132829,
+      "learning_rate": 6.712672835537566e-06,
+      "loss": 1.049,
+      "step": 13385
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.035697471945214,
+      "learning_rate": 6.709916974225181e-06,
+      "loss": 1.0153,
+      "step": 13390
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.64975431761266,
+      "learning_rate": 6.707160524532e-06,
+      "loss": 1.0552,
+      "step": 13395
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.294643581025277,
+      "learning_rate": 6.70440348740651e-06,
+      "loss": 1.073,
+      "step": 13400
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.488330345940135,
+      "learning_rate": 6.70164586379741e-06,
+      "loss": 1.0568,
+      "step": 13405
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.870119310854673,
+      "learning_rate": 6.698887654653593e-06,
+      "loss": 1.0285,
+      "step": 13410
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.912999058837164,
+      "learning_rate": 6.6961288609241555e-06,
+      "loss": 1.0441,
+      "step": 13415
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 26.74928942560138,
+      "learning_rate": 6.693369483558399e-06,
+      "loss": 1.0799,
+      "step": 13420
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.9684024815037,
+      "learning_rate": 6.69060952350582e-06,
+      "loss": 1.016,
+      "step": 13425
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.46064987134036,
+      "learning_rate": 6.687848981716118e-06,
+      "loss": 0.9943,
+      "step": 13430
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.169127122095071,
+      "learning_rate": 6.6850878591391945e-06,
+      "loss": 1.0552,
+      "step": 13435
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.165354843115252,
+      "learning_rate": 6.682326156725147e-06,
+      "loss": 1.0508,
+      "step": 13440
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.2432862349534695,
+      "learning_rate": 6.679563875424276e-06,
+      "loss": 1.0393,
+      "step": 13445
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.935217332921848,
+      "learning_rate": 6.67680101618708e-06,
+      "loss": 1.0513,
+      "step": 13450
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.607537224645915,
+      "learning_rate": 6.674037579964256e-06,
+      "loss": 1.0162,
+      "step": 13455
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.120510077099906,
+      "learning_rate": 6.671273567706699e-06,
+      "loss": 1.0126,
+      "step": 13460
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.850181977228456,
+      "learning_rate": 6.668508980365505e-06,
+      "loss": 1.0038,
+      "step": 13465
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.137039326824293,
+      "learning_rate": 6.665743818891963e-06,
+      "loss": 1.0238,
+      "step": 13470
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.516447441123454,
+      "learning_rate": 6.662978084237565e-06,
+      "loss": 1.0241,
+      "step": 13475
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 27.20639222893311,
+      "learning_rate": 6.6602117773539954e-06,
+      "loss": 0.9892,
+      "step": 13480
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.739972267618672,
+      "learning_rate": 6.65744489919314e-06,
+      "loss": 1.0032,
+      "step": 13485
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 21.551587919524792,
+      "learning_rate": 6.654677450707077e-06,
+      "loss": 1.0572,
+      "step": 13490
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.2760398499502985,
+      "learning_rate": 6.651909432848083e-06,
+      "loss": 1.0114,
+      "step": 13495
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.902126713444899,
+      "learning_rate": 6.6491408465686324e-06,
+      "loss": 1.1127,
+      "step": 13500
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.188235790272506,
+      "learning_rate": 6.646371692821391e-06,
+      "loss": 0.9909,
+      "step": 13505
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.421880256038474,
+      "learning_rate": 6.6436019725592215e-06,
+      "loss": 1.0017,
+      "step": 13510
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.010103002828106,
+      "learning_rate": 6.640831686735186e-06,
+      "loss": 1.0216,
+      "step": 13515
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.120440748200015,
+      "learning_rate": 6.638060836302531e-06,
+      "loss": 1.0354,
+      "step": 13520
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.263586848738726,
+      "learning_rate": 6.635289422214708e-06,
+      "loss": 1.0607,
+      "step": 13525
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 16.134589255668278,
+      "learning_rate": 6.632517445425357e-06,
+      "loss": 1.011,
+      "step": 13530
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.013205345946286,
+      "learning_rate": 6.6297449068883125e-06,
+      "loss": 1.0028,
+      "step": 13535
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.790604203729266,
+      "learning_rate": 6.6269718075576e-06,
+      "loss": 1.0237,
+      "step": 13540
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.71188736972473,
+      "learning_rate": 6.624198148387446e-06,
+      "loss": 1.0279,
+      "step": 13545
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.169575350843859,
+      "learning_rate": 6.621423930332258e-06,
+      "loss": 1.0798,
+      "step": 13550
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.090805139004452,
+      "learning_rate": 6.618649154346645e-06,
+      "loss": 1.0083,
+      "step": 13555
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.675451134812084,
+      "learning_rate": 6.615873821385404e-06,
+      "loss": 1.0365,
+      "step": 13560
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.993575443756557,
+      "learning_rate": 6.613097932403524e-06,
+      "loss": 0.9776,
+      "step": 13565
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.518707514073316,
+      "learning_rate": 6.610321488356186e-06,
+      "loss": 1.0298,
+      "step": 13570
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 16.39762762125881,
+      "learning_rate": 6.607544490198763e-06,
+      "loss": 1.011,
+      "step": 13575
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 19.01928544072449,
+      "learning_rate": 6.6047669388868155e-06,
+      "loss": 1.0303,
+      "step": 13580
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 27.189320273451663,
+      "learning_rate": 6.601988835376096e-06,
+      "loss": 1.0292,
+      "step": 13585
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 39.491268467130276,
+      "learning_rate": 6.599210180622551e-06,
+      "loss": 1.0243,
+      "step": 13590
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 26.732490527271786,
+      "learning_rate": 6.5964309755823075e-06,
+      "loss": 1.0122,
+      "step": 13595
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.135663071949276,
+      "learning_rate": 6.593651221211691e-06,
+      "loss": 1.0185,
+      "step": 13600
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.233287153650046,
+      "learning_rate": 6.590870918467214e-06,
+      "loss": 0.971,
+      "step": 13605
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.495907008406722,
+      "learning_rate": 6.588090068305569e-06,
+      "loss": 1.0558,
+      "step": 13610
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 44.42453635403032,
+      "learning_rate": 6.58530867168365e-06,
+      "loss": 0.9835,
+      "step": 13615
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 28.144903089867967,
+      "learning_rate": 6.582526729558533e-06,
+      "loss": 1.0596,
+      "step": 13620
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.024692701567258,
+      "learning_rate": 6.579744242887478e-06,
+      "loss": 1.0064,
+      "step": 13625
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.463744260982136,
+      "learning_rate": 6.576961212627938e-06,
+      "loss": 1.0177,
+      "step": 13630
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.95274300642296,
+      "learning_rate": 6.574177639737551e-06,
+      "loss": 1.0668,
+      "step": 13635
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.42371767681941,
+      "learning_rate": 6.571393525174142e-06,
+      "loss": 1.0396,
+      "step": 13640
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.357186063769456,
+      "learning_rate": 6.568608869895722e-06,
+      "loss": 1.0231,
+      "step": 13645
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.015801963572391,
+      "learning_rate": 6.565823674860487e-06,
+      "loss": 1.0472,
+      "step": 13650
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.069250439384112,
+      "learning_rate": 6.563037941026822e-06,
+      "loss": 0.981,
+      "step": 13655
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 24.130766918714176,
+      "learning_rate": 6.560251669353294e-06,
+      "loss": 1.0042,
+      "step": 13660
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.899136249675347,
+      "learning_rate": 6.557464860798657e-06,
+      "loss": 1.0442,
+      "step": 13665
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.27856661226327,
+      "learning_rate": 6.554677516321849e-06,
+      "loss": 1.026,
+      "step": 13670
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.703392661787985,
+      "learning_rate": 6.551889636881992e-06,
+      "loss": 1.0072,
+      "step": 13675
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.521669786912787,
+      "learning_rate": 6.549101223438394e-06,
+      "loss": 0.9798,
+      "step": 13680
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.132208029869739,
+      "learning_rate": 6.546312276950544e-06,
+      "loss": 1.0406,
+      "step": 13685
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.337966994467242,
+      "learning_rate": 6.5435227983781155e-06,
+      "loss": 1.0336,
+      "step": 13690
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.018133752379166,
+      "learning_rate": 6.540732788680968e-06,
+      "loss": 1.0596,
+      "step": 13695
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.981410299463178,
+      "learning_rate": 6.537942248819139e-06,
+      "loss": 1.052,
+      "step": 13700
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.807726025401056,
+      "learning_rate": 6.53515117975285e-06,
+      "loss": 1.0451,
+      "step": 13705
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.258256090032113,
+      "learning_rate": 6.532359582442509e-06,
+      "loss": 1.0257,
+      "step": 13710
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.830220039221603,
+      "learning_rate": 6.529567457848697e-06,
+      "loss": 1.0574,
+      "step": 13715
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.951624879155915,
+      "learning_rate": 6.526774806932184e-06,
+      "loss": 1.053,
+      "step": 13720
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.10751827537318,
+      "learning_rate": 6.523981630653918e-06,
+      "loss": 1.0139,
+      "step": 13725
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.374129464914586,
+      "learning_rate": 6.521187929975028e-06,
+      "loss": 1.0713,
+      "step": 13730
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.969936644729158,
+      "learning_rate": 6.518393705856826e-06,
+      "loss": 1.0491,
+      "step": 13735
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.11796650989927,
+      "learning_rate": 6.515598959260798e-06,
+      "loss": 1.0005,
+      "step": 13740
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.871941947917126,
+      "learning_rate": 6.512803691148618e-06,
+      "loss": 1.0545,
+      "step": 13745
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.622679603307276,
+      "learning_rate": 6.510007902482132e-06,
+      "loss": 1.051,
+      "step": 13750
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 61.39754193248918,
+      "learning_rate": 6.507211594223367e-06,
+      "loss": 0.9929,
+      "step": 13755
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.488750011595506,
+      "learning_rate": 6.504414767334535e-06,
+      "loss": 0.9972,
+      "step": 13760
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.12429589602954,
+      "learning_rate": 6.501617422778016e-06,
+      "loss": 1.0398,
+      "step": 13765
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.879413296776791,
+      "learning_rate": 6.498819561516376e-06,
+      "loss": 0.9726,
+      "step": 13770
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 23.75657216326405,
+      "learning_rate": 6.4960211845123574e-06,
+      "loss": 1.0689,
+      "step": 13775
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.45006080589518,
+      "learning_rate": 6.493222292728878e-06,
+      "loss": 1.0506,
+      "step": 13780
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.991676780315553,
+      "learning_rate": 6.490422887129034e-06,
+      "loss": 1.0286,
+      "step": 13785
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.18143877800714,
+      "learning_rate": 6.487622968676098e-06,
+      "loss": 1.0247,
+      "step": 13790
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.793232464061429,
+      "learning_rate": 6.48482253833352e-06,
+      "loss": 1.0389,
+      "step": 13795
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.608126922471388,
+      "learning_rate": 6.482021597064923e-06,
+      "loss": 1.0548,
+      "step": 13800
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.140751653201086,
+      "learning_rate": 6.479220145834111e-06,
+      "loss": 1.0124,
+      "step": 13805
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.505552673280925,
+      "learning_rate": 6.47641818560506e-06,
+      "loss": 1.0583,
+      "step": 13810
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.627856677353117,
+      "learning_rate": 6.47361571734192e-06,
+      "loss": 1.0085,
+      "step": 13815
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.41666298648295,
+      "learning_rate": 6.470812742009021e-06,
+      "loss": 0.9837,
+      "step": 13820
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.724829915820955,
+      "learning_rate": 6.468009260570861e-06,
+      "loss": 1.0064,
+      "step": 13825
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.376678289487328,
+      "learning_rate": 6.4652052739921165e-06,
+      "loss": 1.0592,
+      "step": 13830
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.893900152860663,
+      "learning_rate": 6.462400783237638e-06,
+      "loss": 0.9976,
+      "step": 13835
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.700056112704488,
+      "learning_rate": 6.459595789272446e-06,
+      "loss": 0.9963,
+      "step": 13840
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.192370893705413,
+      "learning_rate": 6.456790293061735e-06,
+      "loss": 1.0182,
+      "step": 13845
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.006904290572265,
+      "learning_rate": 6.453984295570879e-06,
+      "loss": 1.038,
+      "step": 13850
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.040487629761007,
+      "learning_rate": 6.451177797765414e-06,
+      "loss": 1.0622,
+      "step": 13855
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.209879361348221,
+      "learning_rate": 6.4483708006110546e-06,
+      "loss": 1.0178,
+      "step": 13860
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 29.12635124130482,
+      "learning_rate": 6.445563305073685e-06,
+      "loss": 0.996,
+      "step": 13865
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 30.88874303744498,
+      "learning_rate": 6.4427553121193665e-06,
+      "loss": 1.0448,
+      "step": 13870
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 31.535720727730226,
+      "learning_rate": 6.4399468227143204e-06,
+      "loss": 1.0054,
+      "step": 13875
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.864907947690293,
+      "learning_rate": 6.437137837824947e-06,
+      "loss": 1.0738,
+      "step": 13880
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 30.114081794271012,
+      "learning_rate": 6.434328358417819e-06,
+      "loss": 1.0386,
+      "step": 13885
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 23.122982889635825,
+      "learning_rate": 6.43151838545967e-06,
+      "loss": 1.0401,
+      "step": 13890
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 45.08967697873334,
+      "learning_rate": 6.428707919917412e-06,
+      "loss": 1.0247,
+      "step": 13895
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.635129933519156,
+      "learning_rate": 6.425896962758126e-06,
+      "loss": 1.0081,
+      "step": 13900
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.054772921652399,
+      "learning_rate": 6.423085514949055e-06,
+      "loss": 1.0032,
+      "step": 13905
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.190495259014995,
+      "learning_rate": 6.420273577457617e-06,
+      "loss": 1.0165,
+      "step": 13910
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 29.90283692177583,
+      "learning_rate": 6.417461151251399e-06,
+      "loss": 1.1068,
+      "step": 13915
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 23.32826749738727,
+      "learning_rate": 6.414648237298151e-06,
+      "loss": 1.0314,
+      "step": 13920
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.451426247775657,
+      "learning_rate": 6.411834836565797e-06,
+      "loss": 1.018,
+      "step": 13925
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.1216586854247375,
+      "learning_rate": 6.409020950022424e-06,
+      "loss": 1.0708,
+      "step": 13930
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.443637085100315,
+      "learning_rate": 6.406206578636288e-06,
+      "loss": 1.0659,
+      "step": 13935
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 20.288942801522673,
+      "learning_rate": 6.403391723375812e-06,
+      "loss": 1.0014,
+      "step": 13940
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.847174012916593,
+      "learning_rate": 6.400576385209583e-06,
+      "loss": 0.9968,
+      "step": 13945
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 27.161722995764585,
+      "learning_rate": 6.397760565106358e-06,
+      "loss": 1.0816,
+      "step": 13950
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.175698198589409,
+      "learning_rate": 6.394944264035057e-06,
+      "loss": 1.0533,
+      "step": 13955
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.558573764246493,
+      "learning_rate": 6.3921274829647685e-06,
+      "loss": 0.9715,
+      "step": 13960
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.821086188227273,
+      "learning_rate": 6.389310222864742e-06,
+      "loss": 1.022,
+      "step": 13965
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.366573020455437,
+      "learning_rate": 6.386492484704394e-06,
+      "loss": 1.0075,
+      "step": 13970
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.897677438642832,
+      "learning_rate": 6.3836742694533084e-06,
+      "loss": 1.0126,
+      "step": 13975
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.255974567583273,
+      "learning_rate": 6.380855578081227e-06,
+      "loss": 1.0278,
+      "step": 13980
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.620443852704517,
+      "learning_rate": 6.378036411558058e-06,
+      "loss": 1.0286,
+      "step": 13985
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.237252298755164,
+      "learning_rate": 6.3752167708538795e-06,
+      "loss": 1.0324,
+      "step": 13990
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.148748108774727,
+      "learning_rate": 6.3723966569389215e-06,
+      "loss": 0.9985,
+      "step": 13995
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.73717599860813,
+      "learning_rate": 6.369576070783585e-06,
+      "loss": 1.0606,
+      "step": 14000
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.626033599591908,
+      "learning_rate": 6.366755013358431e-06,
+      "loss": 1.0858,
+      "step": 14005
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 13.170414197647256,
+      "learning_rate": 6.3639334856341824e-06,
+      "loss": 1.038,
+      "step": 14010
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.102484949514457,
+      "learning_rate": 6.361111488581724e-06,
+      "loss": 0.9809,
+      "step": 14015
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 28.116597064326502,
+      "learning_rate": 6.358289023172102e-06,
+      "loss": 1.0353,
+      "step": 14020
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 38.69362673107948,
+      "learning_rate": 6.355466090376526e-06,
+      "loss": 1.0562,
+      "step": 14025
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 47.16918166965689,
+      "learning_rate": 6.35264269116636e-06,
+      "loss": 1.0332,
+      "step": 14030
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 17.533420893919676,
+      "learning_rate": 6.349818826513137e-06,
+      "loss": 1.0296,
+      "step": 14035
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.442797392178308,
+      "learning_rate": 6.346994497388545e-06,
+      "loss": 1.0713,
+      "step": 14040
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.04787718380431,
+      "learning_rate": 6.344169704764433e-06,
+      "loss": 1.0068,
+      "step": 14045
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.064768698412545,
+      "learning_rate": 6.341344449612811e-06,
+      "loss": 1.0295,
+      "step": 14050
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.302200586498188,
+      "learning_rate": 6.338518732905844e-06,
+      "loss": 1.0236,
+      "step": 14055
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 21.329774898188415,
+      "learning_rate": 6.3356925556158586e-06,
+      "loss": 1.058,
+      "step": 14060
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 29.134880862354656,
+      "learning_rate": 6.332865918715341e-06,
+      "loss": 1.0199,
+      "step": 14065
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 32.23187860799201,
+      "learning_rate": 6.330038823176935e-06,
+      "loss": 1.048,
+      "step": 14070
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 37.799283905948954,
+      "learning_rate": 6.327211269973439e-06,
+      "loss": 1.0262,
+      "step": 14075
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.704421125137417,
+      "learning_rate": 6.324383260077813e-06,
+      "loss": 1.0037,
+      "step": 14080
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.9294961136118,
+      "learning_rate": 6.321554794463173e-06,
+      "loss": 1.0039,
+      "step": 14085
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.850289258704223,
+      "learning_rate": 6.318725874102789e-06,
+      "loss": 1.0635,
+      "step": 14090
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.226221318889067,
+      "learning_rate": 6.315896499970091e-06,
+      "loss": 0.9888,
+      "step": 14095
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.130575799026504,
+      "learning_rate": 6.313066673038666e-06,
+      "loss": 1.0261,
+      "step": 14100
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.480941531356995,
+      "learning_rate": 6.310236394282252e-06,
+      "loss": 1.0188,
+      "step": 14105
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.185087851042856,
+      "learning_rate": 6.307405664674745e-06,
+      "loss": 1.0114,
+      "step": 14110
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.910149014895005,
+      "learning_rate": 6.304574485190199e-06,
+      "loss": 1.0491,
+      "step": 14115
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.739412132695425,
+      "learning_rate": 6.301742856802819e-06,
+      "loss": 0.9905,
+      "step": 14120
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.0777644734152,
+      "learning_rate": 6.298910780486964e-06,
+      "loss": 1.021,
+      "step": 14125
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.950787765281109,
+      "learning_rate": 6.29607825721715e-06,
+      "loss": 1.036,
+      "step": 14130
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.30674289285351,
+      "learning_rate": 6.2932452879680475e-06,
+      "loss": 1.0568,
+      "step": 14135
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.131594426838518,
+      "learning_rate": 6.290411873714475e-06,
+      "loss": 1.0144,
+      "step": 14140
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.093782127641017,
+      "learning_rate": 6.287578015431409e-06,
+      "loss": 0.9836,
+      "step": 14145
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.006982750133712,
+      "learning_rate": 6.284743714093979e-06,
+      "loss": 1.072,
+      "step": 14150
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.524437370717353,
+      "learning_rate": 6.281908970677463e-06,
+      "loss": 1.0174,
+      "step": 14155
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.981786541530653,
+      "learning_rate": 6.279073786157293e-06,
+      "loss": 0.9756,
+      "step": 14160
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.638712127684034,
+      "learning_rate": 6.276238161509058e-06,
+      "loss": 1.0631,
+      "step": 14165
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.830997597851976,
+      "learning_rate": 6.273402097708488e-06,
+      "loss": 1.0142,
+      "step": 14170
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.764014225155503,
+      "learning_rate": 6.270565595731475e-06,
+      "loss": 1.0186,
+      "step": 14175
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.199229045647108,
+      "learning_rate": 6.267728656554053e-06,
+      "loss": 0.9966,
+      "step": 14180
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.008084456261397,
+      "learning_rate": 6.26489128115241e-06,
+      "loss": 1.0303,
+      "step": 14185
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 21.477202925427804,
+      "learning_rate": 6.262053470502886e-06,
+      "loss": 0.9913,
+      "step": 14190
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.387465020085141,
+      "learning_rate": 6.259215225581968e-06,
+      "loss": 1.0221,
+      "step": 14195
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 19.397843353781308,
+      "learning_rate": 6.256376547366294e-06,
+      "loss": 1.0107,
+      "step": 14200
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.135092010729808,
+      "learning_rate": 6.25353743683265e-06,
+      "loss": 1.0878,
+      "step": 14205
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 16.12569114899103,
+      "learning_rate": 6.250697894957971e-06,
+      "loss": 0.9939,
+      "step": 14210
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.956171692699573,
+      "learning_rate": 6.247857922719341e-06,
+      "loss": 1.0283,
+      "step": 14215
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.561231576101571,
+      "learning_rate": 6.24501752109399e-06,
+      "loss": 1.0195,
+      "step": 14220
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.562739598868662,
+      "learning_rate": 6.242176691059302e-06,
+      "loss": 1.0614,
+      "step": 14225
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.959521963283766,
+      "learning_rate": 6.239335433592799e-06,
+      "loss": 1.0354,
+      "step": 14230
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.3662667179511,
+      "learning_rate": 6.2364937496721575e-06,
+      "loss": 1.0758,
+      "step": 14235
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.379812977336815,
+      "learning_rate": 6.233651640275199e-06,
+      "loss": 1.009,
+      "step": 14240
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.883131577198084,
+      "learning_rate": 6.2308091063798895e-06,
+      "loss": 1.0122,
+      "step": 14245
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.14167183566433,
+      "learning_rate": 6.227966148964339e-06,
+      "loss": 1.018,
+      "step": 14250
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.308290140967893,
+      "learning_rate": 6.2251227690068125e-06,
+      "loss": 1.0288,
+      "step": 14255
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.664006828581652,
+      "learning_rate": 6.22227896748571e-06,
+      "loss": 0.9823,
+      "step": 14260
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.49020895969274,
+      "learning_rate": 6.219434745379582e-06,
+      "loss": 1.0258,
+      "step": 14265
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.663898407814036,
+      "learning_rate": 6.216590103667124e-06,
+      "loss": 1.0244,
+      "step": 14270
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.217872837879128,
+      "learning_rate": 6.213745043327172e-06,
+      "loss": 1.0175,
+      "step": 14275
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.949280368143053,
+      "learning_rate": 6.210899565338711e-06,
+      "loss": 1.001,
+      "step": 14280
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 30.094188801468604,
+      "learning_rate": 6.208053670680864e-06,
+      "loss": 1.048,
+      "step": 14285
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 20.347516427833664,
+      "learning_rate": 6.205207360332906e-06,
+      "loss": 0.9853,
+      "step": 14290
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 28.21486187237754,
+      "learning_rate": 6.202360635274243e-06,
+      "loss": 1.0146,
+      "step": 14295
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.837680300255675,
+      "learning_rate": 6.199513496484436e-06,
+      "loss": 0.9886,
+      "step": 14300
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.318146291443392,
+      "learning_rate": 6.19666594494318e-06,
+      "loss": 1.0477,
+      "step": 14305
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.860643957291648,
+      "learning_rate": 6.193817981630314e-06,
+      "loss": 1.0713,
+      "step": 14310
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 17.737601081480946,
+      "learning_rate": 6.190969607525823e-06,
+      "loss": 1.0437,
+      "step": 14315
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.679315730182553,
+      "learning_rate": 6.188120823609826e-06,
+      "loss": 1.0078,
+      "step": 14320
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.776719899109231,
+      "learning_rate": 6.185271630862588e-06,
+      "loss": 1.0335,
+      "step": 14325
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.095041730152499,
+      "learning_rate": 6.1824220302645165e-06,
+      "loss": 1.0284,
+      "step": 14330
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.704614074308678,
+      "learning_rate": 6.179572022796151e-06,
+      "loss": 1.0643,
+      "step": 14335
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.134028494062003,
+      "learning_rate": 6.17672160943818e-06,
+      "loss": 0.9957,
+      "step": 14340
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.781740867825054,
+      "learning_rate": 6.173870791171428e-06,
+      "loss": 1.0073,
+      "step": 14345
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.8775956473541,
+      "learning_rate": 6.171019568976857e-06,
+      "loss": 0.9979,
+      "step": 14350
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.887612818763753,
+      "learning_rate": 6.16816794383557e-06,
+      "loss": 1.0681,
+      "step": 14355
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.619252812637434,
+      "learning_rate": 6.1653159167288115e-06,
+      "loss": 1.0361,
+      "step": 14360
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.324555085616012,
+      "learning_rate": 6.162463488637957e-06,
+      "loss": 1.0245,
+      "step": 14365
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.315834199483762,
+      "learning_rate": 6.159610660544526e-06,
+      "loss": 1.0556,
+      "step": 14370
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.9917183655457675,
+      "learning_rate": 6.156757433430176e-06,
+      "loss": 1.051,
+      "step": 14375
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.626563270273751,
+      "learning_rate": 6.153903808276698e-06,
+      "loss": 1.0146,
+      "step": 14380
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.43592684093022,
+      "learning_rate": 6.151049786066021e-06,
+      "loss": 1.0138,
+      "step": 14385
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.265247039986576,
+      "learning_rate": 6.148195367780211e-06,
+      "loss": 1.0057,
+      "step": 14390
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.458256378539899,
+      "learning_rate": 6.145340554401473e-06,
+      "loss": 1.0076,
+      "step": 14395
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.2129788646983375,
+      "learning_rate": 6.1424853469121436e-06,
+      "loss": 1.0102,
+      "step": 14400
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.904045579179168,
+      "learning_rate": 6.139629746294698e-06,
+      "loss": 1.0414,
+      "step": 14405
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.254639251979803,
+      "learning_rate": 6.1367737535317455e-06,
+      "loss": 1.0348,
+      "step": 14410
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 24.60089000632015,
+      "learning_rate": 6.133917369606028e-06,
+      "loss": 1.0041,
+      "step": 14415
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 25.11777311760514,
+      "learning_rate": 6.131060595500429e-06,
+      "loss": 1.0148,
+      "step": 14420
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.428895668678699,
+      "learning_rate": 6.128203432197959e-06,
+      "loss": 1.0127,
+      "step": 14425
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.744034932615966,
+      "learning_rate": 6.125345880681763e-06,
+      "loss": 0.9718,
+      "step": 14430
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.820188753420734,
+      "learning_rate": 6.1224879419351265e-06,
+      "loss": 1.0571,
+      "step": 14435
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.721814644490256,
+      "learning_rate": 6.119629616941462e-06,
+      "loss": 1.0497,
+      "step": 14440
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.363061957577708,
+      "learning_rate": 6.116770906684315e-06,
+      "loss": 1.0061,
+      "step": 14445
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.604748909394797,
+      "learning_rate": 6.113911812147364e-06,
+      "loss": 1.0434,
+      "step": 14450
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.423518434542537,
+      "learning_rate": 6.1110523343144245e-06,
+      "loss": 0.9749,
+      "step": 14455
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.908629896757577,
+      "learning_rate": 6.108192474169438e-06,
+      "loss": 1.0043,
+      "step": 14460
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.146388089823004,
+      "learning_rate": 6.105332232696478e-06,
+      "loss": 1.023,
+      "step": 14465
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.164722658583353,
+      "learning_rate": 6.1024716108797536e-06,
+      "loss": 1.0099,
+      "step": 14470
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.13629495963452,
+      "learning_rate": 6.0996106097036005e-06,
+      "loss": 1.0635,
+      "step": 14475
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.800462873014633,
+      "learning_rate": 6.096749230152486e-06,
+      "loss": 1.0582,
+      "step": 14480
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.102963227939757,
+      "learning_rate": 6.093887473211011e-06,
+      "loss": 1.0162,
+      "step": 14485
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.403683996342634,
+      "learning_rate": 6.0910253398638986e-06,
+      "loss": 0.9756,
+      "step": 14490
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.1986272105334885,
+      "learning_rate": 6.08816283109601e-06,
+      "loss": 1.0398,
+      "step": 14495
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.499142846058897,
+      "learning_rate": 6.085299947892331e-06,
+      "loss": 1.0325,
+      "step": 14500
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.443651726931194,
+      "learning_rate": 6.082436691237977e-06,
+      "loss": 1.0316,
+      "step": 14505
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.599973887823316,
+      "learning_rate": 6.079573062118192e-06,
+      "loss": 0.9952,
+      "step": 14510
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.28820497848877,
+      "learning_rate": 6.076709061518345e-06,
+      "loss": 0.9898,
+      "step": 14515
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.505366057732097,
+      "learning_rate": 6.07384469042394e-06,
+      "loss": 1.0557,
+      "step": 14520
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.316945997554746,
+      "learning_rate": 6.070979949820601e-06,
+      "loss": 0.9976,
+      "step": 14525
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.461804826693626,
+      "learning_rate": 6.068114840694085e-06,
+      "loss": 1.0098,
+      "step": 14530
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.95271979417951,
+      "learning_rate": 6.065249364030274e-06,
+      "loss": 1.0732,
+      "step": 14535
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.253690343477784,
+      "learning_rate": 6.062383520815173e-06,
+      "loss": 1.0354,
+      "step": 14540
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.420956881537123,
+      "learning_rate": 6.059517312034916e-06,
+      "loss": 1.0241,
+      "step": 14545
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.335665009559412,
+      "learning_rate": 6.056650738675765e-06,
+      "loss": 1.039,
+      "step": 14550
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.074575706713052,
+      "learning_rate": 6.053783801724104e-06,
+      "loss": 1.026,
+      "step": 14555
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.785882460179483,
+      "learning_rate": 6.050916502166444e-06,
+      "loss": 1.0531,
+      "step": 14560
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.30238163595188,
+      "learning_rate": 6.048048840989419e-06,
+      "loss": 1.0576,
+      "step": 14565
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.506627651398785,
+      "learning_rate": 6.045180819179788e-06,
+      "loss": 1.072,
+      "step": 14570
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.198950377669895,
+      "learning_rate": 6.042312437724436e-06,
+      "loss": 1.0364,
+      "step": 14575
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.304318128207594,
+      "learning_rate": 6.03944369761037e-06,
+      "loss": 1.0088,
+      "step": 14580
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.41513228240807,
+      "learning_rate": 6.036574599824719e-06,
+      "loss": 1.0097,
+      "step": 14585
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.723566109044413,
+      "learning_rate": 6.033705145354739e-06,
+      "loss": 0.9864,
+      "step": 14590
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.220611774840002,
+      "learning_rate": 6.0308353351878066e-06,
+      "loss": 1.0636,
+      "step": 14595
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 28.326960131834948,
+      "learning_rate": 6.0279651703114195e-06,
+      "loss": 1.0307,
+      "step": 14600
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 21.15053825545421,
+      "learning_rate": 6.0250946517131995e-06,
+      "loss": 1.0315,
+      "step": 14605
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.95053160025602,
+      "learning_rate": 6.0222237803808895e-06,
+      "loss": 1.0259,
+      "step": 14610
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 23.706207217394876,
+      "learning_rate": 6.01935255730235e-06,
+      "loss": 1.0002,
+      "step": 14615
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.140859339836725,
+      "learning_rate": 6.016480983465572e-06,
+      "loss": 1.0286,
+      "step": 14620
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.605635830489675,
+      "learning_rate": 6.013609059858657e-06,
+      "loss": 1.0206,
+      "step": 14625
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.045248244303234,
+      "learning_rate": 6.010736787469833e-06,
+      "loss": 1.0038,
+      "step": 14630
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.588855130318489,
+      "learning_rate": 6.007864167287447e-06,
+      "loss": 1.034,
+      "step": 14635
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.09002337269585,
+      "learning_rate": 6.004991200299962e-06,
+      "loss": 0.9912,
+      "step": 14640
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 20.41358760134712,
+      "learning_rate": 6.0021178874959665e-06,
+      "loss": 1.0801,
+      "step": 14645
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.779793206949819,
+      "learning_rate": 5.999244229864162e-06,
+      "loss": 1.0057,
+      "step": 14650
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.243732728824146,
+      "learning_rate": 5.996370228393371e-06,
+      "loss": 1.0523,
+      "step": 14655
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.525122559658719,
+      "learning_rate": 5.993495884072538e-06,
+      "loss": 1.0188,
+      "step": 14660
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.96403205642408,
+      "learning_rate": 5.99062119789072e-06,
+      "loss": 1.0091,
+      "step": 14665
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.281253296587717,
+      "learning_rate": 5.987746170837093e-06,
+      "loss": 1.0365,
+      "step": 14670
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.516804753459836,
+      "learning_rate": 5.984870803900953e-06,
+      "loss": 1.0164,
+      "step": 14675
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.549956303475277,
+      "learning_rate": 5.98199509807171e-06,
+      "loss": 1.0113,
+      "step": 14680
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.293675045161462,
+      "learning_rate": 5.979119054338891e-06,
+      "loss": 1.0273,
+      "step": 14685
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.75212362247369,
+      "learning_rate": 5.976242673692141e-06,
+      "loss": 1.0119,
+      "step": 14690
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.283894924775016,
+      "learning_rate": 5.973365957121219e-06,
+      "loss": 1.0087,
+      "step": 14695
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.99199079388364,
+      "learning_rate": 5.970488905616001e-06,
+      "loss": 0.9819,
+      "step": 14700
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.31338495628191,
+      "learning_rate": 5.967611520166479e-06,
+      "loss": 0.9973,
+      "step": 14705
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.28664721508452,
+      "learning_rate": 5.964733801762754e-06,
+      "loss": 1.0251,
+      "step": 14710
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.402366788820295,
+      "learning_rate": 5.961855751395052e-06,
+      "loss": 1.0345,
+      "step": 14715
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.971032257036951,
+      "learning_rate": 5.9589773700537035e-06,
+      "loss": 1.0322,
+      "step": 14720
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.61183751369133,
+      "learning_rate": 5.956098658729157e-06,
+      "loss": 1.0088,
+      "step": 14725
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.953844617439651,
+      "learning_rate": 5.953219618411976e-06,
+      "loss": 1.0001,
+      "step": 14730
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.375623072845399,
+      "learning_rate": 5.950340250092835e-06,
+      "loss": 1.0226,
+      "step": 14735
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.030057617795244,
+      "learning_rate": 5.947460554762522e-06,
+      "loss": 0.9634,
+      "step": 14740
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.009853891348671,
+      "learning_rate": 5.944580533411935e-06,
+      "loss": 1.0118,
+      "step": 14745
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 28.990989603600664,
+      "learning_rate": 5.9417001870320925e-06,
+      "loss": 1.0167,
+      "step": 14750
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.118613775864574,
+      "learning_rate": 5.9388195166141114e-06,
+      "loss": 1.0082,
+      "step": 14755
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.273742102437229,
+      "learning_rate": 5.935938523149234e-06,
+      "loss": 1.0389,
+      "step": 14760
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 39.17306160606416,
+      "learning_rate": 5.933057207628804e-06,
+      "loss": 1.0493,
+      "step": 14765
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 24.024743927701977,
+      "learning_rate": 5.930175571044279e-06,
+      "loss": 0.9837,
+      "step": 14770
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.705609986527344,
+      "learning_rate": 5.927293614387229e-06,
+      "loss": 1.0272,
+      "step": 14775
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 23.230859114470846,
+      "learning_rate": 5.924411338649332e-06,
+      "loss": 1.0593,
+      "step": 14780
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 35.492015860924646,
+      "learning_rate": 5.921528744822376e-06,
+      "loss": 1.0298,
+      "step": 14785
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.562825349710867,
+      "learning_rate": 5.918645833898257e-06,
+      "loss": 1.0212,
+      "step": 14790
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.27956387033416,
+      "learning_rate": 5.915762606868987e-06,
+      "loss": 1.0027,
+      "step": 14795
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 26.21674719448401,
+      "learning_rate": 5.912879064726679e-06,
+      "loss": 0.98,
+      "step": 14800
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.700641274635641,
+      "learning_rate": 5.909995208463555e-06,
+      "loss": 1.0159,
+      "step": 14805
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 25.915543721817727,
+      "learning_rate": 5.907111039071949e-06,
+      "loss": 1.0333,
+      "step": 14810
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.689572006386276,
+      "learning_rate": 5.9042265575443015e-06,
+      "loss": 1.024,
+      "step": 14815
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.410061579336677,
+      "learning_rate": 5.901341764873159e-06,
+      "loss": 1.0091,
+      "step": 14820
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 20.74331671098658,
+      "learning_rate": 5.898456662051175e-06,
+      "loss": 0.9996,
+      "step": 14825
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.260279666494277,
+      "learning_rate": 5.895571250071114e-06,
+      "loss": 1.0082,
+      "step": 14830
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.515962590425934,
+      "learning_rate": 5.892685529925839e-06,
+      "loss": 1.0382,
+      "step": 14835
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 21.656660220190645,
+      "learning_rate": 5.889799502608326e-06,
+      "loss": 1.0119,
+      "step": 14840
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.943941394119102,
+      "learning_rate": 5.886913169111655e-06,
+      "loss": 1.0148,
+      "step": 14845
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.059556966773174,
+      "learning_rate": 5.88402653042901e-06,
+      "loss": 1.0267,
+      "step": 14850
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.230521020102254,
+      "learning_rate": 5.881139587553679e-06,
+      "loss": 1.0459,
+      "step": 14855
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.029114904796708,
+      "learning_rate": 5.878252341479058e-06,
+      "loss": 1.0302,
+      "step": 14860
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.191819892272424,
+      "learning_rate": 5.8753647931986455e-06,
+      "loss": 1.0411,
+      "step": 14865
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.431883483873603,
+      "learning_rate": 5.872476943706043e-06,
+      "loss": 1.0362,
+      "step": 14870
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.225020541416042,
+      "learning_rate": 5.869588793994958e-06,
+      "loss": 1.0278,
+      "step": 14875
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.718840666658366,
+      "learning_rate": 5.8667003450592e-06,
+      "loss": 0.9902,
+      "step": 14880
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.306096075768597,
+      "learning_rate": 5.863811597892679e-06,
+      "loss": 1.0415,
+      "step": 14885
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 17.330385188221122,
+      "learning_rate": 5.860922553489416e-06,
+      "loss": 1.045,
+      "step": 14890
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.898400490052744,
+      "learning_rate": 5.858033212843521e-06,
+      "loss": 1.0119,
+      "step": 14895
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.821759514758661,
+      "learning_rate": 5.855143576949218e-06,
+      "loss": 1.1096,
+      "step": 14900
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.145556238240854,
+      "learning_rate": 5.852253646800829e-06,
+      "loss": 1.0223,
+      "step": 14905
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.538354365516687,
+      "learning_rate": 5.849363423392771e-06,
+      "loss": 1.0003,
+      "step": 14910
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.376499493234194,
+      "learning_rate": 5.846472907719571e-06,
+      "loss": 1.0383,
+      "step": 14915
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.640406948927293,
+      "learning_rate": 5.843582100775854e-06,
+      "loss": 1.0545,
+      "step": 14920
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 41.71062596817106,
+      "learning_rate": 5.84069100355634e-06,
+      "loss": 1.0365,
+      "step": 14925
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 36.440213900509036,
+      "learning_rate": 5.837799617055855e-06,
+      "loss": 1.0287,
+      "step": 14930
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.231312091439358,
+      "learning_rate": 5.834907942269323e-06,
+      "loss": 1.0196,
+      "step": 14935
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.67987282339489,
+      "learning_rate": 5.832015980191763e-06,
+      "loss": 1.0247,
+      "step": 14940
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.027745730413337,
+      "learning_rate": 5.8291237318183e-06,
+      "loss": 1.0395,
+      "step": 14945
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.944458259992423,
+      "learning_rate": 5.826231198144153e-06,
+      "loss": 0.9978,
+      "step": 14950
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.375857405507643,
+      "learning_rate": 5.82333838016464e-06,
+      "loss": 1.0039,
+      "step": 14955
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 29.02177122478957,
+      "learning_rate": 5.820445278875177e-06,
+      "loss": 1.011,
+      "step": 14960
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.960662825766184,
+      "learning_rate": 5.817551895271277e-06,
+      "loss": 1.036,
+      "step": 14965
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.955402914365195,
+      "learning_rate": 5.81465823034855e-06,
+      "loss": 0.9873,
+      "step": 14970
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 19.03775796460291,
+      "learning_rate": 5.811764285102704e-06,
+      "loss": 1.002,
+      "step": 14975
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 20.372957156704015,
+      "learning_rate": 5.8088700605295445e-06,
+      "loss": 1.0141,
+      "step": 14980
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.777200256501984,
+      "learning_rate": 5.805975557624969e-06,
+      "loss": 1.0528,
+      "step": 14985
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 48.67917693510359,
+      "learning_rate": 5.803080777384973e-06,
+      "loss": 1.0117,
+      "step": 14990
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.497038082024737,
+      "learning_rate": 5.800185720805649e-06,
+      "loss": 0.9775,
+      "step": 14995
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 23.166715738992746,
+      "learning_rate": 5.797290388883183e-06,
+      "loss": 0.9993,
+      "step": 15000
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.057863642501305,
+      "learning_rate": 5.794394782613855e-06,
+      "loss": 1.001,
+      "step": 15005
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.830435844917975,
+      "learning_rate": 5.7914989029940414e-06,
+      "loss": 1.0265,
+      "step": 15010
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.212613349302586,
+      "learning_rate": 5.788602751020211e-06,
+      "loss": 1.0209,
+      "step": 15015
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 17.05136078336535,
+      "learning_rate": 5.785706327688929e-06,
+      "loss": 1.0196,
+      "step": 15020
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.07256425645445,
+      "learning_rate": 5.782809633996849e-06,
+      "loss": 0.9761,
+      "step": 15025
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 25.085596348207954,
+      "learning_rate": 5.7799126709407215e-06,
+      "loss": 1.0027,
+      "step": 15030
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.371754516177725,
+      "learning_rate": 5.777015439517389e-06,
+      "loss": 1.0181,
+      "step": 15035
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 5.5333864685009395,
+      "learning_rate": 5.774117940723784e-06,
+      "loss": 1.0047,
+      "step": 15040
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.500730030934907,
+      "learning_rate": 5.771220175556938e-06,
+      "loss": 1.0336,
+      "step": 15045
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.709625588493253,
+      "learning_rate": 5.768322145013964e-06,
+      "loss": 1.0299,
+      "step": 15050
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.338952099488806,
+      "learning_rate": 5.765423850092073e-06,
+      "loss": 1.036,
+      "step": 15055
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 5.169962548517268,
+      "learning_rate": 5.762525291788569e-06,
+      "loss": 0.982,
+      "step": 15060
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 5.900902537411488,
+      "learning_rate": 5.759626471100839e-06,
+      "loss": 0.9946,
+      "step": 15065
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.1707931642179235,
+      "learning_rate": 5.756727389026365e-06,
+      "loss": 1.0523,
+      "step": 15070
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.353224816483072,
+      "learning_rate": 5.753828046562721e-06,
+      "loss": 0.9908,
+      "step": 15075
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.573037894562587,
+      "learning_rate": 5.7509284447075644e-06,
+      "loss": 0.997,
+      "step": 15080
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.837813817381356,
+      "learning_rate": 5.748028584458648e-06,
+      "loss": 1.0719,
+      "step": 15085
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.876634308775737,
+      "learning_rate": 5.745128466813811e-06,
+      "loss": 1.0037,
+      "step": 15090
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.539536593451036,
+      "learning_rate": 5.742228092770978e-06,
+      "loss": 1.0185,
+      "step": 15095
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 33.397164571504376,
+      "learning_rate": 5.739327463328168e-06,
+      "loss": 1.0098,
+      "step": 15100
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 29.183823725475442,
+      "learning_rate": 5.736426579483486e-06,
+      "loss": 1.0048,
+      "step": 15105
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 24.587574963191283,
+      "learning_rate": 5.733525442235119e-06,
+      "loss": 1.0098,
+      "step": 15110
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.947952312742559,
+      "learning_rate": 5.730624052581349e-06,
+      "loss": 1.0179,
+      "step": 15115
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.89284380916069,
+      "learning_rate": 5.727722411520541e-06,
+      "loss": 0.9783,
+      "step": 15120
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.396496552929893,
+      "learning_rate": 5.724820520051145e-06,
+      "loss": 1.0223,
+      "step": 15125
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 23.512104397954204,
+      "learning_rate": 5.7219183791717004e-06,
+      "loss": 0.998,
+      "step": 15130
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.491487800240288,
+      "learning_rate": 5.719015989880831e-06,
+      "loss": 1.0189,
+      "step": 15135
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.884338842606663,
+      "learning_rate": 5.716113353177247e-06,
+      "loss": 1.0263,
+      "step": 15140
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.324532652016312,
+      "learning_rate": 5.713210470059743e-06,
+      "loss": 1.0832,
+      "step": 15145
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.037924639353008,
+      "learning_rate": 5.710307341527196e-06,
+      "loss": 1.0221,
+      "step": 15150
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.74501667455677,
+      "learning_rate": 5.707403968578572e-06,
+      "loss": 1.0076,
+      "step": 15155
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.5396246138440945,
+      "learning_rate": 5.704500352212918e-06,
+      "loss": 1.067,
+      "step": 15160
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.767467544066514,
+      "learning_rate": 5.701596493429363e-06,
+      "loss": 0.9922,
+      "step": 15165
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.52735107254121,
+      "learning_rate": 5.6986923932271285e-06,
+      "loss": 1.0567,
+      "step": 15170
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.0504532995719345,
+      "learning_rate": 5.695788052605507e-06,
+      "loss": 0.9651,
+      "step": 15175
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.310908479402293,
+      "learning_rate": 5.69288347256388e-06,
+      "loss": 1.038,
+      "step": 15180
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.202300147414778,
+      "learning_rate": 5.689978654101715e-06,
+      "loss": 0.9734,
+      "step": 15185
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.184742770785585,
+      "learning_rate": 5.687073598218551e-06,
+      "loss": 1.0026,
+      "step": 15190
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.543331062085483,
+      "learning_rate": 5.68416830591402e-06,
+      "loss": 1.0343,
+      "step": 15195
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 18.40410285840536,
+      "learning_rate": 5.681262778187828e-06,
+      "loss": 1.0112,
+      "step": 15200
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.552560720211156,
+      "learning_rate": 5.678357016039764e-06,
+      "loss": 0.9836,
+      "step": 15205
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.137282707856617,
+      "learning_rate": 5.675451020469699e-06,
+      "loss": 1.0329,
+      "step": 15210
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.269031411588143,
+      "learning_rate": 5.672544792477584e-06,
+      "loss": 0.9694,
+      "step": 15215
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.489519557977571,
+      "learning_rate": 5.669638333063448e-06,
+      "loss": 0.9874,
+      "step": 15220
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.159324332143338,
+      "learning_rate": 5.666731643227399e-06,
+      "loss": 0.9907,
+      "step": 15225
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 18.394066611868105,
+      "learning_rate": 5.663824723969631e-06,
+      "loss": 0.9899,
+      "step": 15230
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 18.369779247598046,
+      "learning_rate": 5.6609175762904075e-06,
+      "loss": 0.9992,
+      "step": 15235
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.373663946948557,
+      "learning_rate": 5.658010201190078e-06,
+      "loss": 1.0451,
+      "step": 15240
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.41526122202614,
+      "learning_rate": 5.655102599669068e-06,
+      "loss": 1.0139,
+      "step": 15245
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.543128116619436,
+      "learning_rate": 5.652194772727878e-06,
+      "loss": 1.0553,
+      "step": 15250
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.985843352021122,
+      "learning_rate": 5.649286721367089e-06,
+      "loss": 1.0342,
+      "step": 15255
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.4618556132775264,
+      "learning_rate": 5.646378446587362e-06,
+      "loss": 1.0222,
+      "step": 15260
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.018828072452745,
+      "learning_rate": 5.643469949389426e-06,
+      "loss": 1.0022,
+      "step": 15265
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 21.408005050217124,
+      "learning_rate": 5.640561230774095e-06,
+      "loss": 1.0343,
+      "step": 15270
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 27.216153751199354,
+      "learning_rate": 5.6376522917422584e-06,
+      "loss": 1.0465,
+      "step": 15275
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.033256161413907,
+      "learning_rate": 5.634743133294877e-06,
+      "loss": 1.0771,
+      "step": 15280
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.203846527256655,
+      "learning_rate": 5.631833756432988e-06,
+      "loss": 1.0364,
+      "step": 15285
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.766700739594345,
+      "learning_rate": 5.628924162157707e-06,
+      "loss": 1.002,
+      "step": 15290
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.70182695862929,
+      "learning_rate": 5.6260143514702245e-06,
+      "loss": 1.0249,
+      "step": 15295
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.402225039197062,
+      "learning_rate": 5.6231043253718e-06,
+      "loss": 1.0014,
+      "step": 15300
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.84456051133836,
+      "learning_rate": 5.620194084863771e-06,
+      "loss": 1.0315,
+      "step": 15305
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.13298034214026,
+      "learning_rate": 5.617283630947553e-06,
+      "loss": 1.0206,
+      "step": 15310
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.80736730732881,
+      "learning_rate": 5.614372964624625e-06,
+      "loss": 1.0183,
+      "step": 15315
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.23622607920322,
+      "learning_rate": 5.6114620868965466e-06,
+      "loss": 1.0323,
+      "step": 15320
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.502210869656103,
+      "learning_rate": 5.608550998764949e-06,
+      "loss": 1.0368,
+      "step": 15325
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.627686206640388,
+      "learning_rate": 5.605639701231533e-06,
+      "loss": 0.9889,
+      "step": 15330
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.981370174837831,
+      "learning_rate": 5.602728195298075e-06,
+      "loss": 0.9927,
+      "step": 15335
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.225736903417539,
+      "learning_rate": 5.5998164819664215e-06,
+      "loss": 1.0384,
+      "step": 15340
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.113727655453436,
+      "learning_rate": 5.596904562238488e-06,
+      "loss": 1.0588,
+      "step": 15345
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.306783354210584,
+      "learning_rate": 5.593992437116264e-06,
+      "loss": 1.0228,
+      "step": 15350
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.44502839539049,
+      "learning_rate": 5.5910801076018115e-06,
+      "loss": 0.9847,
+      "step": 15355
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.849547074253466,
+      "learning_rate": 5.5881675746972585e-06,
+      "loss": 1.0247,
+      "step": 15360
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.89612174609036,
+      "learning_rate": 5.585254839404804e-06,
+      "loss": 0.9802,
+      "step": 15365
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.112813776458878,
+      "learning_rate": 5.582341902726719e-06,
+      "loss": 1.0123,
+      "step": 15370
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.294517000571133,
+      "learning_rate": 5.57942876566534e-06,
+      "loss": 1.007,
+      "step": 15375
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.6312243252068495,
+      "learning_rate": 5.576515429223077e-06,
+      "loss": 0.9867,
+      "step": 15380
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.158982603092975,
+      "learning_rate": 5.5736018944024065e-06,
+      "loss": 0.9994,
+      "step": 15385
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.584956074326355,
+      "learning_rate": 5.5706881622058704e-06,
+      "loss": 1.0102,
+      "step": 15390
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.795599395721448,
+      "learning_rate": 5.567774233636083e-06,
+      "loss": 1.0619,
+      "step": 15395
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.595874741508421,
+      "learning_rate": 5.564860109695726e-06,
+      "loss": 1.0089,
+      "step": 15400
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.457989774283112,
+      "learning_rate": 5.561945791387543e-06,
+      "loss": 0.9862,
+      "step": 15405
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.683225599592505,
+      "learning_rate": 5.55903127971435e-06,
+      "loss": 1.0738,
+      "step": 15410
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.491692917430592,
+      "learning_rate": 5.556116575679028e-06,
+      "loss": 1.0031,
+      "step": 15415
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.804822153310992,
+      "learning_rate": 5.553201680284523e-06,
+      "loss": 1.042,
+      "step": 15420
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.054414839126639,
+      "learning_rate": 5.550286594533848e-06,
+      "loss": 1.0369,
+      "step": 15425
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.795255075267063,
+      "learning_rate": 5.547371319430083e-06,
+      "loss": 1.0149,
+      "step": 15430
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.29537836031942,
+      "learning_rate": 5.544455855976369e-06,
+      "loss": 1.0145,
+      "step": 15435
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.764843303631773,
+      "learning_rate": 5.541540205175914e-06,
+      "loss": 0.9834,
+      "step": 15440
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 24.348022899174037,
+      "learning_rate": 5.538624368031992e-06,
+      "loss": 1.059,
+      "step": 15445
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 42.15851586163241,
+      "learning_rate": 5.535708345547939e-06,
+      "loss": 1.0674,
+      "step": 15450
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 14.398321831145607,
+      "learning_rate": 5.5327921387271565e-06,
+      "loss": 1.0084,
+      "step": 15455
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.41292785937368,
+      "learning_rate": 5.529875748573109e-06,
+      "loss": 0.9958,
+      "step": 15460
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 31.31049057929927,
+      "learning_rate": 5.52695917608932e-06,
+      "loss": 1.0856,
+      "step": 15465
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.824836231611595,
+      "learning_rate": 5.5240424222793836e-06,
+      "loss": 1.0009,
+      "step": 15470
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.923284968966275,
+      "learning_rate": 5.521125488146951e-06,
+      "loss": 1.0319,
+      "step": 15475
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 25.84663898577755,
+      "learning_rate": 5.5182083746957334e-06,
+      "loss": 1.0175,
+      "step": 15480
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.9546403529747,
+      "learning_rate": 5.51529108292951e-06,
+      "loss": 1.0446,
+      "step": 15485
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.7153867003786205,
+      "learning_rate": 5.512373613852117e-06,
+      "loss": 1.0295,
+      "step": 15490
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.35043341931898,
+      "learning_rate": 5.5094559684674545e-06,
+      "loss": 1.0182,
+      "step": 15495
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.7129822787587905,
+      "learning_rate": 5.506538147779478e-06,
+      "loss": 0.9995,
+      "step": 15500
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.9358006177255,
+      "learning_rate": 5.503620152792208e-06,
+      "loss": 0.9772,
+      "step": 15505
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.139260154917016,
+      "learning_rate": 5.500701984509727e-06,
+      "loss": 1.0554,
+      "step": 15510
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.548520446037127,
+      "learning_rate": 5.497783643936169e-06,
+      "loss": 1.0202,
+      "step": 15515
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.310244375627785,
+      "learning_rate": 5.494865132075735e-06,
+      "loss": 0.9979,
+      "step": 15520
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.879052737594693,
+      "learning_rate": 5.491946449932683e-06,
+      "loss": 1.074,
+      "step": 15525
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 25.970468165459245,
+      "learning_rate": 5.489027598511327e-06,
+      "loss": 1.0097,
+      "step": 15530
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.811799979001137,
+      "learning_rate": 5.48610857881604e-06,
+      "loss": 0.9936,
+      "step": 15535
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.571587571136714,
+      "learning_rate": 5.483189391851258e-06,
+      "loss": 0.9873,
+      "step": 15540
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.477857504477589,
+      "learning_rate": 5.480270038621466e-06,
+      "loss": 0.9792,
+      "step": 15545
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.107218873468646,
+      "learning_rate": 5.4773505201312125e-06,
+      "loss": 1.0711,
+      "step": 15550
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.503662574518724,
+      "learning_rate": 5.474430837385102e-06,
+      "loss": 0.9318,
+      "step": 15555
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 23.904668772962744,
+      "learning_rate": 5.471510991387792e-06,
+      "loss": 1.001,
+      "step": 15560
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.82295750301058,
+      "learning_rate": 5.4685909831439995e-06,
+      "loss": 0.9886,
+      "step": 15565
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.798088528044294,
+      "learning_rate": 5.4656708136584994e-06,
+      "loss": 1.0119,
+      "step": 15570
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 18.4122532450321,
+      "learning_rate": 5.462750483936116e-06,
+      "loss": 1.0103,
+      "step": 15575
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.189506784850868,
+      "learning_rate": 5.459829994981732e-06,
+      "loss": 1.0242,
+      "step": 15580
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.985499988081523,
+      "learning_rate": 5.456909347800289e-06,
+      "loss": 1.0173,
+      "step": 15585
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.435487548336896,
+      "learning_rate": 5.453988543396773e-06,
+      "loss": 1.0303,
+      "step": 15590
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.9345012211472,
+      "learning_rate": 5.451067582776233e-06,
+      "loss": 1.0322,
+      "step": 15595
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.203630037662816,
+      "learning_rate": 5.44814646694377e-06,
+      "loss": 1.0119,
+      "step": 15600
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.700017667723494,
+      "learning_rate": 5.445225196904536e-06,
+      "loss": 1.0121,
+      "step": 15605
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.727849581352104,
+      "learning_rate": 5.4423037736637354e-06,
+      "loss": 0.9896,
+      "step": 15610
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.293525246706005,
+      "learning_rate": 5.4393821982266296e-06,
+      "loss": 0.988,
+      "step": 15615
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.412605276421097,
+      "learning_rate": 5.436460471598528e-06,
+      "loss": 1.0571,
+      "step": 15620
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.940978612544219,
+      "learning_rate": 5.433538594784796e-06,
+      "loss": 1.0098,
+      "step": 15625
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.230723437166615,
+      "learning_rate": 5.4306165687908485e-06,
+      "loss": 1.0514,
+      "step": 15630
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.221237771261272,
+      "learning_rate": 5.4276943946221494e-06,
+      "loss": 1.0028,
+      "step": 15635
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.657311547680987,
+      "learning_rate": 5.424772073284218e-06,
+      "loss": 1.0432,
+      "step": 15640
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.577141376755671,
+      "learning_rate": 5.421849605782622e-06,
+      "loss": 0.9769,
+      "step": 15645
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.860599441000725,
+      "learning_rate": 5.418926993122979e-06,
+      "loss": 1.0185,
+      "step": 15650
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.177441626567626,
+      "learning_rate": 5.4160042363109585e-06,
+      "loss": 0.9806,
+      "step": 15655
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.185151466420002,
+      "learning_rate": 5.413081336352278e-06,
+      "loss": 1.0278,
+      "step": 15660
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.775833291063558,
+      "learning_rate": 5.410158294252704e-06,
+      "loss": 1.0019,
+      "step": 15665
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.249542375778276,
+      "learning_rate": 5.4072351110180525e-06,
+      "loss": 1.0343,
+      "step": 15670
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.721745645435944,
+      "learning_rate": 5.404311787654189e-06,
+      "loss": 1.028,
+      "step": 15675
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.587955552947999,
+      "learning_rate": 5.401388325167024e-06,
+      "loss": 1.0212,
+      "step": 15680
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.269292112699187,
+      "learning_rate": 5.39846472456252e-06,
+      "loss": 1.0303,
+      "step": 15685
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.921660261357361,
+      "learning_rate": 5.3955409868466845e-06,
+      "loss": 0.9814,
+      "step": 15690
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.739963976484743,
+      "learning_rate": 5.392617113025576e-06,
+      "loss": 1.0304,
+      "step": 15695
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.29286986619866,
+      "learning_rate": 5.38969310410529e-06,
+      "loss": 1.0466,
+      "step": 15700
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.065776472665721,
+      "learning_rate": 5.386768961091981e-06,
+      "loss": 0.9981,
+      "step": 15705
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.73112924832002,
+      "learning_rate": 5.3838446849918425e-06,
+      "loss": 1.0516,
+      "step": 15710
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.114222624670669,
+      "learning_rate": 5.3809202768111135e-06,
+      "loss": 0.9975,
+      "step": 15715
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 20.69821552756159,
+      "learning_rate": 5.377995737556081e-06,
+      "loss": 1.0127,
+      "step": 15720
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 27.485717435936895,
+      "learning_rate": 5.375071068233077e-06,
+      "loss": 1.042,
+      "step": 15725
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.7777365075493,
+      "learning_rate": 5.372146269848476e-06,
+      "loss": 1.0256,
+      "step": 15730
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.602185252863674,
+      "learning_rate": 5.3692213434086995e-06,
+      "loss": 1.0205,
+      "step": 15735
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.2234703348929346,
+      "learning_rate": 5.366296289920211e-06,
+      "loss": 0.9911,
+      "step": 15740
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.4351426890756,
+      "learning_rate": 5.363371110389519e-06,
+      "loss": 1.0269,
+      "step": 15745
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.786363333398277,
+      "learning_rate": 5.360445805823174e-06,
+      "loss": 0.9598,
+      "step": 15750
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.97525024859164,
+      "learning_rate": 5.357520377227774e-06,
+      "loss": 0.9675,
+      "step": 15755
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.086909538105326,
+      "learning_rate": 5.354594825609952e-06,
+      "loss": 0.9931,
+      "step": 15760
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.159060953047028,
+      "learning_rate": 5.351669151976389e-06,
+      "loss": 1.0362,
+      "step": 15765
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.916611390361543,
+      "learning_rate": 5.348743357333808e-06,
+      "loss": 1.0337,
+      "step": 15770
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.805481262074903,
+      "learning_rate": 5.34581744268897e-06,
+      "loss": 1.0247,
+      "step": 15775
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 26.194657589325068,
+      "learning_rate": 5.34289140904868e-06,
+      "loss": 0.9847,
+      "step": 15780
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.415964782837731,
+      "learning_rate": 5.339965257419784e-06,
+      "loss": 1.0127,
+      "step": 15785
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.071868588891041,
+      "learning_rate": 5.3370389888091675e-06,
+      "loss": 1.0205,
+      "step": 15790
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.9732842913514395,
+      "learning_rate": 5.334112604223757e-06,
+      "loss": 1.0129,
+      "step": 15795
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.997310476911395,
+      "learning_rate": 5.331186104670518e-06,
+      "loss": 0.992,
+      "step": 15800
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.085613485435752,
+      "learning_rate": 5.328259491156458e-06,
+      "loss": 0.9577,
+      "step": 15805
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.168853592166277,
+      "learning_rate": 5.325332764688619e-06,
+      "loss": 0.9916,
+      "step": 15810
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.324216349508182,
+      "learning_rate": 5.322405926274087e-06,
+      "loss": 1.0157,
+      "step": 15815
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.376033806430183,
+      "learning_rate": 5.319478976919984e-06,
+      "loss": 1.0294,
+      "step": 15820
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.579910155353343,
+      "learning_rate": 5.31655191763347e-06,
+      "loss": 1.0061,
+      "step": 15825
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.2035533184641825,
+      "learning_rate": 5.313624749421743e-06,
+      "loss": 1.0234,
+      "step": 15830
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.575174310526426,
+      "learning_rate": 5.31069747329204e-06,
+      "loss": 0.971,
+      "step": 15835
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.968812764711304,
+      "learning_rate": 5.307770090251633e-06,
+      "loss": 1.0269,
+      "step": 15840
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.944910548216217,
+      "learning_rate": 5.3048426013078315e-06,
+      "loss": 1.0075,
+      "step": 15845
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.062487695154594,
+      "learning_rate": 5.301915007467982e-06,
+      "loss": 1.0303,
+      "step": 15850
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.147205587050276,
+      "learning_rate": 5.298987309739467e-06,
+      "loss": 0.9552,
+      "step": 15855
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.092372722717185,
+      "learning_rate": 5.296059509129704e-06,
+      "loss": 1.0003,
+      "step": 15860
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.38372461132063,
+      "learning_rate": 5.293131606646148e-06,
+      "loss": 0.9896,
+      "step": 15865
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.089507838844908,
+      "learning_rate": 5.290203603296285e-06,
+      "loss": 1.0019,
+      "step": 15870
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.69511411954582,
+      "learning_rate": 5.287275500087639e-06,
+      "loss": 1.024,
+      "step": 15875
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.584071369622801,
+      "learning_rate": 5.284347298027769e-06,
+      "loss": 0.9924,
+      "step": 15880
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.611077298524117,
+      "learning_rate": 5.281418998124264e-06,
+      "loss": 1.0062,
+      "step": 15885
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.579750167447928,
+      "learning_rate": 5.278490601384752e-06,
+      "loss": 1.002,
+      "step": 15890
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.858087425142072,
+      "learning_rate": 5.275562108816889e-06,
+      "loss": 1.003,
+      "step": 15895
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.62984739944366,
+      "learning_rate": 5.2726335214283675e-06,
+      "loss": 1.0321,
+      "step": 15900
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.914241867641904,
+      "learning_rate": 5.269704840226911e-06,
+      "loss": 0.9902,
+      "step": 15905
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.49404515966052,
+      "learning_rate": 5.266776066220278e-06,
+      "loss": 1.0109,
+      "step": 15910
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.567038240824985,
+      "learning_rate": 5.2638472004162545e-06,
+      "loss": 1.0079,
+      "step": 15915
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.009631683944291,
+      "learning_rate": 5.260918243822662e-06,
+      "loss": 1.0159,
+      "step": 15920
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.12079307982005,
+      "learning_rate": 5.257989197447352e-06,
+      "loss": 1.0071,
+      "step": 15925
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.507796359581,
+      "learning_rate": 5.255060062298204e-06,
+      "loss": 0.9743,
+      "step": 15930
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.731754541730131,
+      "learning_rate": 5.252130839383133e-06,
+      "loss": 1.0074,
+      "step": 15935
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.589248267275763,
+      "learning_rate": 5.249201529710079e-06,
+      "loss": 1.0055,
+      "step": 15940
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.784506805665072,
+      "learning_rate": 5.24627213428702e-06,
+      "loss": 0.9896,
+      "step": 15945
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.764375372843931,
+      "learning_rate": 5.243342654121953e-06,
+      "loss": 1.007,
+      "step": 15950
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.327918830755468,
+      "learning_rate": 5.24041309022291e-06,
+      "loss": 0.974,
+      "step": 15955
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.952405391443223,
+      "learning_rate": 5.237483443597954e-06,
+      "loss": 0.9841,
+      "step": 15960
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 17.890437029989172,
+      "learning_rate": 5.234553715255171e-06,
+      "loss": 0.9778,
+      "step": 15965
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.01316538586824,
+      "learning_rate": 5.231623906202677e-06,
+      "loss": 1.0248,
+      "step": 15970
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.853916748104757,
+      "learning_rate": 5.228694017448621e-06,
+      "loss": 1.0252,
+      "step": 15975
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.036828572797781,
+      "learning_rate": 5.2257640500011704e-06,
+      "loss": 1.0223,
+      "step": 15980
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.557133960109277,
+      "learning_rate": 5.222834004868527e-06,
+      "loss": 0.9986,
+      "step": 15985
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.190170767718795,
+      "learning_rate": 5.219903883058916e-06,
+      "loss": 0.9722,
+      "step": 15990
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.968707961937117,
+      "learning_rate": 5.216973685580586e-06,
+      "loss": 1.0007,
+      "step": 15995
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.695239415229787,
+      "learning_rate": 5.214043413441819e-06,
+      "loss": 1.0044,
+      "step": 16000
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.32502215343701,
+      "learning_rate": 5.21111306765092e-06,
+      "loss": 1.0325,
+      "step": 16005
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.064153141750366,
+      "learning_rate": 5.208182649216213e-06,
+      "loss": 1.0135,
+      "step": 16010
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 25.031161662044592,
+      "learning_rate": 5.205252159146057e-06,
+      "loss": 0.987,
+      "step": 16015
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 26.71026392221403,
+      "learning_rate": 5.202321598448829e-06,
+      "loss": 1.0106,
+      "step": 16020
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.861465699216564,
+      "learning_rate": 5.1993909681329325e-06,
+      "loss": 1.0096,
+      "step": 16025
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.95376418351282,
+      "learning_rate": 5.196460269206794e-06,
+      "loss": 1.0328,
+      "step": 16030
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.814262958608504,
+      "learning_rate": 5.193529502678865e-06,
+      "loss": 1.0316,
+      "step": 16035
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.929773352393471,
+      "learning_rate": 5.190598669557618e-06,
+      "loss": 1.0064,
+      "step": 16040
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.322807476578738,
+      "learning_rate": 5.187667770851552e-06,
+      "loss": 1.074,
+      "step": 16045
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 19.974955659049545,
+      "learning_rate": 5.184736807569185e-06,
+      "loss": 1.0602,
+      "step": 16050
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.79221091060615,
+      "learning_rate": 5.1818057807190584e-06,
+      "loss": 0.9562,
+      "step": 16055
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.40831543670176,
+      "learning_rate": 5.178874691309736e-06,
+      "loss": 0.9686,
+      "step": 16060
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.394481238822914,
+      "learning_rate": 5.175943540349804e-06,
+      "loss": 1.0069,
+      "step": 16065
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 26.236056559541815,
+      "learning_rate": 5.173012328847867e-06,
+      "loss": 1.0099,
+      "step": 16070
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 32.35843255079721,
+      "learning_rate": 5.170081057812553e-06,
+      "loss": 1.0193,
+      "step": 16075
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 39.151728665016165,
+      "learning_rate": 5.167149728252511e-06,
+      "loss": 1.0183,
+      "step": 16080
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 36.6253507644504,
+      "learning_rate": 5.164218341176405e-06,
+      "loss": 1.0064,
+      "step": 16085
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.653327592026665,
+      "learning_rate": 5.161286897592925e-06,
+      "loss": 1.0032,
+      "step": 16090
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 40.64514818536792,
+      "learning_rate": 5.1583553985107785e-06,
+      "loss": 1.0449,
+      "step": 16095
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.140483573782942,
+      "learning_rate": 5.15542384493869e-06,
+      "loss": 0.9884,
+      "step": 16100
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.083417716052871,
+      "learning_rate": 5.152492237885405e-06,
+      "loss": 0.9715,
+      "step": 16105
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.234421458690278,
+      "learning_rate": 5.149560578359687e-06,
+      "loss": 1.0033,
+      "step": 16110
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 36.12020745998665,
+      "learning_rate": 5.146628867370316e-06,
+      "loss": 0.9946,
+      "step": 16115
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 43.702150113062764,
+      "learning_rate": 5.143697105926092e-06,
+      "loss": 1.0268,
+      "step": 16120
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 26.741514849915372,
+      "learning_rate": 5.140765295035832e-06,
+      "loss": 0.9801,
+      "step": 16125
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 19.76751637787017,
+      "learning_rate": 5.137833435708368e-06,
+      "loss": 1.0012,
+      "step": 16130
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.159552963703967,
+      "learning_rate": 5.13490152895255e-06,
+      "loss": 0.9978,
+      "step": 16135
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 30.926519715727775,
+      "learning_rate": 5.131969575777245e-06,
+      "loss": 0.9883,
+      "step": 16140
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 25.473821960921036,
+      "learning_rate": 5.129037577191335e-06,
+      "loss": 1.0405,
+      "step": 16145
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 28.72089383268889,
+      "learning_rate": 5.126105534203717e-06,
+      "loss": 1.029,
+      "step": 16150
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.484114156842334,
+      "learning_rate": 5.123173447823308e-06,
+      "loss": 1.0108,
+      "step": 16155
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.445040855239327,
+      "learning_rate": 5.120241319059031e-06,
+      "loss": 1.0394,
+      "step": 16160
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 46.8445866363268,
+      "learning_rate": 5.11730914891983e-06,
+      "loss": 1.0269,
+      "step": 16165
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.866166214418564,
+      "learning_rate": 5.114376938414665e-06,
+      "loss": 1.0182,
+      "step": 16170
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.67096279704452,
+      "learning_rate": 5.1114446885525046e-06,
+      "loss": 1.0391,
+      "step": 16175
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.017881971905954,
+      "learning_rate": 5.108512400342332e-06,
+      "loss": 1.0013,
+      "step": 16180
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.040845594445138,
+      "learning_rate": 5.105580074793146e-06,
+      "loss": 0.9716,
+      "step": 16185
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.396904757008385,
+      "learning_rate": 5.102647712913958e-06,
+      "loss": 1.0135,
+      "step": 16190
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.974532267997327,
+      "learning_rate": 5.09971531571379e-06,
+      "loss": 0.9845,
+      "step": 16195
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.373761571462232,
+      "learning_rate": 5.096782884201676e-06,
+      "loss": 0.9918,
+      "step": 16200
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.083849869450802,
+      "learning_rate": 5.093850419386667e-06,
+      "loss": 1.0365,
+      "step": 16205
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.544053197681697,
+      "learning_rate": 5.090917922277815e-06,
+      "loss": 1.0435,
+      "step": 16210
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.317444178392435,
+      "learning_rate": 5.087985393884194e-06,
+      "loss": 0.9998,
+      "step": 16215
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.149487955374466,
+      "learning_rate": 5.0850528352148846e-06,
+      "loss": 0.9818,
+      "step": 16220
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.896241226392624,
+      "learning_rate": 5.082120247278973e-06,
+      "loss": 1.0041,
+      "step": 16225
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.43839173359589,
+      "learning_rate": 5.079187631085564e-06,
+      "loss": 1.0192,
+      "step": 16230
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.759400539341507,
+      "learning_rate": 5.076254987643767e-06,
+      "loss": 0.9977,
+      "step": 16235
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.259653918952248,
+      "learning_rate": 5.073322317962701e-06,
+      "loss": 1.0643,
+      "step": 16240
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.969349385953867,
+      "learning_rate": 5.070389623051496e-06,
+      "loss": 0.9602,
+      "step": 16245
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.913741738579871,
+      "learning_rate": 5.067456903919289e-06,
+      "loss": 0.9951,
+      "step": 16250
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.38234563013056,
+      "learning_rate": 5.0645241615752246e-06,
+      "loss": 0.973,
+      "step": 16255
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.500639363718501,
+      "learning_rate": 5.0615913970284594e-06,
+      "loss": 0.9942,
+      "step": 16260
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.000246455537164,
+      "learning_rate": 5.058658611288153e-06,
+      "loss": 0.9913,
+      "step": 16265
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 25.602213423387873,
+      "learning_rate": 5.055725805363474e-06,
+      "loss": 0.9963,
+      "step": 16270
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.792623271639881,
+      "learning_rate": 5.052792980263598e-06,
+      "loss": 1.0323,
+      "step": 16275
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.58658288712669,
+      "learning_rate": 5.04986013699771e-06,
+      "loss": 1.043,
+      "step": 16280
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 21.70761618039746,
+      "learning_rate": 5.046927276574994e-06,
+      "loss": 1.0099,
+      "step": 16285
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 19.85364796172033,
+      "learning_rate": 5.043994400004648e-06,
+      "loss": 1.0249,
+      "step": 16290
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.427242979146856,
+      "learning_rate": 5.041061508295872e-06,
+      "loss": 1.0583,
+      "step": 16295
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.200635499849241,
+      "learning_rate": 5.038128602457868e-06,
+      "loss": 1.0306,
+      "step": 16300
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.09617656159858,
+      "learning_rate": 5.0351956834998504e-06,
+      "loss": 0.9774,
+      "step": 16305
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.715743797183803,
+      "learning_rate": 5.032262752431031e-06,
+      "loss": 1.0376,
+      "step": 16310
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.819222953107973,
+      "learning_rate": 5.029329810260629e-06,
+      "loss": 1.0369,
+      "step": 16315
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.80103133398059,
+      "learning_rate": 5.026396857997867e-06,
+      "loss": 0.9847,
+      "step": 16320
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.888074167871263,
+      "learning_rate": 5.023463896651972e-06,
+      "loss": 0.9626,
+      "step": 16325
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.705718315442812,
+      "learning_rate": 5.020530927232173e-06,
+      "loss": 0.9942,
+      "step": 16330
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.41260032660405,
+      "learning_rate": 5.017597950747701e-06,
+      "loss": 0.9902,
+      "step": 16335
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.255042843953591,
+      "learning_rate": 5.014664968207791e-06,
+      "loss": 1.073,
+      "step": 16340
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.170200899510636,
+      "learning_rate": 5.01173198062168e-06,
+      "loss": 1.011,
+      "step": 16345
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.790745864427179,
+      "learning_rate": 5.008798988998605e-06,
+      "loss": 1.0245,
+      "step": 16350
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.923644495165293,
+      "learning_rate": 5.005865994347805e-06,
+      "loss": 0.9935,
+      "step": 16355
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.578460220226166,
+      "learning_rate": 5.002932997678524e-06,
+      "loss": 1.0125,
+      "step": 16360
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.753338369687813,
+      "learning_rate": 5e-06,
+      "loss": 0.956,
+      "step": 16365
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.70576280011481,
+      "learning_rate": 4.997067002321478e-06,
+      "loss": 0.9988,
+      "step": 16370
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.817150720143112,
+      "learning_rate": 4.994134005652196e-06,
+      "loss": 0.9805,
+      "step": 16375
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.45500405763143,
+      "learning_rate": 4.991201011001398e-06,
+      "loss": 0.9904,
+      "step": 16380
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.819425098234933,
+      "learning_rate": 4.988268019378322e-06,
+      "loss": 1.0371,
+      "step": 16385
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.1871866152562625,
+      "learning_rate": 4.9853350317922105e-06,
+      "loss": 1.0262,
+      "step": 16390
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 27.438805065280388,
+      "learning_rate": 4.9824020492523e-06,
+      "loss": 1.0582,
+      "step": 16395
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.955853598441074,
+      "learning_rate": 4.979469072767829e-06,
+      "loss": 0.9541,
+      "step": 16400
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.369021538950003,
+      "learning_rate": 4.976536103348029e-06,
+      "loss": 1.0106,
+      "step": 16405
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 17.15791993164313,
+      "learning_rate": 4.973603142002134e-06,
+      "loss": 0.9559,
+      "step": 16410
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.55419850396348,
+      "learning_rate": 4.9706701897393715e-06,
+      "loss": 1.0312,
+      "step": 16415
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.541969961872092,
+      "learning_rate": 4.967737247568971e-06,
+      "loss": 1.048,
+      "step": 16420
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.509974176445898,
+      "learning_rate": 4.964804316500151e-06,
+      "loss": 0.9561,
+      "step": 16425
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.444622668215799,
+      "learning_rate": 4.961871397542133e-06,
+      "loss": 1.0031,
+      "step": 16430
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 20.414380824876183,
+      "learning_rate": 4.95893849170413e-06,
+      "loss": 0.9993,
+      "step": 16435
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.008312874134626,
+      "learning_rate": 4.956005599995354e-06,
+      "loss": 1.0188,
+      "step": 16440
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.690679034928507,
+      "learning_rate": 4.953072723425007e-06,
+      "loss": 1.0049,
+      "step": 16445
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.541363194151192,
+      "learning_rate": 4.950139863002293e-06,
+      "loss": 1.0243,
+      "step": 16450
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.849347533184488,
+      "learning_rate": 4.947207019736403e-06,
+      "loss": 0.9921,
+      "step": 16455
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 15.513672027070383,
+      "learning_rate": 4.944274194636528e-06,
+      "loss": 1.0574,
+      "step": 16460
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.615937012219703,
+      "learning_rate": 4.941341388711849e-06,
+      "loss": 1.0081,
+      "step": 16465
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.633247658658282,
+      "learning_rate": 4.938408602971543e-06,
+      "loss": 0.9871,
+      "step": 16470
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.3010119117651735,
+      "learning_rate": 4.935475838424775e-06,
+      "loss": 0.9994,
+      "step": 16475
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.134598686246772,
+      "learning_rate": 4.932543096080712e-06,
+      "loss": 1.0322,
+      "step": 16480
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.047466822582225,
+      "learning_rate": 4.929610376948505e-06,
+      "loss": 0.9992,
+      "step": 16485
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.05062321545772,
+      "learning_rate": 4.9266776820373e-06,
+      "loss": 1.0001,
+      "step": 16490
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.863782270164636,
+      "learning_rate": 4.923745012356235e-06,
+      "loss": 1.0463,
+      "step": 16495
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.068238812118306,
+      "learning_rate": 4.920812368914438e-06,
+      "loss": 1.0067,
+      "step": 16500
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.288893208069187,
+      "learning_rate": 4.917879752721029e-06,
+      "loss": 1.0203,
+      "step": 16505
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 26.670296805020982,
+      "learning_rate": 4.914947164785118e-06,
+      "loss": 0.9884,
+      "step": 16510
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.97630357097606,
+      "learning_rate": 4.9120146061158084e-06,
+      "loss": 1.0277,
+      "step": 16515
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.184246433267411,
+      "learning_rate": 4.909082077722186e-06,
+      "loss": 0.9844,
+      "step": 16520
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.626723164196851,
+      "learning_rate": 4.906149580613336e-06,
+      "loss": 1.0073,
+      "step": 16525
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.8935410810085385,
+      "learning_rate": 4.903217115798325e-06,
+      "loss": 1.0257,
+      "step": 16530
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.181904513877318,
+      "learning_rate": 4.900284684286213e-06,
+      "loss": 1.0207,
+      "step": 16535
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.826227520357035,
+      "learning_rate": 4.897352287086043e-06,
+      "loss": 1.006,
+      "step": 16540
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.01286550419651,
+      "learning_rate": 4.894419925206856e-06,
+      "loss": 0.9993,
+      "step": 16545
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.8796431874940085,
+      "learning_rate": 4.8914875996576695e-06,
+      "loss": 1.0159,
+      "step": 16550
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.036244702718315,
+      "learning_rate": 4.888555311447498e-06,
+      "loss": 1.0214,
+      "step": 16555
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.9240077190368465,
+      "learning_rate": 4.885623061585337e-06,
+      "loss": 1.0047,
+      "step": 16560
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.430152786339917,
+      "learning_rate": 4.8826908510801715e-06,
+      "loss": 0.9777,
+      "step": 16565
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.585109078393798,
+      "learning_rate": 4.87975868094097e-06,
+      "loss": 1.009,
+      "step": 16570
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.150641168623665,
+      "learning_rate": 4.876826552176695e-06,
+      "loss": 1.0202,
+      "step": 16575
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.935491981122044,
+      "learning_rate": 4.873894465796283e-06,
+      "loss": 0.9867,
+      "step": 16580
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.611867042469141,
+      "learning_rate": 4.8709624228086665e-06,
+      "loss": 1.0331,
+      "step": 16585
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.611371344783553,
+      "learning_rate": 4.868030424222756e-06,
+      "loss": 1.0573,
+      "step": 16590
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 32.63255101680103,
+      "learning_rate": 4.865098471047452e-06,
+      "loss": 0.9927,
+      "step": 16595
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 19.81342606970215,
+      "learning_rate": 4.862166564291633e-06,
+      "loss": 1.0103,
+      "step": 16600
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.499850254147836,
+      "learning_rate": 4.859234704964169e-06,
+      "loss": 1.0191,
+      "step": 16605
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.785974686070636,
+      "learning_rate": 4.856302894073908e-06,
+      "loss": 1.0459,
+      "step": 16610
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.036809757407398,
+      "learning_rate": 4.853371132629685e-06,
+      "loss": 1.019,
+      "step": 16615
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.013408316700486,
+      "learning_rate": 4.8504394216403145e-06,
+      "loss": 0.9935,
+      "step": 16620
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.371096164511118,
+      "learning_rate": 4.8475077621145965e-06,
+      "loss": 1.0574,
+      "step": 16625
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.6340660387069,
+      "learning_rate": 4.844576155061313e-06,
+      "loss": 1.0259,
+      "step": 16630
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.442063040672251,
+      "learning_rate": 4.841644601489222e-06,
+      "loss": 0.9876,
+      "step": 16635
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.244679072108467,
+      "learning_rate": 4.8387131024070775e-06,
+      "loss": 1.0156,
+      "step": 16640
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.89473569238001,
+      "learning_rate": 4.835781658823597e-06,
+      "loss": 1.008,
+      "step": 16645
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.662598637323155,
+      "learning_rate": 4.832850271747493e-06,
+      "loss": 1.043,
+      "step": 16650
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.7050325167017855,
+      "learning_rate": 4.829918942187449e-06,
+      "loss": 0.9853,
+      "step": 16655
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.552096546462334,
+      "learning_rate": 4.826987671152136e-06,
+      "loss": 0.9813,
+      "step": 16660
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.231427286814666,
+      "learning_rate": 4.8240564596501976e-06,
+      "loss": 1.0655,
+      "step": 16665
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.973960826013965,
+      "learning_rate": 4.821125308690267e-06,
+      "loss": 1.0352,
+      "step": 16670
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.115569281328436,
+      "learning_rate": 4.818194219280943e-06,
+      "loss": 0.991,
+      "step": 16675
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 18.37398551239263,
+      "learning_rate": 4.815263192430818e-06,
+      "loss": 1.022,
+      "step": 16680
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.116026171727619,
+      "learning_rate": 4.81233222914845e-06,
+      "loss": 1.0154,
+      "step": 16685
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.88286013001611,
+      "learning_rate": 4.809401330442384e-06,
+      "loss": 0.985,
+      "step": 16690
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.794187658035513,
+      "learning_rate": 4.806470497321135e-06,
+      "loss": 0.9761,
+      "step": 16695
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.911140347032589,
+      "learning_rate": 4.803539730793207e-06,
+      "loss": 1.0205,
+      "step": 16700
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 25.425147897472836,
+      "learning_rate": 4.8006090318670675e-06,
+      "loss": 1.0148,
+      "step": 16705
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.894959658277786,
+      "learning_rate": 4.797678401551172e-06,
+      "loss": 1.046,
+      "step": 16710
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.720123354590075,
+      "learning_rate": 4.794747840853943e-06,
+      "loss": 0.9894,
+      "step": 16715
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.45932942523887,
+      "learning_rate": 4.791817350783788e-06,
+      "loss": 1.0095,
+      "step": 16720
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.482267961424742,
+      "learning_rate": 4.7888869323490805e-06,
+      "loss": 1.0267,
+      "step": 16725
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.750108166742298,
+      "learning_rate": 4.785956586558182e-06,
+      "loss": 1.01,
+      "step": 16730
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.196071199273943,
+      "learning_rate": 4.783026314419414e-06,
+      "loss": 1.0019,
+      "step": 16735
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.908790043567544,
+      "learning_rate": 4.780096116941087e-06,
+      "loss": 1.0099,
+      "step": 16740
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.650041814165618,
+      "learning_rate": 4.777165995131473e-06,
+      "loss": 1.0018,
+      "step": 16745
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 17.682189805974183,
+      "learning_rate": 4.77423594999883e-06,
+      "loss": 1.0463,
+      "step": 16750
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.3288850258912115,
+      "learning_rate": 4.771305982551381e-06,
+      "loss": 0.9903,
+      "step": 16755
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 26.904240163387783,
+      "learning_rate": 4.7683760937973235e-06,
+      "loss": 1.0455,
+      "step": 16760
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 40.74926258668739,
+      "learning_rate": 4.7654462847448316e-06,
+      "loss": 1.0471,
+      "step": 16765
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.09515246288266,
+      "learning_rate": 4.762516556402048e-06,
+      "loss": 0.9861,
+      "step": 16770
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 23.91326969614342,
+      "learning_rate": 4.759586909777092e-06,
+      "loss": 0.9822,
+      "step": 16775
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.47668710150236,
+      "learning_rate": 4.75665734587805e-06,
+      "loss": 1.0411,
+      "step": 16780
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 25.25118285825452,
+      "learning_rate": 4.753727865712983e-06,
+      "loss": 0.9719,
+      "step": 16785
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.448283026301109,
+      "learning_rate": 4.750798470289922e-06,
+      "loss": 1.0017,
+      "step": 16790
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.430106955844881,
+      "learning_rate": 4.74786916061687e-06,
+      "loss": 0.973,
+      "step": 16795
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.983024090698553,
+      "learning_rate": 4.744939937701797e-06,
+      "loss": 1.0874,
+      "step": 16800
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 48.062147831281614,
+      "learning_rate": 4.74201080255265e-06,
+      "loss": 0.9621,
+      "step": 16805
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.270595501676446,
+      "learning_rate": 4.739081756177339e-06,
+      "loss": 0.9702,
+      "step": 16810
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.539937937696077,
+      "learning_rate": 4.736152799583746e-06,
+      "loss": 0.9727,
+      "step": 16815
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.640223121696199,
+      "learning_rate": 4.733223933779723e-06,
+      "loss": 1.0139,
+      "step": 16820
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.791427548679868,
+      "learning_rate": 4.73029515977309e-06,
+      "loss": 1.0287,
+      "step": 16825
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 20.742115000163974,
+      "learning_rate": 4.727366478571633e-06,
+      "loss": 1.0327,
+      "step": 16830
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.944895323419959,
+      "learning_rate": 4.724437891183112e-06,
+      "loss": 0.9897,
+      "step": 16835
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.526581874208533,
+      "learning_rate": 4.721509398615249e-06,
+      "loss": 0.9737,
+      "step": 16840
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.652363648702273,
+      "learning_rate": 4.718581001875737e-06,
+      "loss": 1.0013,
+      "step": 16845
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.278136771376039,
+      "learning_rate": 4.715652701972233e-06,
+      "loss": 0.9977,
+      "step": 16850
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.61950884273999,
+      "learning_rate": 4.712724499912362e-06,
+      "loss": 1.0335,
+      "step": 16855
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.070495461354499,
+      "learning_rate": 4.709796396703715e-06,
+      "loss": 0.942,
+      "step": 16860
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.222729282641115,
+      "learning_rate": 4.706868393353854e-06,
+      "loss": 0.9613,
+      "step": 16865
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.940206553289965,
+      "learning_rate": 4.703940490870296e-06,
+      "loss": 1.0139,
+      "step": 16870
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.654456357112275,
+      "learning_rate": 4.701012690260534e-06,
+      "loss": 1.0285,
+      "step": 16875
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.150880433070066,
+      "learning_rate": 4.69808499253202e-06,
+      "loss": 0.9728,
+      "step": 16880
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.347270980874507,
+      "learning_rate": 4.695157398692171e-06,
+      "loss": 1.0049,
+      "step": 16885
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.608674093571464,
+      "learning_rate": 4.69222990974837e-06,
+      "loss": 0.9961,
+      "step": 16890
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.716642295325675,
+      "learning_rate": 4.689302526707961e-06,
+      "loss": 0.9758,
+      "step": 16895
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 22.581713134228753,
+      "learning_rate": 4.686375250578259e-06,
+      "loss": 0.9866,
+      "step": 16900
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.87550070685061,
+      "learning_rate": 4.683448082366532e-06,
+      "loss": 1.0443,
+      "step": 16905
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.306310736277971,
+      "learning_rate": 4.680521023080018e-06,
+      "loss": 1.0307,
+      "step": 16910
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.584978128407352,
+      "learning_rate": 4.677594073725915e-06,
+      "loss": 1.0003,
+      "step": 16915
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.88851693358926,
+      "learning_rate": 4.674667235311384e-06,
+      "loss": 1.0235,
+      "step": 16920
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 30.054488713226664,
+      "learning_rate": 4.6717405088435445e-06,
+      "loss": 0.9955,
+      "step": 16925
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.621390724593681,
+      "learning_rate": 4.668813895329483e-06,
+      "loss": 1.0425,
+      "step": 16930
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.266159283078437,
+      "learning_rate": 4.6658873957762445e-06,
+      "loss": 1.0256,
+      "step": 16935
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.748902310168273,
+      "learning_rate": 4.662961011190835e-06,
+      "loss": 0.9987,
+      "step": 16940
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.719847757693675,
+      "learning_rate": 4.660034742580218e-06,
+      "loss": 0.9898,
+      "step": 16945
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.953366372779367,
+      "learning_rate": 4.657108590951323e-06,
+      "loss": 0.9953,
+      "step": 16950
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.829895806922655,
+      "learning_rate": 4.654182557311031e-06,
+      "loss": 1.0197,
+      "step": 16955
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.987562304626971,
+      "learning_rate": 4.651256642666194e-06,
+      "loss": 1.0042,
+      "step": 16960
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 22.439513059825465,
+      "learning_rate": 4.648330848023611e-06,
+      "loss": 1.0018,
+      "step": 16965
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.647793940331765,
+      "learning_rate": 4.645405174390049e-06,
+      "loss": 1.0185,
+      "step": 16970
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.646874515262727,
+      "learning_rate": 4.642479622772227e-06,
+      "loss": 0.9786,
+      "step": 16975
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.023487564809672,
+      "learning_rate": 4.639554194176827e-06,
+      "loss": 1.0247,
+      "step": 16980
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.715892801327604,
+      "learning_rate": 4.636628889610481e-06,
+      "loss": 1.0001,
+      "step": 16985
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 18.11156588494887,
+      "learning_rate": 4.63370371007979e-06,
+      "loss": 1.0271,
+      "step": 16990
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 22.52875419448177,
+      "learning_rate": 4.630778656591301e-06,
+      "loss": 0.9778,
+      "step": 16995
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.726171691868798,
+      "learning_rate": 4.6278537301515256e-06,
+      "loss": 0.967,
+      "step": 17000
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 23.2971827059768,
+      "learning_rate": 4.624928931766924e-06,
+      "loss": 1.0191,
+      "step": 17005
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.843956527421028,
+      "learning_rate": 4.62200426244392e-06,
+      "loss": 1.0643,
+      "step": 17010
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 28.33813073646957,
+      "learning_rate": 4.619079723188889e-06,
+      "loss": 1.001,
+      "step": 17015
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 26.971320195799127,
+      "learning_rate": 4.616155315008159e-06,
+      "loss": 1.0223,
+      "step": 17020
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.941925646816498,
+      "learning_rate": 4.6132310389080205e-06,
+      "loss": 1.0409,
+      "step": 17025
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.057853691194898,
+      "learning_rate": 4.610306895894711e-06,
+      "loss": 1.0512,
+      "step": 17030
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.217344829324652,
+      "learning_rate": 4.607382886974428e-06,
+      "loss": 0.9383,
+      "step": 17035
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.626275290185591,
+      "learning_rate": 4.604459013153316e-06,
+      "loss": 0.9898,
+      "step": 17040
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.647514563751912,
+      "learning_rate": 4.601535275437482e-06,
+      "loss": 0.987,
+      "step": 17045
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.78296404389408,
+      "learning_rate": 4.598611674832977e-06,
+      "loss": 1.033,
+      "step": 17050
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.24405677217857,
+      "learning_rate": 4.595688212345814e-06,
+      "loss": 1.022,
+      "step": 17055
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.862725670052217,
+      "learning_rate": 4.592764888981948e-06,
+      "loss": 1.0246,
+      "step": 17060
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.559220803954123,
+      "learning_rate": 4.589841705747298e-06,
+      "loss": 0.9701,
+      "step": 17065
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.9863393651332055,
+      "learning_rate": 4.586918663647723e-06,
+      "loss": 1.0066,
+      "step": 17070
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.517380585020579,
+      "learning_rate": 4.583995763689043e-06,
+      "loss": 0.994,
+      "step": 17075
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.356856092450435,
+      "learning_rate": 4.581073006877021e-06,
+      "loss": 0.9981,
+      "step": 17080
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.891020181152935,
+      "learning_rate": 4.578150394217379e-06,
+      "loss": 0.9576,
+      "step": 17085
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.722072676995166,
+      "learning_rate": 4.575227926715783e-06,
+      "loss": 0.9896,
+      "step": 17090
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.102300230727144,
+      "learning_rate": 4.572305605377852e-06,
+      "loss": 0.9656,
+      "step": 17095
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.58749119120385,
+      "learning_rate": 4.569383431209153e-06,
+      "loss": 0.9775,
+      "step": 17100
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.591727232593307,
+      "learning_rate": 4.5664614052152056e-06,
+      "loss": 0.9955,
+      "step": 17105
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.519427533241876,
+      "learning_rate": 4.5635395284014714e-06,
+      "loss": 0.9577,
+      "step": 17110
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.03651336287879,
+      "learning_rate": 4.560617801773371e-06,
+      "loss": 0.9762,
+      "step": 17115
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.72300338152943,
+      "learning_rate": 4.557696226336265e-06,
+      "loss": 0.9808,
+      "step": 17120
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.967811199066373,
+      "learning_rate": 4.554774803095467e-06,
+      "loss": 0.9577,
+      "step": 17125
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 20.01930160997758,
+      "learning_rate": 4.551853533056231e-06,
+      "loss": 0.9873,
+      "step": 17130
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.00087246915051,
+      "learning_rate": 4.548932417223768e-06,
+      "loss": 0.9588,
+      "step": 17135
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 18.4835082019977,
+      "learning_rate": 4.546011456603229e-06,
+      "loss": 1.0342,
+      "step": 17140
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.10114229581885,
+      "learning_rate": 4.543090652199713e-06,
+      "loss": 0.9772,
+      "step": 17145
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.719237844124875,
+      "learning_rate": 4.540170005018269e-06,
+      "loss": 0.9921,
+      "step": 17150
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 38.72733323117287,
+      "learning_rate": 4.537249516063886e-06,
+      "loss": 1.0065,
+      "step": 17155
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 18.655554225283357,
+      "learning_rate": 4.534329186341503e-06,
+      "loss": 1.0126,
+      "step": 17160
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.583481223929379,
+      "learning_rate": 4.531409016856001e-06,
+      "loss": 1.0153,
+      "step": 17165
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.630329560996714,
+      "learning_rate": 4.528489008612211e-06,
+      "loss": 0.9809,
+      "step": 17170
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 19.1968269463294,
+      "learning_rate": 4.5255691626149e-06,
+      "loss": 1.0535,
+      "step": 17175
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.514058147973323,
+      "learning_rate": 4.52264947986879e-06,
+      "loss": 1.0545,
+      "step": 17180
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 37.06831696340187,
+      "learning_rate": 4.5197299613785355e-06,
+      "loss": 0.9911,
+      "step": 17185
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 18.270934411108115,
+      "learning_rate": 4.516810608148744e-06,
+      "loss": 1.0212,
+      "step": 17190
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.937408149311286,
+      "learning_rate": 4.5138914211839605e-06,
+      "loss": 0.9908,
+      "step": 17195
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 13.308259612931357,
+      "learning_rate": 4.510972401488675e-06,
+      "loss": 1.0033,
+      "step": 17200
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.430432297635973,
+      "learning_rate": 4.5080535500673175e-06,
+      "loss": 1.0123,
+      "step": 17205
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.6980510237613915,
+      "learning_rate": 4.505134867924266e-06,
+      "loss": 1.0147,
+      "step": 17210
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.142734641134295,
+      "learning_rate": 4.502216356063831e-06,
+      "loss": 0.9704,
+      "step": 17215
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.755769632860416,
+      "learning_rate": 4.499298015490275e-06,
+      "loss": 0.9695,
+      "step": 17220
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.707542636435186,
+      "learning_rate": 4.496379847207792e-06,
+      "loss": 1.0009,
+      "step": 17225
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.107552234617984,
+      "learning_rate": 4.493461852220524e-06,
+      "loss": 1.0208,
+      "step": 17230
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.540283502674953,
+      "learning_rate": 4.490544031532546e-06,
+      "loss": 0.9819,
+      "step": 17235
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.8252134143751455,
+      "learning_rate": 4.487626386147884e-06,
+      "loss": 1.0146,
+      "step": 17240
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.68033910105838,
+      "learning_rate": 4.48470891707049e-06,
+      "loss": 1.0048,
+      "step": 17245
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.716785015753265,
+      "learning_rate": 4.481791625304267e-06,
+      "loss": 1.0389,
+      "step": 17250
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.748595545093872,
+      "learning_rate": 4.478874511853051e-06,
+      "loss": 0.9763,
+      "step": 17255
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.787979772469156,
+      "learning_rate": 4.475957577720617e-06,
+      "loss": 0.9967,
+      "step": 17260
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.623623483227947,
+      "learning_rate": 4.473040823910681e-06,
+      "loss": 1.051,
+      "step": 17265
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.899670736913321,
+      "learning_rate": 4.4701242514268925e-06,
+      "loss": 1.0022,
+      "step": 17270
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.208714786916086,
+      "learning_rate": 4.467207861272846e-06,
+      "loss": 1.0399,
+      "step": 17275
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 22.947292874334988,
+      "learning_rate": 4.464291654452062e-06,
+      "loss": 1.0067,
+      "step": 17280
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.629056253001238,
+      "learning_rate": 4.46137563196801e-06,
+      "loss": 0.9718,
+      "step": 17285
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.206480636864152,
+      "learning_rate": 4.458459794824088e-06,
+      "loss": 0.9993,
+      "step": 17290
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 4.803311960029917,
+      "learning_rate": 4.455544144023635e-06,
+      "loss": 0.9975,
+      "step": 17295
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.348121825749278,
+      "learning_rate": 4.452628680569919e-06,
+      "loss": 0.9796,
+      "step": 17300
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.288188813858446,
+      "learning_rate": 4.449713405466154e-06,
+      "loss": 0.9764,
+      "step": 17305
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.461168584339323,
+      "learning_rate": 4.446798319715478e-06,
+      "loss": 0.9525,
+      "step": 17310
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.714426835122893,
+      "learning_rate": 4.443883424320974e-06,
+      "loss": 1.0277,
+      "step": 17315
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.288258486608056,
+      "learning_rate": 4.440968720285651e-06,
+      "loss": 0.9845,
+      "step": 17320
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.975300294674621,
+      "learning_rate": 4.438054208612459e-06,
+      "loss": 0.9864,
+      "step": 17325
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.98775495317326,
+      "learning_rate": 4.435139890304274e-06,
+      "loss": 1.0285,
+      "step": 17330
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.307857755049275,
+      "learning_rate": 4.4322257663639186e-06,
+      "loss": 0.9848,
+      "step": 17335
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.699145118913506,
+      "learning_rate": 4.4293118377941295e-06,
+      "loss": 1.0009,
+      "step": 17340
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.902087498401224,
+      "learning_rate": 4.426398105597595e-06,
+      "loss": 1.0017,
+      "step": 17345
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.285934871439602,
+      "learning_rate": 4.423484570776923e-06,
+      "loss": 0.9993,
+      "step": 17350
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.455064209350546,
+      "learning_rate": 4.420571234334661e-06,
+      "loss": 1.0108,
+      "step": 17355
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.4557331429864115,
+      "learning_rate": 4.417658097273282e-06,
+      "loss": 0.9708,
+      "step": 17360
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.605121792110902,
+      "learning_rate": 4.414745160595198e-06,
+      "loss": 0.9765,
+      "step": 17365
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.619675388451704,
+      "learning_rate": 4.411832425302742e-06,
+      "loss": 1.0003,
+      "step": 17370
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.416456959181561,
+      "learning_rate": 4.408919892398189e-06,
+      "loss": 0.9591,
+      "step": 17375
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.601776990215138,
+      "learning_rate": 4.406007562883736e-06,
+      "loss": 1.0323,
+      "step": 17380
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.827378110856624,
+      "learning_rate": 4.403095437761514e-06,
+      "loss": 0.9753,
+      "step": 17385
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.286835453274794,
+      "learning_rate": 4.400183518033579e-06,
+      "loss": 0.9774,
+      "step": 17390
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.106658867927537,
+      "learning_rate": 4.397271804701926e-06,
+      "loss": 0.9667,
+      "step": 17395
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.385230411470802,
+      "learning_rate": 4.394360298768469e-06,
+      "loss": 0.9837,
+      "step": 17400
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.907928573203082,
+      "learning_rate": 4.391449001235052e-06,
+      "loss": 0.9549,
+      "step": 17405
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.133982511906204,
+      "learning_rate": 4.388537913103454e-06,
+      "loss": 1.0296,
+      "step": 17410
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.0215145597485265,
+      "learning_rate": 4.385627035375377e-06,
+      "loss": 1.0016,
+      "step": 17415
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.266710216270274,
+      "learning_rate": 4.38271636905245e-06,
+      "loss": 1.0154,
+      "step": 17420
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.678286453039773,
+      "learning_rate": 4.3798059151362295e-06,
+      "loss": 0.9848,
+      "step": 17425
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.549703960442,
+      "learning_rate": 4.3768956746282026e-06,
+      "loss": 1.0005,
+      "step": 17430
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 29.5138649807499,
+      "learning_rate": 4.373985648529777e-06,
+      "loss": 0.9626,
+      "step": 17435
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.480192541997506,
+      "learning_rate": 4.371075837842294e-06,
+      "loss": 0.9756,
+      "step": 17440
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.568325674625413,
+      "learning_rate": 4.368166243567013e-06,
+      "loss": 1.006,
+      "step": 17445
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.115886297919321,
+      "learning_rate": 4.365256866705126e-06,
+      "loss": 0.9651,
+      "step": 17450
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.275895266537964,
+      "learning_rate": 4.362347708257743e-06,
+      "loss": 0.962,
+      "step": 17455
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.259529619695632,
+      "learning_rate": 4.359438769225906e-06,
+      "loss": 1.0017,
+      "step": 17460
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.857933774426776,
+      "learning_rate": 4.3565300506105745e-06,
+      "loss": 0.9852,
+      "step": 17465
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.629255297123736,
+      "learning_rate": 4.3536215534126404e-06,
+      "loss": 0.9695,
+      "step": 17470
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.3825534811967355,
+      "learning_rate": 4.350713278632911e-06,
+      "loss": 0.9809,
+      "step": 17475
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.394150217809622,
+      "learning_rate": 4.3478052272721234e-06,
+      "loss": 0.9942,
+      "step": 17480
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.22363972614146,
+      "learning_rate": 4.344897400330933e-06,
+      "loss": 0.9782,
+      "step": 17485
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.202572061556888,
+      "learning_rate": 4.341989798809923e-06,
+      "loss": 0.9811,
+      "step": 17490
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.541440915741455,
+      "learning_rate": 4.339082423709592e-06,
+      "loss": 0.9589,
+      "step": 17495
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.146462234492245,
+      "learning_rate": 4.33617527603037e-06,
+      "loss": 0.9939,
+      "step": 17500
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.2775085198018195,
+      "learning_rate": 4.3332683567726e-06,
+      "loss": 0.9663,
+      "step": 17505
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.61356411742695,
+      "learning_rate": 4.330361666936555e-06,
+      "loss": 0.9924,
+      "step": 17510
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.77351936985382,
+      "learning_rate": 4.327455207522417e-06,
+      "loss": 1.0308,
+      "step": 17515
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.68995331942081,
+      "learning_rate": 4.3245489795303025e-06,
+      "loss": 1.0281,
+      "step": 17520
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.917899439391954,
+      "learning_rate": 4.321642983960238e-06,
+      "loss": 0.9828,
+      "step": 17525
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.713513046333969,
+      "learning_rate": 4.318737221812173e-06,
+      "loss": 1.0316,
+      "step": 17530
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.220594770385739,
+      "learning_rate": 4.315831694085982e-06,
+      "loss": 0.9288,
+      "step": 17535
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.360087324298462,
+      "learning_rate": 4.31292640178145e-06,
+      "loss": 0.9691,
+      "step": 17540
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.104011531131553,
+      "learning_rate": 4.310021345898288e-06,
+      "loss": 0.9856,
+      "step": 17545
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.394733795721605,
+      "learning_rate": 4.307116527436121e-06,
+      "loss": 0.9483,
+      "step": 17550
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.6137553038406685,
+      "learning_rate": 4.304211947394496e-06,
+      "loss": 0.9874,
+      "step": 17555
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 17.249399564508376,
+      "learning_rate": 4.301307606772873e-06,
+      "loss": 0.9847,
+      "step": 17560
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.970658585095424,
+      "learning_rate": 4.298403506570638e-06,
+      "loss": 0.992,
+      "step": 17565
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 20.018871333149153,
+      "learning_rate": 4.295499647787085e-06,
+      "loss": 0.9623,
+      "step": 17570
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.33446100071459,
+      "learning_rate": 4.292596031421431e-06,
+      "loss": 0.9659,
+      "step": 17575
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.655812375433922,
+      "learning_rate": 4.289692658472806e-06,
+      "loss": 0.9549,
+      "step": 17580
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.006061782628333,
+      "learning_rate": 4.28678952994026e-06,
+      "loss": 0.978,
+      "step": 17585
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.31919048127266,
+      "learning_rate": 4.283886646822753e-06,
+      "loss": 0.9939,
+      "step": 17590
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.898513394142511,
+      "learning_rate": 4.2809840101191695e-06,
+      "loss": 0.9559,
+      "step": 17595
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 22.10211628788823,
+      "learning_rate": 4.2780816208282995e-06,
+      "loss": 0.9865,
+      "step": 17600
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.142924272084487,
+      "learning_rate": 4.275179479948856e-06,
+      "loss": 0.9863,
+      "step": 17605
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 43.70523431723558,
+      "learning_rate": 4.272277588479461e-06,
+      "loss": 0.9565,
+      "step": 17610
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 22.507175546613677,
+      "learning_rate": 4.269375947418652e-06,
+      "loss": 0.9726,
+      "step": 17615
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 17.30146559379677,
+      "learning_rate": 4.2664745577648806e-06,
+      "loss": 1.0624,
+      "step": 17620
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 25.06644250902353,
+      "learning_rate": 4.2635734205165155e-06,
+      "loss": 1.0045,
+      "step": 17625
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.910629056259536,
+      "learning_rate": 4.260672536671832e-06,
+      "loss": 0.9652,
+      "step": 17630
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.806887353642972,
+      "learning_rate": 4.257771907229023e-06,
+      "loss": 0.969,
+      "step": 17635
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.1094952043062385,
+      "learning_rate": 4.254871533186191e-06,
+      "loss": 1.003,
+      "step": 17640
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 20.289905829013996,
+      "learning_rate": 4.2519714155413534e-06,
+      "loss": 1.0062,
+      "step": 17645
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.23840709833831,
+      "learning_rate": 4.249071555292438e-06,
+      "loss": 0.964,
+      "step": 17650
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 37.41253753110922,
+      "learning_rate": 4.246171953437281e-06,
+      "loss": 1.0014,
+      "step": 17655
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.888293950500067,
+      "learning_rate": 4.2432726109736365e-06,
+      "loss": 0.9665,
+      "step": 17660
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.644402360932366,
+      "learning_rate": 4.240373528899163e-06,
+      "loss": 0.9564,
+      "step": 17665
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.026582771228052,
+      "learning_rate": 4.237474708211434e-06,
+      "loss": 0.9931,
+      "step": 17670
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.193252882210064,
+      "learning_rate": 4.234576149907928e-06,
+      "loss": 1.0125,
+      "step": 17675
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.234158378729202,
+      "learning_rate": 4.2316778549860395e-06,
+      "loss": 0.9782,
+      "step": 17680
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.290440088877322,
+      "learning_rate": 4.228779824443065e-06,
+      "loss": 1.0064,
+      "step": 17685
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.854939976517208,
+      "learning_rate": 4.225882059276217e-06,
+      "loss": 1.0144,
+      "step": 17690
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.235898705036122,
+      "learning_rate": 4.222984560482614e-06,
+      "loss": 0.9722,
+      "step": 17695
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.231748698058492,
+      "learning_rate": 4.220087329059281e-06,
+      "loss": 0.9929,
+      "step": 17700
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.612949224990907,
+      "learning_rate": 4.2171903660031535e-06,
+      "loss": 0.9802,
+      "step": 17705
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.64871835788384,
+      "learning_rate": 4.214293672311073e-06,
+      "loss": 1.0134,
+      "step": 17710
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 27.804267649031544,
+      "learning_rate": 4.211397248979789e-06,
+      "loss": 0.9994,
+      "step": 17715
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.232331289327623,
+      "learning_rate": 4.208501097005959e-06,
+      "loss": 0.9865,
+      "step": 17720
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.36852730763666,
+      "learning_rate": 4.205605217386145e-06,
+      "loss": 0.9561,
+      "step": 17725
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.900038904042403,
+      "learning_rate": 4.202709611116818e-06,
+      "loss": 0.996,
+      "step": 17730
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 29.413539099777385,
+      "learning_rate": 4.1998142791943515e-06,
+      "loss": 1.0042,
+      "step": 17735
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.892005516743971,
+      "learning_rate": 4.196919222615029e-06,
+      "loss": 1.0257,
+      "step": 17740
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.966387701303805,
+      "learning_rate": 4.194024442375032e-06,
+      "loss": 0.9833,
+      "step": 17745
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.231042995618557,
+      "learning_rate": 4.191129939470457e-06,
+      "loss": 0.9778,
+      "step": 17750
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.639381946636583,
+      "learning_rate": 4.188235714897296e-06,
+      "loss": 0.9758,
+      "step": 17755
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.232707031871469,
+      "learning_rate": 4.185341769651451e-06,
+      "loss": 0.9965,
+      "step": 17760
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.799057501236086,
+      "learning_rate": 4.1824481047287244e-06,
+      "loss": 0.9631,
+      "step": 17765
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.93211267186369,
+      "learning_rate": 4.179554721124825e-06,
+      "loss": 1.0005,
+      "step": 17770
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.342829499951731,
+      "learning_rate": 4.17666161983536e-06,
+      "loss": 0.9595,
+      "step": 17775
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.298086003533476,
+      "learning_rate": 4.1737688018558476e-06,
+      "loss": 1.0069,
+      "step": 17780
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.309454412926847,
+      "learning_rate": 4.170876268181703e-06,
+      "loss": 0.9926,
+      "step": 17785
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.48349237724496,
+      "learning_rate": 4.167984019808238e-06,
+      "loss": 0.9817,
+      "step": 17790
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.696426460645399,
+      "learning_rate": 4.165092057730681e-06,
+      "loss": 1.0428,
+      "step": 17795
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.215360712972935,
+      "learning_rate": 4.1622003829441474e-06,
+      "loss": 0.987,
+      "step": 17800
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.644196565159671,
+      "learning_rate": 4.1593089964436625e-06,
+      "loss": 0.977,
+      "step": 17805
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.107974839971601,
+      "learning_rate": 4.156417899224147e-06,
+      "loss": 0.9941,
+      "step": 17810
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.301233945541358,
+      "learning_rate": 4.153527092280431e-06,
+      "loss": 1.0134,
+      "step": 17815
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.220120395735668,
+      "learning_rate": 4.15063657660723e-06,
+      "loss": 1.0002,
+      "step": 17820
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 15.811694128818516,
+      "learning_rate": 4.147746353199174e-06,
+      "loss": 0.9888,
+      "step": 17825
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.713258453209393,
+      "learning_rate": 4.144856423050783e-06,
+      "loss": 0.9514,
+      "step": 17830
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.175345447281716,
+      "learning_rate": 4.14196678715648e-06,
+      "loss": 0.9518,
+      "step": 17835
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.196682747430522,
+      "learning_rate": 4.139077446510585e-06,
+      "loss": 1.0259,
+      "step": 17840
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.721920938939746,
+      "learning_rate": 4.136188402107322e-06,
+      "loss": 0.9831,
+      "step": 17845
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 16.312284928694034,
+      "learning_rate": 4.133299654940801e-06,
+      "loss": 0.9578,
+      "step": 17850
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.727399474249944,
+      "learning_rate": 4.130411206005043e-06,
+      "loss": 1.0198,
+      "step": 17855
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.203930546045322,
+      "learning_rate": 4.127523056293958e-06,
+      "loss": 1.0151,
+      "step": 17860
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.905780390122556,
+      "learning_rate": 4.124635206801356e-06,
+      "loss": 0.9554,
+      "step": 17865
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.270094254344347,
+      "learning_rate": 4.121747658520942e-06,
+      "loss": 0.9901,
+      "step": 17870
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 24.142268165349467,
+      "learning_rate": 4.1188604124463224e-06,
+      "loss": 1.0144,
+      "step": 17875
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.472023048193844,
+      "learning_rate": 4.115973469570991e-06,
+      "loss": 0.9952,
+      "step": 17880
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.9995434492814255,
+      "learning_rate": 4.113086830888346e-06,
+      "loss": 1.0237,
+      "step": 17885
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.160793829862443,
+      "learning_rate": 4.1102004973916744e-06,
+      "loss": 0.9899,
+      "step": 17890
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.314565949813037,
+      "learning_rate": 4.107314470074163e-06,
+      "loss": 0.9722,
+      "step": 17895
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.753414499732502,
+      "learning_rate": 4.104428749928887e-06,
+      "loss": 0.9672,
+      "step": 17900
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.277581642321612,
+      "learning_rate": 4.101543337948826e-06,
+      "loss": 0.9993,
+      "step": 17905
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.02303347969287,
+      "learning_rate": 4.0986582351268445e-06,
+      "loss": 0.9497,
+      "step": 17910
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 22.410472744076394,
+      "learning_rate": 4.0957734424557e-06,
+      "loss": 1.0122,
+      "step": 17915
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 25.44505269635536,
+      "learning_rate": 4.092888960928053e-06,
+      "loss": 0.9531,
+      "step": 17920
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.758375815804422,
+      "learning_rate": 4.090004791536447e-06,
+      "loss": 1.0082,
+      "step": 17925
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.558547984958102,
+      "learning_rate": 4.0871209352733244e-06,
+      "loss": 1.0401,
+      "step": 17930
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.957332675445743,
+      "learning_rate": 4.084237393131015e-06,
+      "loss": 1.0092,
+      "step": 17935
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.890734746420278,
+      "learning_rate": 4.081354166101744e-06,
+      "loss": 0.9885,
+      "step": 17940
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.54145049689193,
+      "learning_rate": 4.078471255177626e-06,
+      "loss": 0.981,
+      "step": 17945
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 23.910132442804926,
+      "learning_rate": 4.075588661350671e-06,
+      "loss": 1.0282,
+      "step": 17950
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.24292684826029,
+      "learning_rate": 4.072706385612773e-06,
+      "loss": 0.9975,
+      "step": 17955
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 19.115389829935094,
+      "learning_rate": 4.069824428955724e-06,
+      "loss": 0.9957,
+      "step": 17960
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.945914484492379,
+      "learning_rate": 4.066942792371198e-06,
+      "loss": 0.9641,
+      "step": 17965
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.521339685687494,
+      "learning_rate": 4.064061476850769e-06,
+      "loss": 0.9963,
+      "step": 17970
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.894472343268108,
+      "learning_rate": 4.0611804833858885e-06,
+      "loss": 1.0377,
+      "step": 17975
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.576633426606861,
+      "learning_rate": 4.05829981296791e-06,
+      "loss": 0.9905,
+      "step": 17980
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 24.05532145531544,
+      "learning_rate": 4.055419466588064e-06,
+      "loss": 1.0365,
+      "step": 17985
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.015289000674614,
+      "learning_rate": 4.05253944523748e-06,
+      "loss": 0.9566,
+      "step": 17990
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.899950989787076,
+      "learning_rate": 4.049659749907165e-06,
+      "loss": 1.0058,
+      "step": 17995
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.243197684418895,
+      "learning_rate": 4.046780381588025e-06,
+      "loss": 1.0171,
+      "step": 18000
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.474829652542478,
+      "learning_rate": 4.043901341270843e-06,
+      "loss": 0.9419,
+      "step": 18005
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.102354302507712,
+      "learning_rate": 4.041022629946298e-06,
+      "loss": 0.9974,
+      "step": 18010
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.581577273091294,
+      "learning_rate": 4.038144248604949e-06,
+      "loss": 0.9694,
+      "step": 18015
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.7366265831491345,
+      "learning_rate": 4.035266198237247e-06,
+      "loss": 1.0029,
+      "step": 18020
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.450022518303586,
+      "learning_rate": 4.032388479833522e-06,
+      "loss": 0.9699,
+      "step": 18025
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 21.668572220867887,
+      "learning_rate": 4.029511094384001e-06,
+      "loss": 0.9723,
+      "step": 18030
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.873050740877582,
+      "learning_rate": 4.026634042878782e-06,
+      "loss": 0.9608,
+      "step": 18035
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.936992079591061,
+      "learning_rate": 4.02375732630786e-06,
+      "loss": 0.9841,
+      "step": 18040
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.356020950439156,
+      "learning_rate": 4.020880945661111e-06,
+      "loss": 0.9837,
+      "step": 18045
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.460548595726973,
+      "learning_rate": 4.0180049019282916e-06,
+      "loss": 0.9915,
+      "step": 18050
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.417525899749307,
+      "learning_rate": 4.015129196099049e-06,
+      "loss": 1.0166,
+      "step": 18055
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.20607296161159,
+      "learning_rate": 4.012253829162908e-06,
+      "loss": 0.9829,
+      "step": 18060
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.930258224866766,
+      "learning_rate": 4.0093788021092826e-06,
+      "loss": 0.991,
+      "step": 18065
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.401745585192158,
+      "learning_rate": 4.006504115927463e-06,
+      "loss": 1.0106,
+      "step": 18070
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.793097611353607,
+      "learning_rate": 4.0036297716066295e-06,
+      "loss": 1.0181,
+      "step": 18075
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.639441474692842,
+      "learning_rate": 4.0007557701358405e-06,
+      "loss": 0.9854,
+      "step": 18080
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.979689799633668,
+      "learning_rate": 3.997882112504036e-06,
+      "loss": 0.9966,
+      "step": 18085
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.30019551431489,
+      "learning_rate": 3.995008799700039e-06,
+      "loss": 0.9399,
+      "step": 18090
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.010236910045,
+      "learning_rate": 3.992135832712555e-06,
+      "loss": 0.973,
+      "step": 18095
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.357174340154753,
+      "learning_rate": 3.9892632125301664e-06,
+      "loss": 0.9703,
+      "step": 18100
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.800647275884193,
+      "learning_rate": 3.986390940141344e-06,
+      "loss": 1.0361,
+      "step": 18105
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.66733919132119,
+      "learning_rate": 3.983519016534429e-06,
+      "loss": 0.9551,
+      "step": 18110
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.08796648741792,
+      "learning_rate": 3.980647442697651e-06,
+      "loss": 0.9354,
+      "step": 18115
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.8986084699880825,
+      "learning_rate": 3.977776219619113e-06,
+      "loss": 0.9813,
+      "step": 18120
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.779198222654169,
+      "learning_rate": 3.974905348286803e-06,
+      "loss": 1.004,
+      "step": 18125
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.207525881092016,
+      "learning_rate": 3.9720348296885805e-06,
+      "loss": 0.9679,
+      "step": 18130
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.756377143981297,
+      "learning_rate": 3.969164664812194e-06,
+      "loss": 0.9359,
+      "step": 18135
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.282128100521323,
+      "learning_rate": 3.966294854645261e-06,
+      "loss": 0.943,
+      "step": 18140
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.343543523016564,
+      "learning_rate": 3.963425400175282e-06,
+      "loss": 0.9858,
+      "step": 18145
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 14.959195280855614,
+      "learning_rate": 3.960556302389632e-06,
+      "loss": 0.9732,
+      "step": 18150
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.4525478859237255,
+      "learning_rate": 3.957687562275566e-06,
+      "loss": 0.9777,
+      "step": 18155
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.0295155709128565,
+      "learning_rate": 3.954819180820213e-06,
+      "loss": 1.0042,
+      "step": 18160
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.130634304881189,
+      "learning_rate": 3.951951159010583e-06,
+      "loss": 0.9856,
+      "step": 18165
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.513383734699328,
+      "learning_rate": 3.949083497833558e-06,
+      "loss": 0.9726,
+      "step": 18170
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.615005606728285,
+      "learning_rate": 3.946216198275897e-06,
+      "loss": 0.9881,
+      "step": 18175
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.967220890591053,
+      "learning_rate": 3.943349261324237e-06,
+      "loss": 0.9645,
+      "step": 18180
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.03922077754858,
+      "learning_rate": 3.940482687965085e-06,
+      "loss": 1.0003,
+      "step": 18185
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.307895107855172,
+      "learning_rate": 3.937616479184831e-06,
+      "loss": 0.9933,
+      "step": 18190
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 18.744700072412474,
+      "learning_rate": 3.934750635969728e-06,
+      "loss": 0.9654,
+      "step": 18195
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.952335132138158,
+      "learning_rate": 3.931885159305916e-06,
+      "loss": 1.0467,
+      "step": 18200
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.245737402574543,
+      "learning_rate": 3.9290200501794e-06,
+      "loss": 0.9912,
+      "step": 18205
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.934123527505806,
+      "learning_rate": 3.926155309576063e-06,
+      "loss": 0.9621,
+      "step": 18210
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.822756484981494,
+      "learning_rate": 3.923290938481657e-06,
+      "loss": 0.9913,
+      "step": 18215
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.3094605135069015,
+      "learning_rate": 3.920426937881812e-06,
+      "loss": 1.0135,
+      "step": 18220
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.234278005327626,
+      "learning_rate": 3.917563308762024e-06,
+      "loss": 1.024,
+      "step": 18225
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 14.143253937630613,
+      "learning_rate": 3.9147000521076695e-06,
+      "loss": 1.0049,
+      "step": 18230
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.190834230959963,
+      "learning_rate": 3.9118371689039905e-06,
+      "loss": 1.0115,
+      "step": 18235
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.641550092853329,
+      "learning_rate": 3.908974660136102e-06,
+      "loss": 0.9952,
+      "step": 18240
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.542498068012963,
+      "learning_rate": 3.906112526788991e-06,
+      "loss": 0.9804,
+      "step": 18245
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.868475394958528,
+      "learning_rate": 3.903250769847515e-06,
+      "loss": 1.0248,
+      "step": 18250
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.90066644420374,
+      "learning_rate": 3.9003893902964e-06,
+      "loss": 0.9451,
+      "step": 18255
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 27.968509274783298,
+      "learning_rate": 3.897528389120247e-06,
+      "loss": 0.9905,
+      "step": 18260
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.71773024020709,
+      "learning_rate": 3.894667767303523e-06,
+      "loss": 0.9664,
+      "step": 18265
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.158395474912115,
+      "learning_rate": 3.891807525830564e-06,
+      "loss": 1.0286,
+      "step": 18270
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.322537717175965,
+      "learning_rate": 3.888947665685576e-06,
+      "loss": 1.0008,
+      "step": 18275
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.961458444995351,
+      "learning_rate": 3.8860881878526365e-06,
+      "loss": 0.9912,
+      "step": 18280
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.145722663367105,
+      "learning_rate": 3.883229093315686e-06,
+      "loss": 1.0161,
+      "step": 18285
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.074742049472983,
+      "learning_rate": 3.880370383058539e-06,
+      "loss": 0.9978,
+      "step": 18290
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.970232085990174,
+      "learning_rate": 3.877512058064876e-06,
+      "loss": 0.9639,
+      "step": 18295
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.412604579770981,
+      "learning_rate": 3.8746541193182375e-06,
+      "loss": 1.0341,
+      "step": 18300
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.680626298857828,
+      "learning_rate": 3.871796567802044e-06,
+      "loss": 0.9469,
+      "step": 18305
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.796765209774283,
+      "learning_rate": 3.868939404499573e-06,
+      "loss": 0.942,
+      "step": 18310
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.149810215370284,
+      "learning_rate": 3.8660826303939745e-06,
+      "loss": 0.9614,
+      "step": 18315
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 21.602876098567382,
+      "learning_rate": 3.863226246468256e-06,
+      "loss": 1.0041,
+      "step": 18320
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.5164022506358705,
+      "learning_rate": 3.860370253705304e-06,
+      "loss": 0.9953,
+      "step": 18325
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.346002360848893,
+      "learning_rate": 3.857514653087857e-06,
+      "loss": 0.9913,
+      "step": 18330
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.696269583932353,
+      "learning_rate": 3.854659445598529e-06,
+      "loss": 0.9501,
+      "step": 18335
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.613799479606872,
+      "learning_rate": 3.85180463221979e-06,
+      "loss": 0.9782,
+      "step": 18340
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.906200366057462,
+      "learning_rate": 3.848950213933982e-06,
+      "loss": 0.9995,
+      "step": 18345
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.985103632638992,
+      "learning_rate": 3.846096191723303e-06,
+      "loss": 1.0518,
+      "step": 18350
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.450855140786176,
+      "learning_rate": 3.843242566569826e-06,
+      "loss": 1.0015,
+      "step": 18355
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.17042311212853,
+      "learning_rate": 3.840389339455474e-06,
+      "loss": 0.9941,
+      "step": 18360
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.19466857070708,
+      "learning_rate": 3.837536511362045e-06,
+      "loss": 0.9699,
+      "step": 18365
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.901155997440986,
+      "learning_rate": 3.834684083271191e-06,
+      "loss": 0.9548,
+      "step": 18370
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 19.352982354625738,
+      "learning_rate": 3.831832056164431e-06,
+      "loss": 0.9756,
+      "step": 18375
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.70998177042718,
+      "learning_rate": 3.8289804310231434e-06,
+      "loss": 1.0063,
+      "step": 18380
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 25.9351320245653,
+      "learning_rate": 3.826129208828573e-06,
+      "loss": 0.9915,
+      "step": 18385
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.987625138296798,
+      "learning_rate": 3.82327839056182e-06,
+      "loss": 0.9812,
+      "step": 18390
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.41267269943096,
+      "learning_rate": 3.8204279772038495e-06,
+      "loss": 0.9616,
+      "step": 18395
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.9943162502046015,
+      "learning_rate": 3.817577969735485e-06,
+      "loss": 0.9764,
+      "step": 18400
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.261829409492166,
+      "learning_rate": 3.8147283691374124e-06,
+      "loss": 0.9682,
+      "step": 18405
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.718804953227591,
+      "learning_rate": 3.811879176390174e-06,
+      "loss": 0.9882,
+      "step": 18410
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.7449916557750855,
+      "learning_rate": 3.8090303924741784e-06,
+      "loss": 1.0081,
+      "step": 18415
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 19.46955376180154,
+      "learning_rate": 3.8061820183696857e-06,
+      "loss": 0.9514,
+      "step": 18420
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.0383605638679505,
+      "learning_rate": 3.803334055056822e-06,
+      "loss": 0.9281,
+      "step": 18425
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.50864837032302,
+      "learning_rate": 3.8004865035155662e-06,
+      "loss": 0.9906,
+      "step": 18430
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.858201015924298,
+      "learning_rate": 3.797639364725758e-06,
+      "loss": 0.928,
+      "step": 18435
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.82481969041467,
+      "learning_rate": 3.7947926396670975e-06,
+      "loss": 0.9848,
+      "step": 18440
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.861048594296718,
+      "learning_rate": 3.791946329319136e-06,
+      "loss": 0.9813,
+      "step": 18445
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.975396611645755,
+      "learning_rate": 3.7891004346612925e-06,
+      "loss": 0.9723,
+      "step": 18450
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 20.182853926554543,
+      "learning_rate": 3.786254956672829e-06,
+      "loss": 0.9931,
+      "step": 18455
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.837429575125533,
+      "learning_rate": 3.7834098963328784e-06,
+      "loss": 0.9738,
+      "step": 18460
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.216565893064262,
+      "learning_rate": 3.7805652546204193e-06,
+      "loss": 0.9949,
+      "step": 18465
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.355112575571148,
+      "learning_rate": 3.777721032514292e-06,
+      "loss": 0.9918,
+      "step": 18470
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.094017204819921,
+      "learning_rate": 3.7748772309931887e-06,
+      "loss": 0.9381,
+      "step": 18475
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.374752647571407,
+      "learning_rate": 3.7720338510356623e-06,
+      "loss": 0.9747,
+      "step": 18480
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.5966625675434125,
+      "learning_rate": 3.7691908936201126e-06,
+      "loss": 0.9333,
+      "step": 18485
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.305843062394117,
+      "learning_rate": 3.7663483597248023e-06,
+      "loss": 0.9548,
+      "step": 18490
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.114471686460963,
+      "learning_rate": 3.7635062503278425e-06,
+      "loss": 1.0119,
+      "step": 18495
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.816051820282711,
+      "learning_rate": 3.7606645664072016e-06,
+      "loss": 0.9708,
+      "step": 18500
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.49667170674841,
+      "learning_rate": 3.7578233089406984e-06,
+      "loss": 0.9589,
+      "step": 18505
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.761558730111664,
+      "learning_rate": 3.754982478906011e-06,
+      "loss": 0.9756,
+      "step": 18510
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.967302332582332,
+      "learning_rate": 3.75214207728066e-06,
+      "loss": 0.9411,
+      "step": 18515
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.116253224622286,
+      "learning_rate": 3.7493021050420308e-06,
+      "loss": 0.9907,
+      "step": 18520
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 24.120440143339238,
+      "learning_rate": 3.746462563167351e-06,
+      "loss": 0.9318,
+      "step": 18525
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.41418126468614,
+      "learning_rate": 3.7436234526337077e-06,
+      "loss": 0.9539,
+      "step": 18530
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.873694965923555,
+      "learning_rate": 3.740784774418032e-06,
+      "loss": 1.0013,
+      "step": 18535
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 21.799567648823835,
+      "learning_rate": 3.7379465294971164e-06,
+      "loss": 1.0051,
+      "step": 18540
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.05982153807311,
+      "learning_rate": 3.7351087188475904e-06,
+      "loss": 0.9677,
+      "step": 18545
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 32.014472362161165,
+      "learning_rate": 3.732271343445949e-06,
+      "loss": 0.9886,
+      "step": 18550
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.146122922065649,
+      "learning_rate": 3.729434404268527e-06,
+      "loss": 0.9481,
+      "step": 18555
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 17.909251593077418,
+      "learning_rate": 3.726597902291512e-06,
+      "loss": 1.0004,
+      "step": 18560
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.9456126927850335,
+      "learning_rate": 3.7237618384909446e-06,
+      "loss": 0.9513,
+      "step": 18565
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.361620530332281,
+      "learning_rate": 3.720926213842707e-06,
+      "loss": 0.9708,
+      "step": 18570
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.811469339075542,
+      "learning_rate": 3.7180910293225404e-06,
+      "loss": 1.0052,
+      "step": 18575
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.513163674141524,
+      "learning_rate": 3.715256285906023e-06,
+      "loss": 1.0281,
+      "step": 18580
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.774195223815436,
+      "learning_rate": 3.712421984568593e-06,
+      "loss": 1.0158,
+      "step": 18585
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.960282979334194,
+      "learning_rate": 3.7095881262855267e-06,
+      "loss": 0.9662,
+      "step": 18590
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.978757212662668,
+      "learning_rate": 3.7067547120319554e-06,
+      "loss": 0.9774,
+      "step": 18595
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.034834780122345,
+      "learning_rate": 3.7039217427828513e-06,
+      "loss": 0.9752,
+      "step": 18600
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.004516099561922,
+      "learning_rate": 3.701089219513038e-06,
+      "loss": 0.9734,
+      "step": 18605
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.402085580760845,
+      "learning_rate": 3.6982571431971824e-06,
+      "loss": 1.0251,
+      "step": 18610
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 23.56500346844539,
+      "learning_rate": 3.695425514809802e-06,
+      "loss": 0.9798,
+      "step": 18615
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.346783743132763,
+      "learning_rate": 3.692594335325255e-06,
+      "loss": 1.0118,
+      "step": 18620
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.117378850707686,
+      "learning_rate": 3.68976360571775e-06,
+      "loss": 0.9884,
+      "step": 18625
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.6563959867572,
+      "learning_rate": 3.686933326961335e-06,
+      "loss": 0.9778,
+      "step": 18630
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.718362628760115,
+      "learning_rate": 3.6841035000299104e-06,
+      "loss": 0.9712,
+      "step": 18635
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.901977228691198,
+      "learning_rate": 3.681274125897212e-06,
+      "loss": 1.0074,
+      "step": 18640
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.376219367954663,
+      "learning_rate": 3.6784452055368293e-06,
+      "loss": 0.9927,
+      "step": 18645
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.261738860162723,
+      "learning_rate": 3.6756167399221875e-06,
+      "loss": 0.9726,
+      "step": 18650
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.322575516017178,
+      "learning_rate": 3.6727887300265623e-06,
+      "loss": 0.9967,
+      "step": 18655
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.396128479592039,
+      "learning_rate": 3.669961176823065e-06,
+      "loss": 1.0004,
+      "step": 18660
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.572859527055436,
+      "learning_rate": 3.66713408128466e-06,
+      "loss": 0.9574,
+      "step": 18665
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.566152355775521,
+      "learning_rate": 3.664307444384141e-06,
+      "loss": 0.944,
+      "step": 18670
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.814616898071068,
+      "learning_rate": 3.661481267094157e-06,
+      "loss": 0.9645,
+      "step": 18675
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.4491753760139545,
+      "learning_rate": 3.658655550387191e-06,
+      "loss": 0.971,
+      "step": 18680
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.650043757641736,
+      "learning_rate": 3.655830295235567e-06,
+      "loss": 0.994,
+      "step": 18685
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.950233639909765,
+      "learning_rate": 3.6530055026114564e-06,
+      "loss": 0.9695,
+      "step": 18690
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.336979569142324,
+      "learning_rate": 3.6501811734868644e-06,
+      "loss": 0.9697,
+      "step": 18695
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.485667474533593,
+      "learning_rate": 3.6473573088336424e-06,
+      "loss": 0.9788,
+      "step": 18700
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.737953968885742,
+      "learning_rate": 3.6445339096234765e-06,
+      "loss": 0.9528,
+      "step": 18705
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.420294666661196,
+      "learning_rate": 3.6417109768279004e-06,
+      "loss": 0.9939,
+      "step": 18710
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.063368906404644,
+      "learning_rate": 3.638888511418278e-06,
+      "loss": 0.9813,
+      "step": 18715
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.256498293961954,
+      "learning_rate": 3.63606651436582e-06,
+      "loss": 0.946,
+      "step": 18720
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.775343917226046,
+      "learning_rate": 3.633244986641571e-06,
+      "loss": 1.0092,
+      "step": 18725
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.327990203243036,
+      "learning_rate": 3.630423929216417e-06,
+      "loss": 0.9956,
+      "step": 18730
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.684176316275288,
+      "learning_rate": 3.6276033430610793e-06,
+      "loss": 0.9586,
+      "step": 18735
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.370734052190713,
+      "learning_rate": 3.6247832291461226e-06,
+      "loss": 0.9927,
+      "step": 18740
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.603847329973096,
+      "learning_rate": 3.621963588441942e-06,
+      "loss": 0.9735,
+      "step": 18745
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 14.70117140483938,
+      "learning_rate": 3.6191444219187753e-06,
+      "loss": 0.913,
+      "step": 18750
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.259241140398004,
+      "learning_rate": 3.6163257305466937e-06,
+      "loss": 1.0054,
+      "step": 18755
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.831064986749914,
+      "learning_rate": 3.6135075152956074e-06,
+      "loss": 0.9717,
+      "step": 18760
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.268942377717585,
+      "learning_rate": 3.610689777135259e-06,
+      "loss": 0.9348,
+      "step": 18765
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.077651255287886,
+      "learning_rate": 3.6078725170352323e-06,
+      "loss": 0.9825,
+      "step": 18770
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.9833842942122235,
+      "learning_rate": 3.605055735964943e-06,
+      "loss": 0.9822,
+      "step": 18775
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.798592166369944,
+      "learning_rate": 3.6022394348936436e-06,
+      "loss": 0.9778,
+      "step": 18780
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.34851456599785,
+      "learning_rate": 3.5994236147904182e-06,
+      "loss": 0.9623,
+      "step": 18785
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.684841424069073,
+      "learning_rate": 3.5966082766241904e-06,
+      "loss": 0.9787,
+      "step": 18790
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.479047892827804,
+      "learning_rate": 3.593793421363712e-06,
+      "loss": 0.9816,
+      "step": 18795
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.040538898804096,
+      "learning_rate": 3.5909790499775765e-06,
+      "loss": 0.9905,
+      "step": 18800
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.958641544227882,
+      "learning_rate": 3.588165163434203e-06,
+      "loss": 0.9728,
+      "step": 18805
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.054065495967254,
+      "learning_rate": 3.5853517627018495e-06,
+      "loss": 0.9511,
+      "step": 18810
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.890771687952744,
+      "learning_rate": 3.5825388487486034e-06,
+      "loss": 0.995,
+      "step": 18815
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.0730729877051965,
+      "learning_rate": 3.5797264225423846e-06,
+      "loss": 0.9923,
+      "step": 18820
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.636257745017197,
+      "learning_rate": 3.576914485050948e-06,
+      "loss": 0.9761,
+      "step": 18825
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.803076414174225,
+      "learning_rate": 3.5741030372418763e-06,
+      "loss": 0.9786,
+      "step": 18830
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.692665844872918,
+      "learning_rate": 3.571292080082589e-06,
+      "loss": 0.9742,
+      "step": 18835
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.049139761426119,
+      "learning_rate": 3.568481614540332e-06,
+      "loss": 0.961,
+      "step": 18840
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.999274994098982,
+      "learning_rate": 3.5656716415821848e-06,
+      "loss": 0.9545,
+      "step": 18845
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.444020322274279,
+      "learning_rate": 3.5628621621750546e-06,
+      "loss": 0.9868,
+      "step": 18850
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.935898973123527,
+      "learning_rate": 3.560053177285683e-06,
+      "loss": 1.0229,
+      "step": 18855
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.652059327456081,
+      "learning_rate": 3.5572446878806356e-06,
+      "loss": 0.9985,
+      "step": 18860
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.470800195437701,
+      "learning_rate": 3.5544366949263154e-06,
+      "loss": 0.974,
+      "step": 18865
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.540500894353744,
+      "learning_rate": 3.5516291993889463e-06,
+      "loss": 0.9712,
+      "step": 18870
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.903932248809819,
+      "learning_rate": 3.548822202234588e-06,
+      "loss": 1.0247,
+      "step": 18875
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.255685901803376,
+      "learning_rate": 3.546015704429123e-06,
+      "loss": 0.9893,
+      "step": 18880
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.7442920466087815,
+      "learning_rate": 3.5432097069382652e-06,
+      "loss": 0.9865,
+      "step": 18885
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.451898088603783,
+      "learning_rate": 3.540404210727555e-06,
+      "loss": 1.0302,
+      "step": 18890
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.078002477744629,
+      "learning_rate": 3.5375992167623635e-06,
+      "loss": 0.9517,
+      "step": 18895
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.8441422291882414,
+      "learning_rate": 3.534794726007884e-06,
+      "loss": 1.0022,
+      "step": 18900
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.820516107934559,
+      "learning_rate": 3.531990739429141e-06,
+      "loss": 0.9455,
+      "step": 18905
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.986413654613702,
+      "learning_rate": 3.52918725799098e-06,
+      "loss": 0.9679,
+      "step": 18910
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.527089802422374,
+      "learning_rate": 3.5263842826580813e-06,
+      "loss": 0.9618,
+      "step": 18915
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.444281911919659,
+      "learning_rate": 3.523581814394941e-06,
+      "loss": 0.993,
+      "step": 18920
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.782005747340329,
+      "learning_rate": 3.52077985416589e-06,
+      "loss": 0.9753,
+      "step": 18925
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.74891044127738,
+      "learning_rate": 3.517978402935077e-06,
+      "loss": 0.9723,
+      "step": 18930
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.758541435324947,
+      "learning_rate": 3.5151774616664825e-06,
+      "loss": 0.9603,
+      "step": 18935
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.337683199317084,
+      "learning_rate": 3.5123770313239035e-06,
+      "loss": 0.9969,
+      "step": 18940
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.297019442567877,
+      "learning_rate": 3.5095771128709678e-06,
+      "loss": 0.9789,
+      "step": 18945
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.913586941747046,
+      "learning_rate": 3.506777707271124e-06,
+      "loss": 0.9492,
+      "step": 18950
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.048817929045729,
+      "learning_rate": 3.503978815487643e-06,
+      "loss": 1.0014,
+      "step": 18955
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.630358819181739,
+      "learning_rate": 3.501180438483626e-06,
+      "loss": 0.9702,
+      "step": 18960
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.888623974867344,
+      "learning_rate": 3.4983825772219858e-06,
+      "loss": 0.9936,
+      "step": 18965
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.889481060326786,
+      "learning_rate": 3.4955852326654683e-06,
+      "loss": 0.987,
+      "step": 18970
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.851092223043249,
+      "learning_rate": 3.4927884057766344e-06,
+      "loss": 0.9532,
+      "step": 18975
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 21.96452203997625,
+      "learning_rate": 3.4899920975178714e-06,
+      "loss": 0.9801,
+      "step": 18980
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.662043974615656,
+      "learning_rate": 3.4871963088513833e-06,
+      "loss": 0.9786,
+      "step": 18985
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.3324544231113515,
+      "learning_rate": 3.4844010407392035e-06,
+      "loss": 0.9684,
+      "step": 18990
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.410229456737007,
+      "learning_rate": 3.481606294143175e-06,
+      "loss": 0.9266,
+      "step": 18995
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.723097172662212,
+      "learning_rate": 3.4788120700249726e-06,
+      "loss": 0.967,
+      "step": 19000
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.042198027911561,
+      "learning_rate": 3.476018369346083e-06,
+      "loss": 1.0244,
+      "step": 19005
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.48724264317641,
+      "learning_rate": 3.4732251930678184e-06,
+      "loss": 0.9755,
+      "step": 19010
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.070287251272303,
+      "learning_rate": 3.470432542151304e-06,
+      "loss": 0.9775,
+      "step": 19015
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.767740288674974,
+      "learning_rate": 3.467640417557493e-06,
+      "loss": 0.983,
+      "step": 19020
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.745129466956522,
+      "learning_rate": 3.4648488202471497e-06,
+      "loss": 0.9818,
+      "step": 19025
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.653727961877179,
+      "learning_rate": 3.4620577511808623e-06,
+      "loss": 0.9817,
+      "step": 19030
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.797091615802747,
+      "learning_rate": 3.459267211319033e-06,
+      "loss": 1.0586,
+      "step": 19035
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.293391404977797,
+      "learning_rate": 3.456477201621885e-06,
+      "loss": 0.9574,
+      "step": 19040
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.905429069319764,
+      "learning_rate": 3.4536877230494562e-06,
+      "loss": 0.9757,
+      "step": 19045
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.478601961947565,
+      "learning_rate": 3.4508987765616075e-06,
+      "loss": 0.9561,
+      "step": 19050
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.63894396760112,
+      "learning_rate": 3.4481103631180087e-06,
+      "loss": 0.9583,
+      "step": 19055
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.789057388553128,
+      "learning_rate": 3.4453224836781526e-06,
+      "loss": 1.0379,
+      "step": 19060
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.728543474370344,
+      "learning_rate": 3.442535139201345e-06,
+      "loss": 0.9697,
+      "step": 19065
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.968113376227555,
+      "learning_rate": 3.439748330646707e-06,
+      "loss": 0.9938,
+      "step": 19070
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.356592255128312,
+      "learning_rate": 3.43696205897318e-06,
+      "loss": 0.9788,
+      "step": 19075
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.07420793847312,
+      "learning_rate": 3.434176325139513e-06,
+      "loss": 0.9548,
+      "step": 19080
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.565583992430048,
+      "learning_rate": 3.4313911301042813e-06,
+      "loss": 1.0289,
+      "step": 19085
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.439259933471051,
+      "learning_rate": 3.4286064748258597e-06,
+      "loss": 0.9554,
+      "step": 19090
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.951354380025263,
+      "learning_rate": 3.4258223602624507e-06,
+      "loss": 1.0033,
+      "step": 19095
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.20930073969771,
+      "learning_rate": 3.4230387873720637e-06,
+      "loss": 0.9792,
+      "step": 19100
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.283083435209268,
+      "learning_rate": 3.4202557571125245e-06,
+      "loss": 0.9718,
+      "step": 19105
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.211683982481918,
+      "learning_rate": 3.417473270441468e-06,
+      "loss": 0.9909,
+      "step": 19110
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.417928327129392,
+      "learning_rate": 3.414691328316352e-06,
+      "loss": 0.9361,
+      "step": 19115
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.890075038626392,
+      "learning_rate": 3.4119099316944315e-06,
+      "loss": 0.9219,
+      "step": 19120
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.139042437693465,
+      "learning_rate": 3.4091290815327893e-06,
+      "loss": 0.9656,
+      "step": 19125
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.149139570733197,
+      "learning_rate": 3.40634877878831e-06,
+      "loss": 0.9365,
+      "step": 19130
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.084743495998689,
+      "learning_rate": 3.4035690244176946e-06,
+      "loss": 0.9949,
+      "step": 19135
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.122993624208971,
+      "learning_rate": 3.4007898193774503e-06,
+      "loss": 0.9489,
+      "step": 19140
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.45783174606669,
+      "learning_rate": 3.3980111646239056e-06,
+      "loss": 0.9574,
+      "step": 19145
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.3945960690069,
+      "learning_rate": 3.3952330611131857e-06,
+      "loss": 1.0276,
+      "step": 19150
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.510608770178116,
+      "learning_rate": 3.392455509801239e-06,
+      "loss": 0.9959,
+      "step": 19155
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.704575515763283,
+      "learning_rate": 3.3896785116438145e-06,
+      "loss": 0.9894,
+      "step": 19160
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.801312030785498,
+      "learning_rate": 3.3869020675964777e-06,
+      "loss": 1.0108,
+      "step": 19165
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.595579495569428,
+      "learning_rate": 3.3841261786145963e-06,
+      "loss": 0.9596,
+      "step": 19170
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.823529261050941,
+      "learning_rate": 3.3813508456533566e-06,
+      "loss": 1.0178,
+      "step": 19175
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.252521117776325,
+      "learning_rate": 3.3785760696677424e-06,
+      "loss": 0.966,
+      "step": 19180
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.811292947117277,
+      "learning_rate": 3.3758018516125556e-06,
+      "loss": 0.9748,
+      "step": 19185
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.916508964969845,
+      "learning_rate": 3.3730281924423993e-06,
+      "loss": 0.9212,
+      "step": 19190
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 26.873127984778893,
+      "learning_rate": 3.370255093111689e-06,
+      "loss": 0.9369,
+      "step": 19195
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 14.070726015621453,
+      "learning_rate": 3.367482554574645e-06,
+      "loss": 1.0289,
+      "step": 19200
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.370934105921357,
+      "learning_rate": 3.364710577785293e-06,
+      "loss": 0.9708,
+      "step": 19205
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.821227512347984,
+      "learning_rate": 3.361939163697471e-06,
+      "loss": 0.9436,
+      "step": 19210
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.814182159827276,
+      "learning_rate": 3.3591683132648156e-06,
+      "loss": 0.9094,
+      "step": 19215
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.957443256059856,
+      "learning_rate": 3.3563980274407793e-06,
+      "loss": 0.9984,
+      "step": 19220
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.935732607255355,
+      "learning_rate": 3.3536283071786103e-06,
+      "loss": 1.0013,
+      "step": 19225
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.855885109582275,
+      "learning_rate": 3.3508591534313697e-06,
+      "loss": 0.9671,
+      "step": 19230
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.1394970315511355,
+      "learning_rate": 3.3480905671519183e-06,
+      "loss": 0.9849,
+      "step": 19235
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.813034310776297,
+      "learning_rate": 3.3453225492929255e-06,
+      "loss": 1.0069,
+      "step": 19240
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.7170498460960255,
+      "learning_rate": 3.3425551008068613e-06,
+      "loss": 0.9643,
+      "step": 19245
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.363155157270702,
+      "learning_rate": 3.339788222646006e-06,
+      "loss": 0.9958,
+      "step": 19250
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.452011498593757,
+      "learning_rate": 3.3370219157624363e-06,
+      "loss": 1.0531,
+      "step": 19255
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.577825743787345,
+      "learning_rate": 3.334256181108039e-06,
+      "loss": 0.9736,
+      "step": 19260
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.352973183288714,
+      "learning_rate": 3.331491019634496e-06,
+      "loss": 0.987,
+      "step": 19265
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.162757418382164,
+      "learning_rate": 3.3287264322933023e-06,
+      "loss": 0.9249,
+      "step": 19270
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 20.131892472404356,
+      "learning_rate": 3.3259624200357444e-06,
+      "loss": 0.979,
+      "step": 19275
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.198175991509779,
+      "learning_rate": 3.323198983812921e-06,
+      "loss": 0.953,
+      "step": 19280
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.437106390168473,
+      "learning_rate": 3.3204361245757242e-06,
+      "loss": 0.9871,
+      "step": 19285
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.5062723968359535,
+      "learning_rate": 3.3176738432748547e-06,
+      "loss": 1.0045,
+      "step": 19290
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.170068060454293,
+      "learning_rate": 3.3149121408608063e-06,
+      "loss": 0.9387,
+      "step": 19295
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 13.939999849704586,
+      "learning_rate": 3.3121510182838833e-06,
+      "loss": 0.9707,
+      "step": 19300
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 13.588659684745384,
+      "learning_rate": 3.3093904764941804e-06,
+      "loss": 1.0645,
+      "step": 19305
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.260854066268349,
+      "learning_rate": 3.306630516441602e-06,
+      "loss": 1.0016,
+      "step": 19310
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.626575610265271,
+      "learning_rate": 3.3038711390758436e-06,
+      "loss": 0.9482,
+      "step": 19315
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.659383220388749,
+      "learning_rate": 3.3011123453464088e-06,
+      "loss": 0.9535,
+      "step": 19320
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.15180703898682,
+      "learning_rate": 3.298354136202592e-06,
+      "loss": 0.9649,
+      "step": 19325
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.692113458158882,
+      "learning_rate": 3.2955965125934913e-06,
+      "loss": 0.9702,
+      "step": 19330
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.830959718212891,
+      "learning_rate": 3.2928394754680037e-06,
+      "loss": 0.9656,
+      "step": 19335
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.298851888358607,
+      "learning_rate": 3.29008302577482e-06,
+      "loss": 1.0396,
+      "step": 19340
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.562326510048589,
+      "learning_rate": 3.287327164462436e-06,
+      "loss": 0.9582,
+      "step": 19345
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.80680385409349,
+      "learning_rate": 3.284571892479138e-06,
+      "loss": 0.9311,
+      "step": 19350
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.586249629628685,
+      "learning_rate": 3.281817210773014e-06,
+      "loss": 0.9851,
+      "step": 19355
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.72997718645396,
+      "learning_rate": 3.279063120291946e-06,
+      "loss": 0.9776,
+      "step": 19360
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.134953220297068,
+      "learning_rate": 3.2763096219836153e-06,
+      "loss": 1.0368,
+      "step": 19365
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 20.821681569213887,
+      "learning_rate": 3.273556716795496e-06,
+      "loss": 0.9414,
+      "step": 19370
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.563927082040128,
+      "learning_rate": 3.270804405674863e-06,
+      "loss": 0.968,
+      "step": 19375
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 15.408947715661137,
+      "learning_rate": 3.2680526895687813e-06,
+      "loss": 0.9527,
+      "step": 19380
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.582628546020413,
+      "learning_rate": 3.2653015694241166e-06,
+      "loss": 0.9667,
+      "step": 19385
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.582224579638049,
+      "learning_rate": 3.2625510461875238e-06,
+      "loss": 0.9966,
+      "step": 19390
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.985900200531787,
+      "learning_rate": 3.259801120805457e-06,
+      "loss": 0.9541,
+      "step": 19395
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.757010012138148,
+      "learning_rate": 3.257051794224161e-06,
+      "loss": 0.9379,
+      "step": 19400
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.71433219862553,
+      "learning_rate": 3.25430306738968e-06,
+      "loss": 0.944,
+      "step": 19405
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.147343628896916,
+      "learning_rate": 3.2515549412478446e-06,
+      "loss": 0.9509,
+      "step": 19410
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.409899027959961,
+      "learning_rate": 3.2488074167442852e-06,
+      "loss": 0.9639,
+      "step": 19415
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.426059907250027,
+      "learning_rate": 3.24606049482442e-06,
+      "loss": 0.9638,
+      "step": 19420
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.580312941706315,
+      "learning_rate": 3.2433141764334654e-06,
+      "loss": 0.914,
+      "step": 19425
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.71500348245436,
+      "learning_rate": 3.2405684625164213e-06,
+      "loss": 0.9696,
+      "step": 19430
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.001431989529735,
+      "learning_rate": 3.2378233540180913e-06,
+      "loss": 0.938,
+      "step": 19435
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 25.185513924432797,
+      "learning_rate": 3.2350788518830617e-06,
+      "loss": 0.9514,
+      "step": 19440
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.790210054110683,
+      "learning_rate": 3.232334957055714e-06,
+      "loss": 0.9617,
+      "step": 19445
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.369432007064662,
+      "learning_rate": 3.2295916704802194e-06,
+      "loss": 0.9363,
+      "step": 19450
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.8460514741934,
+      "learning_rate": 3.2268489931005385e-06,
+      "loss": 0.9848,
+      "step": 19455
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 14.77954889982426,
+      "learning_rate": 3.2241069258604273e-06,
+      "loss": 0.9394,
+      "step": 19460
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.971049752470999,
+      "learning_rate": 3.2213654697034246e-06,
+      "loss": 0.9953,
+      "step": 19465
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.583984995729754,
+      "learning_rate": 3.218624625572867e-06,
+      "loss": 0.9636,
+      "step": 19470
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.056659589103473,
+      "learning_rate": 3.2158843944118735e-06,
+      "loss": 1.0098,
+      "step": 19475
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.1372502108362665,
+      "learning_rate": 3.213144777163356e-06,
+      "loss": 0.9576,
+      "step": 19480
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.742973523888802,
+      "learning_rate": 3.2104057747700133e-06,
+      "loss": 1.012,
+      "step": 19485
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.8678377346229755,
+      "learning_rate": 3.207667388174336e-06,
+      "loss": 0.9343,
+      "step": 19490
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.993960911914766,
+      "learning_rate": 3.204929618318596e-06,
+      "loss": 0.9503,
+      "step": 19495
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.897187382469572,
+      "learning_rate": 3.2021924661448617e-06,
+      "loss": 0.9685,
+      "step": 19500
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.06865849915388,
+      "learning_rate": 3.199455932594982e-06,
+      "loss": 0.9807,
+      "step": 19505
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.445746655511793,
+      "learning_rate": 3.1967200186105975e-06,
+      "loss": 0.9827,
+      "step": 19510
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.303285449271074,
+      "learning_rate": 3.19398472513313e-06,
+      "loss": 0.9557,
+      "step": 19515
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.096354209566563,
+      "learning_rate": 3.191250053103796e-06,
+      "loss": 1.0006,
+      "step": 19520
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.044059243281657,
+      "learning_rate": 3.1885160034635885e-06,
+      "loss": 0.9825,
+      "step": 19525
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.238962956982464,
+      "learning_rate": 3.185782577153296e-06,
+      "loss": 0.986,
+      "step": 19530
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.294757392118468,
+      "learning_rate": 3.183049775113485e-06,
+      "loss": 0.947,
+      "step": 19535
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.90709327082817,
+      "learning_rate": 3.1803175982845124e-06,
+      "loss": 0.9887,
+      "step": 19540
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.1987411375622825,
+      "learning_rate": 3.1775860476065147e-06,
+      "loss": 0.9221,
+      "step": 19545
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.084108859024125,
+      "learning_rate": 3.1748551240194192e-06,
+      "loss": 0.992,
+      "step": 19550
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.930419426018079,
+      "learning_rate": 3.1721248284629293e-06,
+      "loss": 0.9964,
+      "step": 19555
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.018806857675019,
+      "learning_rate": 3.169395161876543e-06,
+      "loss": 0.9697,
+      "step": 19560
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.210110614847695,
+      "learning_rate": 3.1666661251995313e-06,
+      "loss": 0.9453,
+      "step": 19565
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.061761204356091,
+      "learning_rate": 3.1639377193709564e-06,
+      "loss": 0.9771,
+      "step": 19570
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.907670275453538,
+      "learning_rate": 3.1612099453296575e-06,
+      "loss": 0.9937,
+      "step": 19575
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 6.268881229619832,
+      "learning_rate": 3.158482804014261e-06,
+      "loss": 0.9415,
+      "step": 19580
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 6.410057104629809,
+      "learning_rate": 3.1557562963631728e-06,
+      "loss": 0.9235,
+      "step": 19585
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.225154512292546,
+      "learning_rate": 3.153030423314579e-06,
+      "loss": 0.9581,
+      "step": 19590
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.5527394175980005,
+      "learning_rate": 3.150305185806457e-06,
+      "loss": 0.9861,
+      "step": 19595
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.332123472990733,
+      "learning_rate": 3.1475805847765497e-06,
+      "loss": 0.993,
+      "step": 19600
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.367055598439357,
+      "learning_rate": 3.1448566211623964e-06,
+      "loss": 1.009,
+      "step": 19605
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.04981083164468,
+      "learning_rate": 3.1421332959013063e-06,
+      "loss": 0.9901,
+      "step": 19610
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.850619560943523,
+      "learning_rate": 3.1394106099303766e-06,
+      "loss": 0.9711,
+      "step": 19615
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.190983677356039,
+      "learning_rate": 3.1366885641864764e-06,
+      "loss": 0.974,
+      "step": 19620
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.919928057436149,
+      "learning_rate": 3.1339671596062642e-06,
+      "loss": 0.9464,
+      "step": 19625
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.208963956216042,
+      "learning_rate": 3.1312463971261675e-06,
+      "loss": 0.9159,
+      "step": 19630
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.998212896605247,
+      "learning_rate": 3.1285262776824033e-06,
+      "loss": 0.918,
+      "step": 19635
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.484371252595751,
+      "learning_rate": 3.125806802210958e-06,
+      "loss": 0.9785,
+      "step": 19640
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.07006168798367,
+      "learning_rate": 3.1230879716476027e-06,
+      "loss": 0.9834,
+      "step": 19645
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.719973146813349,
+      "learning_rate": 3.1203697869278815e-06,
+      "loss": 0.9671,
+      "step": 19650
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.906113261291717,
+      "learning_rate": 3.117652248987125e-06,
+      "loss": 0.9734,
+      "step": 19655
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.461870871497509,
+      "learning_rate": 3.1149353587604282e-06,
+      "loss": 0.9461,
+      "step": 19660
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.962793748592961,
+      "learning_rate": 3.1122191171826765e-06,
+      "loss": 0.974,
+      "step": 19665
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.555865802855921,
+      "learning_rate": 3.109503525188522e-06,
+      "loss": 0.9898,
+      "step": 19670
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.682060107258698,
+      "learning_rate": 3.1067885837124003e-06,
+      "loss": 0.9276,
+      "step": 19675
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.964687635868797,
+      "learning_rate": 3.104074293688517e-06,
+      "loss": 0.9366,
+      "step": 19680
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.943606394724827,
+      "learning_rate": 3.1013606560508606e-06,
+      "loss": 0.9712,
+      "step": 19685
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.623945247762496,
+      "learning_rate": 3.098647671733187e-06,
+      "loss": 0.9687,
+      "step": 19690
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.5336412205470475,
+      "learning_rate": 3.0959353416690354e-06,
+      "loss": 0.9732,
+      "step": 19695
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 21.16146004449491,
+      "learning_rate": 3.0932236667917128e-06,
+      "loss": 0.989,
+      "step": 19700
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.607128787122527,
+      "learning_rate": 3.0905126480343057e-06,
+      "loss": 0.9715,
+      "step": 19705
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.665532346332486,
+      "learning_rate": 3.087802286329673e-06,
+      "loss": 0.9226,
+      "step": 19710
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.992099425488396,
+      "learning_rate": 3.085092582610446e-06,
+      "loss": 0.9462,
+      "step": 19715
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.808277797298123,
+      "learning_rate": 3.0823835378090344e-06,
+      "loss": 1.006,
+      "step": 19720
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.481945211686297,
+      "learning_rate": 3.0796751528576137e-06,
+      "loss": 0.9797,
+      "step": 19725
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.420100717721782,
+      "learning_rate": 3.07696742868814e-06,
+      "loss": 0.9554,
+      "step": 19730
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.261498471883031,
+      "learning_rate": 3.074260366232337e-06,
+      "loss": 0.9386,
+      "step": 19735
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.262869495772408,
+      "learning_rate": 3.071553966421703e-06,
+      "loss": 0.9791,
+      "step": 19740
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.01776762367467,
+      "learning_rate": 3.068848230187506e-06,
+      "loss": 0.9706,
+      "step": 19745
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 16.350092017822924,
+      "learning_rate": 3.0661431584607917e-06,
+      "loss": 0.9205,
+      "step": 19750
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.02862210737073,
+      "learning_rate": 3.063438752172366e-06,
+      "loss": 0.9602,
+      "step": 19755
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.624701360428293,
+      "learning_rate": 3.060735012252818e-06,
+      "loss": 0.9675,
+      "step": 19760
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.022757663300407,
+      "learning_rate": 3.058031939632499e-06,
+      "loss": 0.9546,
+      "step": 19765
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.299124192747938,
+      "learning_rate": 3.0553295352415356e-06,
+      "loss": 0.9571,
+      "step": 19770
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.8246198096307005,
+      "learning_rate": 3.0526278000098208e-06,
+      "loss": 0.9435,
+      "step": 19775
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 18.1215474019368,
+      "learning_rate": 3.0499267348670225e-06,
+      "loss": 0.9766,
+      "step": 19780
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.868272802586499,
+      "learning_rate": 3.047226340742569e-06,
+      "loss": 0.9472,
+      "step": 19785
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.174261883529199,
+      "learning_rate": 3.0445266185656685e-06,
+      "loss": 0.9222,
+      "step": 19790
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.064743033532646,
+      "learning_rate": 3.04182756926529e-06,
+      "loss": 0.986,
+      "step": 19795
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.872229895463509,
+      "learning_rate": 3.0391291937701755e-06,
+      "loss": 0.9669,
+      "step": 19800
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.4018475622049,
+      "learning_rate": 3.0364314930088312e-06,
+      "loss": 0.9524,
+      "step": 19805
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 25.0373690123754,
+      "learning_rate": 3.0337344679095373e-06,
+      "loss": 0.9465,
+      "step": 19810
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.512944296977459,
+      "learning_rate": 3.0310381194003312e-06,
+      "loss": 1.0136,
+      "step": 19815
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 26.48421656057948,
+      "learning_rate": 3.0283424484090306e-06,
+      "loss": 0.9404,
+      "step": 19820
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.807803981551405,
+      "learning_rate": 3.0256474558632093e-06,
+      "loss": 0.9365,
+      "step": 19825
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.631833071132508,
+      "learning_rate": 3.022953142690214e-06,
+      "loss": 0.9636,
+      "step": 19830
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.990611013751563,
+      "learning_rate": 3.0202595098171527e-06,
+      "loss": 0.9332,
+      "step": 19835
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.0710164308547974,
+      "eval_runtime": 25.431,
+      "eval_samples_per_second": 31.694,
+      "eval_steps_per_second": 3.972,
+      "step": 19837
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.07847082531632,
+      "learning_rate": 3.0175665581709035e-06,
+      "loss": 0.8202,
+      "step": 19840
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.554418428839272,
+      "learning_rate": 3.014874288678109e-06,
+      "loss": 0.708,
+      "step": 19845
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.866761074974754,
+      "learning_rate": 3.012182702265173e-06,
+      "loss": 0.6866,
+      "step": 19850
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.599634041133625,
+      "learning_rate": 3.0094917998582737e-06,
+      "loss": 0.693,
+      "step": 19855
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.495053387713696,
+      "learning_rate": 3.006801582383343e-06,
+      "loss": 0.6998,
+      "step": 19860
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 17.031744168199783,
+      "learning_rate": 3.0041120507660854e-06,
+      "loss": 0.6845,
+      "step": 19865
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.8690008816324,
+      "learning_rate": 3.0014232059319636e-06,
+      "loss": 0.6585,
+      "step": 19870
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.786581846929866,
+      "learning_rate": 2.9987350488062088e-06,
+      "loss": 0.7137,
+      "step": 19875
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.322402483787055,
+      "learning_rate": 2.9960475803138105e-06,
+      "loss": 0.6975,
+      "step": 19880
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.447412321402005,
+      "learning_rate": 2.993360801379528e-06,
+      "loss": 0.6839,
+      "step": 19885
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.652872882405284,
+      "learning_rate": 2.9906747129278756e-06,
+      "loss": 0.6754,
+      "step": 19890
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.959284438123414,
+      "learning_rate": 2.9879893158831368e-06,
+      "loss": 0.7099,
+      "step": 19895
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.276126342509745,
+      "learning_rate": 2.985304611169352e-06,
+      "loss": 0.7156,
+      "step": 19900
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.270239118964066,
+      "learning_rate": 2.9826205997103264e-06,
+      "loss": 0.7028,
+      "step": 19905
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.870637246845116,
+      "learning_rate": 2.979937282429625e-06,
+      "loss": 0.6825,
+      "step": 19910
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.8819565680372135,
+      "learning_rate": 2.9772546602505758e-06,
+      "loss": 0.6809,
+      "step": 19915
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.735775614992868,
+      "learning_rate": 2.974572734096266e-06,
+      "loss": 0.6958,
+      "step": 19920
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.125437627956788,
+      "learning_rate": 2.9718915048895442e-06,
+      "loss": 0.7067,
+      "step": 19925
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.669127427169447,
+      "learning_rate": 2.9692109735530183e-06,
+      "loss": 0.6444,
+      "step": 19930
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 12.186094264136873,
+      "learning_rate": 2.966531141009057e-06,
+      "loss": 0.6882,
+      "step": 19935
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.473372562403895,
+      "learning_rate": 2.9638520081797862e-06,
+      "loss": 0.6951,
+      "step": 19940
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.946284342434288,
+      "learning_rate": 2.9611735759870963e-06,
+      "loss": 0.7039,
+      "step": 19945
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.49126548825676,
+      "learning_rate": 2.9584958453526315e-06,
+      "loss": 0.7073,
+      "step": 19950
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 21.897553417108384,
+      "learning_rate": 2.955818817197797e-06,
+      "loss": 0.6855,
+      "step": 19955
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.509726325927025,
+      "learning_rate": 2.9531424924437535e-06,
+      "loss": 0.6387,
+      "step": 19960
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.647114585979262,
+      "learning_rate": 2.950466872011425e-06,
+      "loss": 0.7173,
+      "step": 19965
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 19.553340817495304,
+      "learning_rate": 2.9477919568214886e-06,
+      "loss": 0.7056,
+      "step": 19970
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 7.480061284531499,
+      "learning_rate": 2.9451177477943772e-06,
+      "loss": 0.6694,
+      "step": 19975
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.73237332843965,
+      "learning_rate": 2.9424442458502884e-06,
+      "loss": 0.6902,
+      "step": 19980
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.257596743266556,
+      "learning_rate": 2.939771451909168e-06,
+      "loss": 0.6621,
+      "step": 19985
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 7.007980217929251,
+      "learning_rate": 2.937099366890725e-06,
+      "loss": 0.6742,
+      "step": 19990
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 6.684722771405095,
+      "learning_rate": 2.934427991714418e-06,
+      "loss": 0.6924,
+      "step": 19995
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 6.459688562079147,
+      "learning_rate": 2.931757327299467e-06,
+      "loss": 0.6764,
+      "step": 20000
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 7.166925967390257,
+      "learning_rate": 2.929087374564843e-06,
+      "loss": 0.7016,
+      "step": 20005
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.216661904637062,
+      "learning_rate": 2.9264181344292775e-06,
+      "loss": 0.688,
+      "step": 20010
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.799432873080406,
+      "learning_rate": 2.9237496078112493e-06,
+      "loss": 0.6843,
+      "step": 20015
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.919357986122606,
+      "learning_rate": 2.921081795629e-06,
+      "loss": 0.6726,
+      "step": 20020
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.222286865949679,
+      "learning_rate": 2.918414698800517e-06,
+      "loss": 0.6846,
+      "step": 20025
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.497129639848889,
+      "learning_rate": 2.91574831824355e-06,
+      "loss": 0.6942,
+      "step": 20030
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.741374508164342,
+      "learning_rate": 2.913082654875596e-06,
+      "loss": 0.722,
+      "step": 20035
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.86965920427053,
+      "learning_rate": 2.9104177096139073e-06,
+      "loss": 0.6915,
+      "step": 20040
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.343518561020797,
+      "learning_rate": 2.9077534833754866e-06,
+      "loss": 0.685,
+      "step": 20045
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.7595663507858905,
+      "learning_rate": 2.9050899770770964e-06,
+      "loss": 0.6682,
+      "step": 20050
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.569107304442746,
+      "learning_rate": 2.902427191635242e-06,
+      "loss": 0.6701,
+      "step": 20055
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.941058043606878,
+      "learning_rate": 2.8997651279661893e-06,
+      "loss": 0.6646,
+      "step": 20060
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.2021071826259995,
+      "learning_rate": 2.8971037869859503e-06,
+      "loss": 0.6754,
+      "step": 20065
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 23.181178572182155,
+      "learning_rate": 2.8944431696102898e-06,
+      "loss": 0.6937,
+      "step": 20070
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.262592504800057,
+      "learning_rate": 2.8917832767547215e-06,
+      "loss": 0.6674,
+      "step": 20075
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.445147355846368,
+      "learning_rate": 2.8891241093345163e-06,
+      "loss": 0.7094,
+      "step": 20080
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.449205286580153,
+      "learning_rate": 2.886465668264686e-06,
+      "loss": 0.6716,
+      "step": 20085
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 15.904881029885123,
+      "learning_rate": 2.8838079544600038e-06,
+      "loss": 0.6855,
+      "step": 20090
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 7.317897665887902,
+      "learning_rate": 2.881150968834983e-06,
+      "loss": 0.6476,
+      "step": 20095
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 22.960176156048963,
+      "learning_rate": 2.87849471230389e-06,
+      "loss": 0.6739,
+      "step": 20100
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.81723029382713,
+      "learning_rate": 2.87583918578074e-06,
+      "loss": 0.7035,
+      "step": 20105
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.429570917860615,
+      "learning_rate": 2.8731843901792955e-06,
+      "loss": 0.6859,
+      "step": 20110
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 16.864528516068667,
+      "learning_rate": 2.8705303264130734e-06,
+      "loss": 0.6718,
+      "step": 20115
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.239644650544476,
+      "learning_rate": 2.86787699539533e-06,
+      "loss": 0.655,
+      "step": 20120
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.1466601613579686,
+      "learning_rate": 2.8652243980390778e-06,
+      "loss": 0.7119,
+      "step": 20125
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.394354689435088,
+      "learning_rate": 2.862572535257071e-06,
+      "loss": 0.7041,
+      "step": 20130
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 19.978558529974322,
+      "learning_rate": 2.8599214079618137e-06,
+      "loss": 0.6777,
+      "step": 20135
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.013234273687217,
+      "learning_rate": 2.857271017065554e-06,
+      "loss": 0.6787,
+      "step": 20140
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 7.637030733686688,
+      "learning_rate": 2.854621363480292e-06,
+      "loss": 0.7093,
+      "step": 20145
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.422025071823492,
+      "learning_rate": 2.851972448117769e-06,
+      "loss": 0.7019,
+      "step": 20150
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 9.61149127644432,
+      "learning_rate": 2.8493242718894765e-06,
+      "loss": 0.6949,
+      "step": 20155
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.036412049392237,
+      "learning_rate": 2.8466768357066475e-06,
+      "loss": 0.6843,
+      "step": 20160
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.36780256872287,
+      "learning_rate": 2.844030140480263e-06,
+      "loss": 0.7137,
+      "step": 20165
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.8956494643916875,
+      "learning_rate": 2.8413841871210456e-06,
+      "loss": 0.6885,
+      "step": 20170
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.501107518944112,
+      "learning_rate": 2.83873897653947e-06,
+      "loss": 0.6672,
+      "step": 20175
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.34334702345557,
+      "learning_rate": 2.836094509645747e-06,
+      "loss": 0.6899,
+      "step": 20180
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 7.976060684480714,
+      "learning_rate": 2.8334507873498373e-06,
+      "loss": 0.6979,
+      "step": 20185
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.765949760846961,
+      "learning_rate": 2.8308078105614433e-06,
+      "loss": 0.6678,
+      "step": 20190
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.483903474155636,
+      "learning_rate": 2.8281655801900098e-06,
+      "loss": 0.6894,
+      "step": 20195
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.11059505030005,
+      "learning_rate": 2.8255240971447227e-06,
+      "loss": 0.6772,
+      "step": 20200
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.625996552622085,
+      "learning_rate": 2.8228833623345205e-06,
+      "loss": 0.71,
+      "step": 20205
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.572562001223029,
+      "learning_rate": 2.820243376668071e-06,
+      "loss": 0.6595,
+      "step": 20210
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.759923303950589,
+      "learning_rate": 2.817604141053796e-06,
+      "loss": 0.7204,
+      "step": 20215
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.061007775325533,
+      "learning_rate": 2.8149656563998513e-06,
+      "loss": 0.6661,
+      "step": 20220
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.473550927111044,
+      "learning_rate": 2.8123279236141378e-06,
+      "loss": 0.6577,
+      "step": 20225
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.747404862096733,
+      "learning_rate": 2.809690943604295e-06,
+      "loss": 0.6944,
+      "step": 20230
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 12.424278453040909,
+      "learning_rate": 2.8070547172777063e-06,
+      "loss": 0.6621,
+      "step": 20235
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.392454233816305,
+      "learning_rate": 2.8044192455414955e-06,
+      "loss": 0.675,
+      "step": 20240
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.785097695138289,
+      "learning_rate": 2.8017845293025243e-06,
+      "loss": 0.6736,
+      "step": 20245
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.860239130549238,
+      "learning_rate": 2.7991505694673983e-06,
+      "loss": 0.7173,
+      "step": 20250
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.029730559126252,
+      "learning_rate": 2.7965173669424596e-06,
+      "loss": 0.6879,
+      "step": 20255
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.141573323808583,
+      "learning_rate": 2.7938849226337905e-06,
+      "loss": 0.6831,
+      "step": 20260
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.803515943071768,
+      "learning_rate": 2.79125323744721e-06,
+      "loss": 0.6667,
+      "step": 20265
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.768831960299563,
+      "learning_rate": 2.7886223122882833e-06,
+      "loss": 0.6589,
+      "step": 20270
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.363161119443387,
+      "learning_rate": 2.7859921480623048e-06,
+      "loss": 0.6545,
+      "step": 20275
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.738430484972305,
+      "learning_rate": 2.783362745674315e-06,
+      "loss": 0.6556,
+      "step": 20280
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.157022650903519,
+      "learning_rate": 2.780734106029088e-06,
+      "loss": 0.6538,
+      "step": 20285
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.027913613784868,
+      "learning_rate": 2.7781062300311357e-06,
+      "loss": 0.6428,
+      "step": 20290
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.702997038291723,
+      "learning_rate": 2.7754791185847045e-06,
+      "loss": 0.6968,
+      "step": 20295
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.703632225890155,
+      "learning_rate": 2.7728527725937866e-06,
+      "loss": 0.7008,
+      "step": 20300
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.326646266181257,
+      "learning_rate": 2.7702271929621004e-06,
+      "loss": 0.6359,
+      "step": 20305
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.596957107489121,
+      "learning_rate": 2.7676023805931096e-06,
+      "loss": 0.6543,
+      "step": 20310
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.4689293897811275,
+      "learning_rate": 2.764978336390007e-06,
+      "loss": 0.6543,
+      "step": 20315
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.592063799128114,
+      "learning_rate": 2.762355061255724e-06,
+      "loss": 0.6456,
+      "step": 20320
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.724734726353038,
+      "learning_rate": 2.7597325560929254e-06,
+      "loss": 0.6652,
+      "step": 20325
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.5520198384323365,
+      "learning_rate": 2.7571108218040172e-06,
+      "loss": 0.6667,
+      "step": 20330
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.7006276737748305,
+      "learning_rate": 2.7544898592911305e-06,
+      "loss": 0.6714,
+      "step": 20335
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.035710997650199,
+      "learning_rate": 2.7518696694561402e-06,
+      "loss": 0.663,
+      "step": 20340
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.984785241111086,
+      "learning_rate": 2.7492502532006498e-06,
+      "loss": 0.6972,
+      "step": 20345
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.48809346091373,
+      "learning_rate": 2.746631611425997e-06,
+      "loss": 0.6877,
+      "step": 20350
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.822505788690949,
+      "learning_rate": 2.744013745033256e-06,
+      "loss": 0.6788,
+      "step": 20355
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.309126432519314,
+      "learning_rate": 2.7413966549232274e-06,
+      "loss": 0.6215,
+      "step": 20360
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.376856836130756,
+      "learning_rate": 2.7387803419964547e-06,
+      "loss": 0.6749,
+      "step": 20365
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.955550699410014,
+      "learning_rate": 2.736164807153205e-06,
+      "loss": 0.6753,
+      "step": 20370
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.158760477300898,
+      "learning_rate": 2.7335500512934844e-06,
+      "loss": 0.657,
+      "step": 20375
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.839000626662179,
+      "learning_rate": 2.7309360753170267e-06,
+      "loss": 0.698,
+      "step": 20380
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.465886089969563,
+      "learning_rate": 2.7283228801232985e-06,
+      "loss": 0.6795,
+      "step": 20385
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 11.638003524690347,
+      "learning_rate": 2.725710466611495e-06,
+      "loss": 0.6578,
+      "step": 20390
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.555120482903076,
+      "learning_rate": 2.7230988356805502e-06,
+      "loss": 0.6611,
+      "step": 20395
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.007908263827729,
+      "learning_rate": 2.7204879882291195e-06,
+      "loss": 0.711,
+      "step": 20400
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.52058789317427,
+      "learning_rate": 2.7178779251555963e-06,
+      "loss": 0.686,
+      "step": 20405
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.290434833099377,
+      "learning_rate": 2.7152686473581e-06,
+      "loss": 0.6597,
+      "step": 20410
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.865850991334291,
+      "learning_rate": 2.71266015573448e-06,
+      "loss": 0.6921,
+      "step": 20415
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.093957021899156,
+      "learning_rate": 2.7100524511823134e-06,
+      "loss": 0.6589,
+      "step": 20420
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 12.990284071684462,
+      "learning_rate": 2.7074455345989128e-06,
+      "loss": 0.6583,
+      "step": 20425
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.045778809474987,
+      "learning_rate": 2.7048394068813127e-06,
+      "loss": 0.6724,
+      "step": 20430
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.497086333725814,
+      "learning_rate": 2.7022340689262827e-06,
+      "loss": 0.6372,
+      "step": 20435
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 12.37157407740874,
+      "learning_rate": 2.6996295216303147e-06,
+      "loss": 0.6974,
+      "step": 20440
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.893442888000223,
+      "learning_rate": 2.697025765889631e-06,
+      "loss": 0.6669,
+      "step": 20445
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.408198005557214,
+      "learning_rate": 2.6944228026001795e-06,
+      "loss": 0.6511,
+      "step": 20450
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.167101911384638,
+      "learning_rate": 2.6918206326576424e-06,
+      "loss": 0.6442,
+      "step": 20455
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.727500287228669,
+      "learning_rate": 2.6892192569574183e-06,
+      "loss": 0.6843,
+      "step": 20460
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.898389179651797,
+      "learning_rate": 2.6866186763946426e-06,
+      "loss": 0.6685,
+      "step": 20465
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.218251143268944,
+      "learning_rate": 2.684018891864172e-06,
+      "loss": 0.6747,
+      "step": 20470
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.851991764395169,
+      "learning_rate": 2.681419904260587e-06,
+      "loss": 0.6963,
+      "step": 20475
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.618853238599201,
+      "learning_rate": 2.6788217144781995e-06,
+      "loss": 0.6651,
+      "step": 20480
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.94027051266893,
+      "learning_rate": 2.6762243234110415e-06,
+      "loss": 0.688,
+      "step": 20485
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.675509761541878,
+      "learning_rate": 2.6736277319528757e-06,
+      "loss": 0.7261,
+      "step": 20490
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.93388076006707,
+      "learning_rate": 2.6710319409971837e-06,
+      "loss": 0.7099,
+      "step": 20495
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.709835978825093,
+      "learning_rate": 2.66843695143718e-06,
+      "loss": 0.7102,
+      "step": 20500
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.807707734312976,
+      "learning_rate": 2.6658427641657903e-06,
+      "loss": 0.6731,
+      "step": 20505
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.902988562763788,
+      "learning_rate": 2.6632493800756787e-06,
+      "loss": 0.6528,
+      "step": 20510
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.921829168490332,
+      "learning_rate": 2.6606568000592214e-06,
+      "loss": 0.697,
+      "step": 20515
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.402784949687757,
+      "learning_rate": 2.658065025008526e-06,
+      "loss": 0.6611,
+      "step": 20520
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.516261871930669,
+      "learning_rate": 2.655474055815418e-06,
+      "loss": 0.6624,
+      "step": 20525
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.161141461701427,
+      "learning_rate": 2.6528838933714514e-06,
+      "loss": 0.6762,
+      "step": 20530
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.941845927863708,
+      "learning_rate": 2.6502945385678913e-06,
+      "loss": 0.6946,
+      "step": 20535
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.659199780653852,
+      "learning_rate": 2.6477059922957384e-06,
+      "loss": 0.6783,
+      "step": 20540
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 7.124417827033684,
+      "learning_rate": 2.645118255445705e-06,
+      "loss": 0.6573,
+      "step": 20545
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.85559312595928,
+      "learning_rate": 2.642531328908232e-06,
+      "loss": 0.6704,
+      "step": 20550
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.770810220827577,
+      "learning_rate": 2.639945213573476e-06,
+      "loss": 0.6672,
+      "step": 20555
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.784653605858367,
+      "learning_rate": 2.637359910331322e-06,
+      "loss": 0.6597,
+      "step": 20560
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.517438292447058,
+      "learning_rate": 2.6347754200713616e-06,
+      "loss": 0.6762,
+      "step": 20565
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 7.121687089288357,
+      "learning_rate": 2.6321917436829226e-06,
+      "loss": 0.676,
+      "step": 20570
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.757356735754648,
+      "learning_rate": 2.6296088820550402e-06,
+      "loss": 0.6445,
+      "step": 20575
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.034457247639295,
+      "learning_rate": 2.6270268360764807e-06,
+      "loss": 0.689,
+      "step": 20580
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.2361163711605565,
+      "learning_rate": 2.6244456066357176e-06,
+      "loss": 0.6888,
+      "step": 20585
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.735153898853402,
+      "learning_rate": 2.6218651946209543e-06,
+      "loss": 0.7077,
+      "step": 20590
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.132323732813553,
+      "learning_rate": 2.619285600920107e-06,
+      "loss": 0.6814,
+      "step": 20595
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.56491215584217,
+      "learning_rate": 2.6167068264208096e-06,
+      "loss": 0.6309,
+      "step": 20600
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.420679368081689,
+      "learning_rate": 2.6141288720104153e-06,
+      "loss": 0.6567,
+      "step": 20605
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.678522405810549,
+      "learning_rate": 2.611551738576e-06,
+      "loss": 0.6773,
+      "step": 20610
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.886865696835705,
+      "learning_rate": 2.6089754270043493e-06,
+      "loss": 0.6665,
+      "step": 20615
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.956127785995881,
+      "learning_rate": 2.6063999381819686e-06,
+      "loss": 0.6539,
+      "step": 20620
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.519226867216945,
+      "learning_rate": 2.6038252729950877e-06,
+      "loss": 0.695,
+      "step": 20625
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.7267256993890925,
+      "learning_rate": 2.601251432329637e-06,
+      "loss": 0.6967,
+      "step": 20630
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.823013090894927,
+      "learning_rate": 2.5986784170712797e-06,
+      "loss": 0.7371,
+      "step": 20635
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.526737098118042,
+      "learning_rate": 2.5961062281053838e-06,
+      "loss": 0.6591,
+      "step": 20640
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.339338277834594,
+      "learning_rate": 2.59353486631704e-06,
+      "loss": 0.6972,
+      "step": 20645
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.65317210480881,
+      "learning_rate": 2.5909643325910495e-06,
+      "loss": 0.6889,
+      "step": 20650
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.956832370518871,
+      "learning_rate": 2.588394627811934e-06,
+      "loss": 0.709,
+      "step": 20655
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.067865571460534,
+      "learning_rate": 2.5858257528639208e-06,
+      "loss": 0.713,
+      "step": 20660
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.12917377194101,
+      "learning_rate": 2.583257708630962e-06,
+      "loss": 0.626,
+      "step": 20665
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.907861023980704,
+      "learning_rate": 2.580690495996715e-06,
+      "loss": 0.6633,
+      "step": 20670
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.7192338329918355,
+      "learning_rate": 2.578124115844559e-06,
+      "loss": 0.6593,
+      "step": 20675
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 10.403310083930933,
+      "learning_rate": 2.5755585690575808e-06,
+      "loss": 0.6623,
+      "step": 20680
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.040667531817449,
+      "learning_rate": 2.572993856518587e-06,
+      "loss": 0.6759,
+      "step": 20685
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.139888145277165,
+      "learning_rate": 2.570429979110085e-06,
+      "loss": 0.6965,
+      "step": 20690
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 12.342407045299352,
+      "learning_rate": 2.5678669377143082e-06,
+      "loss": 0.6686,
+      "step": 20695
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 8.900244543868103,
+      "learning_rate": 2.5653047332131943e-06,
+      "loss": 0.6473,
+      "step": 20700
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.137426285227035,
+      "learning_rate": 2.562743366488397e-06,
+      "loss": 0.6422,
+      "step": 20705
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.895202332079677,
+      "learning_rate": 2.560182838421279e-06,
+      "loss": 0.633,
+      "step": 20710
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.653129772734897,
+      "learning_rate": 2.5576231498929187e-06,
+      "loss": 0.6608,
+      "step": 20715
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.0133160033021715,
+      "learning_rate": 2.5550643017840966e-06,
+      "loss": 0.6672,
+      "step": 20720
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.671373077072103,
+      "learning_rate": 2.5525062949753143e-06,
+      "loss": 0.6669,
+      "step": 20725
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.510727721272181,
+      "learning_rate": 2.549949130346777e-06,
+      "loss": 0.6768,
+      "step": 20730
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.834607513541353,
+      "learning_rate": 2.547392808778405e-06,
+      "loss": 0.7069,
+      "step": 20735
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.60320911064297,
+      "learning_rate": 2.544837331149826e-06,
+      "loss": 0.6845,
+      "step": 20740
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.333675816136024,
+      "learning_rate": 2.542282698340376e-06,
+      "loss": 0.6757,
+      "step": 20745
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.605720986577917,
+      "learning_rate": 2.5397289112291016e-06,
+      "loss": 0.6295,
+      "step": 20750
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.79014322646118,
+      "learning_rate": 2.5371759706947575e-06,
+      "loss": 0.6647,
+      "step": 20755
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 8.1075613008192,
+      "learning_rate": 2.5346238776158115e-06,
+      "loss": 0.6826,
+      "step": 20760
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.475051681081209,
+      "learning_rate": 2.532072632870433e-06,
+      "loss": 0.6881,
+      "step": 20765
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.665453634646515,
+      "learning_rate": 2.5295222373365056e-06,
+      "loss": 0.6741,
+      "step": 20770
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.6573931101364465,
+      "learning_rate": 2.526972691891617e-06,
+      "loss": 0.661,
+      "step": 20775
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 8.297920740915522,
+      "learning_rate": 2.5244239974130637e-06,
+      "loss": 0.6929,
+      "step": 20780
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.549276471188224,
+      "learning_rate": 2.5218761547778457e-06,
+      "loss": 0.6911,
+      "step": 20785
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.854245160085682,
+      "learning_rate": 2.519329164862678e-06,
+      "loss": 0.6927,
+      "step": 20790
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.596145059140424,
+      "learning_rate": 2.5167830285439743e-06,
+      "loss": 0.6333,
+      "step": 20795
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.27937273157563,
+      "learning_rate": 2.5142377466978598e-06,
+      "loss": 0.688,
+      "step": 20800
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 8.834287310597626,
+      "learning_rate": 2.5116933202001627e-06,
+      "loss": 0.683,
+      "step": 20805
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.1546276439154175,
+      "learning_rate": 2.509149749926417e-06,
+      "loss": 0.6488,
+      "step": 20810
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.963786781997544,
+      "learning_rate": 2.5066070367518624e-06,
+      "loss": 0.6578,
+      "step": 20815
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.901247568524524,
+      "learning_rate": 2.5040651815514465e-06,
+      "loss": 0.6575,
+      "step": 20820
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.8133936919905835,
+      "learning_rate": 2.5015241851998156e-06,
+      "loss": 0.6654,
+      "step": 20825
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 9.8028683121827,
+      "learning_rate": 2.4989840485713287e-06,
+      "loss": 0.6362,
+      "step": 20830
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 11.757108818862323,
+      "learning_rate": 2.4964447725400417e-06,
+      "loss": 0.6458,
+      "step": 20835
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.1336993367309995,
+      "learning_rate": 2.493906357979717e-06,
+      "loss": 0.6796,
+      "step": 20840
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.6868173155668975,
+      "learning_rate": 2.491368805763819e-06,
+      "loss": 0.6749,
+      "step": 20845
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.181757307196955,
+      "learning_rate": 2.488832116765521e-06,
+      "loss": 0.6935,
+      "step": 20850
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.879231867941652,
+      "learning_rate": 2.486296291857691e-06,
+      "loss": 0.6466,
+      "step": 20855
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.210616346246946,
+      "learning_rate": 2.4837613319129082e-06,
+      "loss": 0.6493,
+      "step": 20860
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.092954661412917,
+      "learning_rate": 2.481227237803448e-06,
+      "loss": 0.686,
+      "step": 20865
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.902697895006476,
+      "learning_rate": 2.4786940104012897e-06,
+      "loss": 0.6926,
+      "step": 20870
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.0086173784932,
+      "learning_rate": 2.4761616505781137e-06,
+      "loss": 0.6361,
+      "step": 20875
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.193635820587068,
+      "learning_rate": 2.473630159205302e-06,
+      "loss": 0.6792,
+      "step": 20880
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.165631840813355,
+      "learning_rate": 2.471099537153941e-06,
+      "loss": 0.654,
+      "step": 20885
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.861817478106466,
+      "learning_rate": 2.468569785294812e-06,
+      "loss": 0.6763,
+      "step": 20890
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.372317983836856,
+      "learning_rate": 2.466040904498404e-06,
+      "loss": 0.6674,
+      "step": 20895
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.755747744451209,
+      "learning_rate": 2.463512895634901e-06,
+      "loss": 0.6508,
+      "step": 20900
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.040387676166835,
+      "learning_rate": 2.460985759574187e-06,
+      "loss": 0.6563,
+      "step": 20905
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.968898382731428,
+      "learning_rate": 2.4584594971858473e-06,
+      "loss": 0.6549,
+      "step": 20910
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 8.786369127215906,
+      "learning_rate": 2.455934109339168e-06,
+      "loss": 0.6701,
+      "step": 20915
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.362652153951624,
+      "learning_rate": 2.453409596903131e-06,
+      "loss": 0.6565,
+      "step": 20920
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.037224859768098,
+      "learning_rate": 2.4508859607464203e-06,
+      "loss": 0.69,
+      "step": 20925
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.183266255986752,
+      "learning_rate": 2.4483632017374167e-06,
+      "loss": 0.6915,
+      "step": 20930
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.745109900608311,
+      "learning_rate": 2.445841320744198e-06,
+      "loss": 0.705,
+      "step": 20935
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 10.187762151918701,
+      "learning_rate": 2.443320318634539e-06,
+      "loss": 0.6788,
+      "step": 20940
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.243981756066541,
+      "learning_rate": 2.4408001962759187e-06,
+      "loss": 0.6579,
+      "step": 20945
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.436368028087012,
+      "learning_rate": 2.4382809545355046e-06,
+      "loss": 0.6857,
+      "step": 20950
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.060669721055994,
+      "learning_rate": 2.4357625942801694e-06,
+      "loss": 0.658,
+      "step": 20955
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.9540665812641596,
+      "learning_rate": 2.4332451163764765e-06,
+      "loss": 0.6839,
+      "step": 20960
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.879592711398314,
+      "learning_rate": 2.4307285216906866e-06,
+      "loss": 0.6852,
+      "step": 20965
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.395266201019412,
+      "learning_rate": 2.4282128110887575e-06,
+      "loss": 0.6396,
+      "step": 20970
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 10.715951603539619,
+      "learning_rate": 2.4256979854363453e-06,
+      "loss": 0.6795,
+      "step": 20975
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.113152703376329,
+      "learning_rate": 2.423184045598796e-06,
+      "loss": 0.6761,
+      "step": 20980
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.185557904992484,
+      "learning_rate": 2.420670992441157e-06,
+      "loss": 0.7158,
+      "step": 20985
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.8309932391804375,
+      "learning_rate": 2.4181588268281656e-06,
+      "loss": 0.6775,
+      "step": 20990
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.352726397807516,
+      "learning_rate": 2.4156475496242563e-06,
+      "loss": 0.6462,
+      "step": 20995
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.074028530793119,
+      "learning_rate": 2.413137161693557e-06,
+      "loss": 0.6394,
+      "step": 21000
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.806118859847987,
+      "learning_rate": 2.410627663899888e-06,
+      "loss": 0.6965,
+      "step": 21005
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.303205623748911,
+      "learning_rate": 2.4081190571067688e-06,
+      "loss": 0.6612,
+      "step": 21010
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.7635101057624665,
+      "learning_rate": 2.4056113421774036e-06,
+      "loss": 0.6534,
+      "step": 21015
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.0749282619508,
+      "learning_rate": 2.4031045199746998e-06,
+      "loss": 0.7072,
+      "step": 21020
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.909661154884081,
+      "learning_rate": 2.4005985913612507e-06,
+      "loss": 0.6823,
+      "step": 21025
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.926500818645366,
+      "learning_rate": 2.3980935571993435e-06,
+      "loss": 0.6779,
+      "step": 21030
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.879990313993776,
+      "learning_rate": 2.3955894183509566e-06,
+      "loss": 0.685,
+      "step": 21035
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.49919898506855,
+      "learning_rate": 2.3930861756777648e-06,
+      "loss": 0.6762,
+      "step": 21040
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.825971877972342,
+      "learning_rate": 2.390583830041128e-06,
+      "loss": 0.6662,
+      "step": 21045
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.836307842150723,
+      "learning_rate": 2.3880823823021056e-06,
+      "loss": 0.7139,
+      "step": 21050
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.8872685890080385,
+      "learning_rate": 2.38558183332144e-06,
+      "loss": 0.6456,
+      "step": 21055
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.874725834505177,
+      "learning_rate": 2.3830821839595684e-06,
+      "loss": 0.7174,
+      "step": 21060
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.197819594155209,
+      "learning_rate": 2.3805834350766165e-06,
+      "loss": 0.6593,
+      "step": 21065
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.793047824829741,
+      "learning_rate": 2.3780855875324047e-06,
+      "loss": 0.6978,
+      "step": 21070
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.829755969981174,
+      "learning_rate": 2.375588642186436e-06,
+      "loss": 0.663,
+      "step": 21075
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.64441800256596,
+      "learning_rate": 2.373092599897911e-06,
+      "loss": 0.6674,
+      "step": 21080
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.047361648257937,
+      "learning_rate": 2.3705974615257133e-06,
+      "loss": 0.7065,
+      "step": 21085
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.95270047732342,
+      "learning_rate": 2.3681032279284173e-06,
+      "loss": 0.7028,
+      "step": 21090
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.204323988308421,
+      "learning_rate": 2.365609899964285e-06,
+      "loss": 0.6651,
+      "step": 21095
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.271678402978469,
+      "learning_rate": 2.3631174784912723e-06,
+      "loss": 0.6666,
+      "step": 21100
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.063459933192673,
+      "learning_rate": 2.3606259643670148e-06,
+      "loss": 0.6482,
+      "step": 21105
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.437559796028627,
+      "learning_rate": 2.3581353584488437e-06,
+      "loss": 0.6586,
+      "step": 21110
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.825669261536088,
+      "learning_rate": 2.355645661593773e-06,
+      "loss": 0.641,
+      "step": 21115
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.667516137702388,
+      "learning_rate": 2.3531568746585036e-06,
+      "loss": 0.6698,
+      "step": 21120
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.988785674417255,
+      "learning_rate": 2.3506689984994265e-06,
+      "loss": 0.6739,
+      "step": 21125
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.727327892790064,
+      "learning_rate": 2.3481820339726142e-06,
+      "loss": 0.6624,
+      "step": 21130
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.8673197563247435,
+      "learning_rate": 2.345695981933833e-06,
+      "loss": 0.6868,
+      "step": 21135
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.490877814256852,
+      "learning_rate": 2.3432108432385274e-06,
+      "loss": 0.6523,
+      "step": 21140
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.706242854874962,
+      "learning_rate": 2.3407266187418354e-06,
+      "loss": 0.6702,
+      "step": 21145
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.630820647418488,
+      "learning_rate": 2.338243309298574e-06,
+      "loss": 0.6594,
+      "step": 21150
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.784604042705804,
+      "learning_rate": 2.3357609157632473e-06,
+      "loss": 0.734,
+      "step": 21155
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.414196654666086,
+      "learning_rate": 2.3332794389900434e-06,
+      "loss": 0.6885,
+      "step": 21160
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 7.1018738978504,
+      "learning_rate": 2.3307988798328384e-06,
+      "loss": 0.7051,
+      "step": 21165
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.906674381033124,
+      "learning_rate": 2.3283192391451885e-06,
+      "loss": 0.6619,
+      "step": 21170
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.469245284173604,
+      "learning_rate": 2.3258405177803386e-06,
+      "loss": 0.6902,
+      "step": 21175
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.014548678716852,
+      "learning_rate": 2.323362716591212e-06,
+      "loss": 0.6439,
+      "step": 21180
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.395091828408313,
+      "learning_rate": 2.320885836430418e-06,
+      "loss": 0.6858,
+      "step": 21185
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.338385671125915,
+      "learning_rate": 2.3184098781502473e-06,
+      "loss": 0.6797,
+      "step": 21190
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.440486546092224,
+      "learning_rate": 2.3159348426026784e-06,
+      "loss": 0.6535,
+      "step": 21195
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.014703336012933,
+      "learning_rate": 2.313460730639364e-06,
+      "loss": 0.6694,
+      "step": 21200
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.0268000255109815,
+      "learning_rate": 2.3109875431116485e-06,
+      "loss": 0.6431,
+      "step": 21205
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.294111410480581,
+      "learning_rate": 2.308515280870551e-06,
+      "loss": 0.6721,
+      "step": 21210
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.596658628899066,
+      "learning_rate": 2.3060439447667755e-06,
+      "loss": 0.64,
+      "step": 21215
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.359281340269409,
+      "learning_rate": 2.3035735356507027e-06,
+      "loss": 0.6634,
+      "step": 21220
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 8.878999044428964,
+      "learning_rate": 2.3011040543724035e-06,
+      "loss": 0.6545,
+      "step": 21225
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.7306168814791745,
+      "learning_rate": 2.2986355017816194e-06,
+      "loss": 0.7132,
+      "step": 21230
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.440527418729273,
+      "learning_rate": 2.2961678787277803e-06,
+      "loss": 0.6431,
+      "step": 21235
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.095974682993984,
+      "learning_rate": 2.2937011860599924e-06,
+      "loss": 0.6931,
+      "step": 21240
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.959130597361517,
+      "learning_rate": 2.2912354246270406e-06,
+      "loss": 0.6827,
+      "step": 21245
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.820875700283996,
+      "learning_rate": 2.2887705952773933e-06,
+      "loss": 0.6873,
+      "step": 21250
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.121696731093606,
+      "learning_rate": 2.286306698859192e-06,
+      "loss": 0.6799,
+      "step": 21255
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 10.373115426094676,
+      "learning_rate": 2.2838437362202653e-06,
+      "loss": 0.7123,
+      "step": 21260
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.701640475740954,
+      "learning_rate": 2.2813817082081135e-06,
+      "loss": 0.6797,
+      "step": 21265
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.262566730988645,
+      "learning_rate": 2.278920615669921e-06,
+      "loss": 0.7041,
+      "step": 21270
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 8.108524486232833,
+      "learning_rate": 2.276460459452546e-06,
+      "loss": 0.6351,
+      "step": 21275
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 9.204313406839686,
+      "learning_rate": 2.2740012404025253e-06,
+      "loss": 0.6517,
+      "step": 21280
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.876521529639379,
+      "learning_rate": 2.2715429593660722e-06,
+      "loss": 0.6841,
+      "step": 21285
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.1636493192147785,
+      "learning_rate": 2.269085617189083e-06,
+      "loss": 0.6706,
+      "step": 21290
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 7.354043706285142,
+      "learning_rate": 2.266629214717122e-06,
+      "loss": 0.6583,
+      "step": 21295
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.656344874179418,
+      "learning_rate": 2.2641737527954405e-06,
+      "loss": 0.6794,
+      "step": 21300
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 15.927842070258235,
+      "learning_rate": 2.261719232268957e-06,
+      "loss": 0.6469,
+      "step": 21305
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.574361597789768,
+      "learning_rate": 2.2592656539822706e-06,
+      "loss": 0.67,
+      "step": 21310
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.458625032162326,
+      "learning_rate": 2.256813018779653e-06,
+      "loss": 0.656,
+      "step": 21315
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 15.504041246695413,
+      "learning_rate": 2.254361327505057e-06,
+      "loss": 0.6734,
+      "step": 21320
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.920432068704808,
+      "learning_rate": 2.251910581002104e-06,
+      "loss": 0.6801,
+      "step": 21325
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.977985796519088,
+      "learning_rate": 2.2494607801140977e-06,
+      "loss": 0.6701,
+      "step": 21330
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.521974127049712,
+      "learning_rate": 2.2470119256840102e-06,
+      "loss": 0.6597,
+      "step": 21335
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.2108291136682245,
+      "learning_rate": 2.2445640185544887e-06,
+      "loss": 0.6738,
+      "step": 21340
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.578370170317782,
+      "learning_rate": 2.2421170595678554e-06,
+      "loss": 0.6956,
+      "step": 21345
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.455209357833909,
+      "learning_rate": 2.2396710495661096e-06,
+      "loss": 0.6872,
+      "step": 21350
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.932387669873154,
+      "learning_rate": 2.2372259893909177e-06,
+      "loss": 0.6931,
+      "step": 21355
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 7.383305059308782,
+      "learning_rate": 2.234781879883625e-06,
+      "loss": 0.6765,
+      "step": 21360
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.588357606015773,
+      "learning_rate": 2.2323387218852472e-06,
+      "loss": 0.6468,
+      "step": 21365
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.9486101441179855,
+      "learning_rate": 2.229896516236472e-06,
+      "loss": 0.7126,
+      "step": 21370
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.831380879476231,
+      "learning_rate": 2.227455263777657e-06,
+      "loss": 0.6925,
+      "step": 21375
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.772545065182918,
+      "learning_rate": 2.22501496534884e-06,
+      "loss": 0.6487,
+      "step": 21380
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.355784462134895,
+      "learning_rate": 2.2225756217897232e-06,
+      "loss": 0.6881,
+      "step": 21385
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 10.180633022588685,
+      "learning_rate": 2.2201372339396797e-06,
+      "loss": 0.7179,
+      "step": 21390
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.906657785907644,
+      "learning_rate": 2.2176998026377615e-06,
+      "loss": 0.6773,
+      "step": 21395
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.973334214591822,
+      "learning_rate": 2.2152633287226836e-06,
+      "loss": 0.6862,
+      "step": 21400
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.857519742476949,
+      "learning_rate": 2.2128278130328346e-06,
+      "loss": 0.6675,
+      "step": 21405
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.331979105826995,
+      "learning_rate": 2.2103932564062706e-06,
+      "loss": 0.6346,
+      "step": 21410
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.850241441793481,
+      "learning_rate": 2.207959659680726e-06,
+      "loss": 0.6459,
+      "step": 21415
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.663222394532015,
+      "learning_rate": 2.2055270236935927e-06,
+      "loss": 0.6659,
+      "step": 21420
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.714394073555861,
+      "learning_rate": 2.2030953492819435e-06,
+      "loss": 0.6453,
+      "step": 21425
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.421109528249489,
+      "learning_rate": 2.2006646372825138e-06,
+      "loss": 0.6648,
+      "step": 21430
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.463981661729482,
+      "learning_rate": 2.198234888531708e-06,
+      "loss": 0.6567,
+      "step": 21435
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.074988279798165,
+      "learning_rate": 2.1958061038656e-06,
+      "loss": 0.6448,
+      "step": 21440
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.73004122724363,
+      "learning_rate": 2.193378284119934e-06,
+      "loss": 0.6768,
+      "step": 21445
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.659328962468312,
+      "learning_rate": 2.190951430130119e-06,
+      "loss": 0.6546,
+      "step": 21450
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.614004198385022,
+      "learning_rate": 2.188525542731236e-06,
+      "loss": 0.6502,
+      "step": 21455
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.497775382681463,
+      "learning_rate": 2.1861006227580276e-06,
+      "loss": 0.7165,
+      "step": 21460
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.542848170679907,
+      "learning_rate": 2.1836766710449077e-06,
+      "loss": 0.6685,
+      "step": 21465
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.649025832208719,
+      "learning_rate": 2.1812536884259537e-06,
+      "loss": 0.6761,
+      "step": 21470
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.801979439674982,
+      "learning_rate": 2.178831675734915e-06,
+      "loss": 0.6833,
+      "step": 21475
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.821275013329016,
+      "learning_rate": 2.1764106338052005e-06,
+      "loss": 0.6756,
+      "step": 21480
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.596268528847123,
+      "learning_rate": 2.1739905634698916e-06,
+      "loss": 0.6768,
+      "step": 21485
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.495715820222872,
+      "learning_rate": 2.171571465561731e-06,
+      "loss": 0.6667,
+      "step": 21490
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.776997921508625,
+      "learning_rate": 2.169153340913127e-06,
+      "loss": 0.6944,
+      "step": 21495
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.506795003368372,
+      "learning_rate": 2.1667361903561534e-06,
+      "loss": 0.6646,
+      "step": 21500
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 7.392404856257032,
+      "learning_rate": 2.1643200147225523e-06,
+      "loss": 0.6782,
+      "step": 21505
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.477503623359234,
+      "learning_rate": 2.1619048148437256e-06,
+      "loss": 0.6601,
+      "step": 21510
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.524025133682762,
+      "learning_rate": 2.1594905915507397e-06,
+      "loss": 0.6839,
+      "step": 21515
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.615708947012337,
+      "learning_rate": 2.1570773456743305e-06,
+      "loss": 0.6494,
+      "step": 21520
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.095074229625856,
+      "learning_rate": 2.1546650780448907e-06,
+      "loss": 0.6696,
+      "step": 21525
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.528698210233045,
+      "learning_rate": 2.15225378949248e-06,
+      "loss": 0.6469,
+      "step": 21530
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 7.061796993105184,
+      "learning_rate": 2.149843480846819e-06,
+      "loss": 0.6828,
+      "step": 21535
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.608756274550869,
+      "learning_rate": 2.1474341529372955e-06,
+      "loss": 0.6519,
+      "step": 21540
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 7.6559465404075455,
+      "learning_rate": 2.1450258065929536e-06,
+      "loss": 0.6905,
+      "step": 21545
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.611780806892584,
+      "learning_rate": 2.1426184426425073e-06,
+      "loss": 0.6165,
+      "step": 21550
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.775707939635718,
+      "learning_rate": 2.1402120619143254e-06,
+      "loss": 0.6955,
+      "step": 21555
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.858396353644519,
+      "learning_rate": 2.137806665236441e-06,
+      "loss": 0.6687,
+      "step": 21560
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.800894336334279,
+      "learning_rate": 2.135402253436548e-06,
+      "loss": 0.6648,
+      "step": 21565
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.6098765956852015,
+      "learning_rate": 2.1329988273420055e-06,
+      "loss": 0.6861,
+      "step": 21570
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 8.6380129196337,
+      "learning_rate": 2.1305963877798265e-06,
+      "loss": 0.6491,
+      "step": 21575
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.510593341834106,
+      "learning_rate": 2.128194935576692e-06,
+      "loss": 0.6633,
+      "step": 21580
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.012673378533721,
+      "learning_rate": 2.1257944715589363e-06,
+      "loss": 0.6853,
+      "step": 21585
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.940600401787411,
+      "learning_rate": 2.1233949965525585e-06,
+      "loss": 0.6819,
+      "step": 21590
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 8.435816299728263,
+      "learning_rate": 2.120996511383213e-06,
+      "loss": 0.6768,
+      "step": 21595
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 7.472662445866399,
+      "learning_rate": 2.1185990168762193e-06,
+      "loss": 0.6378,
+      "step": 21600
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.412920367145333,
+      "learning_rate": 2.1162025138565505e-06,
+      "loss": 0.691,
+      "step": 21605
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.563470399698418,
+      "learning_rate": 2.1138070031488445e-06,
+      "loss": 0.7107,
+      "step": 21610
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.746862880742315,
+      "learning_rate": 2.1114124855773915e-06,
+      "loss": 0.6786,
+      "step": 21615
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.143965721903495,
+      "learning_rate": 2.1090189619661437e-06,
+      "loss": 0.651,
+      "step": 21620
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.909432206635309,
+      "learning_rate": 2.1066264331387084e-06,
+      "loss": 0.6533,
+      "step": 21625
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.953769822474178,
+      "learning_rate": 2.104234899918355e-06,
+      "loss": 0.6591,
+      "step": 21630
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.648977379946307,
+      "learning_rate": 2.101844363128007e-06,
+      "loss": 0.6816,
+      "step": 21635
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 7.201087140446208,
+      "learning_rate": 2.099454823590244e-06,
+      "loss": 0.6705,
+      "step": 21640
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.959974141273814,
+      "learning_rate": 2.0970662821273074e-06,
+      "loss": 0.6469,
+      "step": 21645
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.664286681471639,
+      "learning_rate": 2.094678739561091e-06,
+      "loss": 0.6646,
+      "step": 21650
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.485570321475418,
+      "learning_rate": 2.092292196713145e-06,
+      "loss": 0.6523,
+      "step": 21655
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.368909470789177,
+      "learning_rate": 2.0899066544046754e-06,
+      "loss": 0.6451,
+      "step": 21660
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.97877059094573,
+      "learning_rate": 2.087522113456548e-06,
+      "loss": 0.674,
+      "step": 21665
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.508107160264964,
+      "learning_rate": 2.085138574689278e-06,
+      "loss": 0.6405,
+      "step": 21670
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.801473722359958,
+      "learning_rate": 2.082756038923042e-06,
+      "loss": 0.6738,
+      "step": 21675
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.701163787304521,
+      "learning_rate": 2.080374506977667e-06,
+      "loss": 0.6915,
+      "step": 21680
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.0583630847683345,
+      "learning_rate": 2.0779939796726358e-06,
+      "loss": 0.6789,
+      "step": 21685
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.989863272716518,
+      "learning_rate": 2.075614457827083e-06,
+      "loss": 0.6593,
+      "step": 21690
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.457297521696792,
+      "learning_rate": 2.073235942259804e-06,
+      "loss": 0.6472,
+      "step": 21695
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.769915672849963,
+      "learning_rate": 2.0708584337892406e-06,
+      "loss": 0.6985,
+      "step": 21700
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.0036448831642755,
+      "learning_rate": 2.0684819332334937e-06,
+      "loss": 0.6477,
+      "step": 21705
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.353427945270811,
+      "learning_rate": 2.066106441410314e-06,
+      "loss": 0.716,
+      "step": 21710
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.403640101289655,
+      "learning_rate": 2.0637319591371057e-06,
+      "loss": 0.6456,
+      "step": 21715
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.799200550599031,
+      "learning_rate": 2.0613584872309238e-06,
+      "loss": 0.6445,
+      "step": 21720
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 10.13065659228395,
+      "learning_rate": 2.058986026508482e-06,
+      "loss": 0.6908,
+      "step": 21725
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.577943061340765,
+      "learning_rate": 2.0566145777861374e-06,
+      "loss": 0.6751,
+      "step": 21730
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.263454894906639,
+      "learning_rate": 2.054244141879907e-06,
+      "loss": 0.6474,
+      "step": 21735
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.782587443226552,
+      "learning_rate": 2.0518747196054533e-06,
+      "loss": 0.633,
+      "step": 21740
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.484463749033461,
+      "learning_rate": 2.049506311778094e-06,
+      "loss": 0.6482,
+      "step": 21745
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.316010453674646,
+      "learning_rate": 2.047138919212792e-06,
+      "loss": 0.6819,
+      "step": 21750
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.308040713505686,
+      "learning_rate": 2.04477254272417e-06,
+      "loss": 0.6636,
+      "step": 21755
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.735685322131239,
+      "learning_rate": 2.0424071831264913e-06,
+      "loss": 0.6882,
+      "step": 21760
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 8.13529321248825,
+      "learning_rate": 2.0400428412336776e-06,
+      "loss": 0.6645,
+      "step": 21765
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 9.906802535189406,
+      "learning_rate": 2.0376795178592973e-06,
+      "loss": 0.66,
+      "step": 21770
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.087737709787373,
+      "learning_rate": 2.035317213816562e-06,
+      "loss": 0.6657,
+      "step": 21775
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.619965083742663,
+      "learning_rate": 2.0329559299183438e-06,
+      "loss": 0.6584,
+      "step": 21780
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.693608908068327,
+      "learning_rate": 2.0305956669771544e-06,
+      "loss": 0.692,
+      "step": 21785
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.339632482464978,
+      "learning_rate": 2.0282364258051614e-06,
+      "loss": 0.6453,
+      "step": 21790
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.938669128359225,
+      "learning_rate": 2.025878207214174e-06,
+      "loss": 0.6991,
+      "step": 21795
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.791176392690036,
+      "learning_rate": 2.023521012015659e-06,
+      "loss": 0.6592,
+      "step": 21800
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.366344972490452,
+      "learning_rate": 2.021164841020717e-06,
+      "loss": 0.6674,
+      "step": 21805
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.120904543191494,
+      "learning_rate": 2.0188096950401097e-06,
+      "loss": 0.6761,
+      "step": 21810
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 8.949775455094654,
+      "learning_rate": 2.016455574884238e-06,
+      "loss": 0.668,
+      "step": 21815
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.4831923815287515,
+      "learning_rate": 2.0141024813631543e-06,
+      "loss": 0.6894,
+      "step": 21820
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 8.6838611621769,
+      "learning_rate": 2.0117504152865535e-06,
+      "loss": 0.677,
+      "step": 21825
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 11.12976987528081,
+      "learning_rate": 2.0093993774637844e-06,
+      "loss": 0.6721,
+      "step": 21830
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.923204024712102,
+      "learning_rate": 2.007049368703829e-06,
+      "loss": 0.6547,
+      "step": 21835
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.444648074512381,
+      "learning_rate": 2.0047003898153294e-06,
+      "loss": 0.7173,
+      "step": 21840
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.512895550568181,
+      "learning_rate": 2.002352441606563e-06,
+      "loss": 0.6774,
+      "step": 21845
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.120862716747655,
+      "learning_rate": 2.00000552488546e-06,
+      "loss": 0.6994,
+      "step": 21850
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.1405050814060465,
+      "learning_rate": 1.9976596404595896e-06,
+      "loss": 0.6761,
+      "step": 21855
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.729094027264364,
+      "learning_rate": 1.995314789136172e-06,
+      "loss": 0.6588,
+      "step": 21860
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.843415886395163,
+      "learning_rate": 1.9929709717220618e-06,
+      "loss": 0.6514,
+      "step": 21865
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.202932791193621,
+      "learning_rate": 1.9906281890237713e-06,
+      "loss": 0.6734,
+      "step": 21870
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.752555156991824,
+      "learning_rate": 1.9882864418474433e-06,
+      "loss": 0.6801,
+      "step": 21875
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.546389634068141,
+      "learning_rate": 1.985945730998877e-06,
+      "loss": 0.712,
+      "step": 21880
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.450342124329078,
+      "learning_rate": 1.9836060572835043e-06,
+      "loss": 0.6513,
+      "step": 21885
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.011460680897503,
+      "learning_rate": 1.981267421506409e-06,
+      "loss": 0.7039,
+      "step": 21890
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.361626448788882,
+      "learning_rate": 1.9789298244723094e-06,
+      "loss": 0.647,
+      "step": 21895
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.543949883889084,
+      "learning_rate": 1.9765932669855696e-06,
+      "loss": 0.7214,
+      "step": 21900
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.629136822219546,
+      "learning_rate": 1.9742577498502015e-06,
+      "loss": 0.6905,
+      "step": 21905
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 8.094072104483322,
+      "learning_rate": 1.9719232738698496e-06,
+      "loss": 0.7063,
+      "step": 21910
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 6.180828173160776,
+      "learning_rate": 1.9695898398478087e-06,
+      "loss": 0.6723,
+      "step": 21915
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.900414331532556,
+      "learning_rate": 1.9672574485870095e-06,
+      "loss": 0.6339,
+      "step": 21920
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.887237517635425,
+      "learning_rate": 1.9649261008900256e-06,
+      "loss": 0.6656,
+      "step": 21925
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.3677286120369345,
+      "learning_rate": 1.9625957975590697e-06,
+      "loss": 0.6835,
+      "step": 21930
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.997891580541595,
+      "learning_rate": 1.9602665393960006e-06,
+      "loss": 0.6776,
+      "step": 21935
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.034257581641024,
+      "learning_rate": 1.9579383272023106e-06,
+      "loss": 0.6598,
+      "step": 21940
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.095295571247911,
+      "learning_rate": 1.9556111617791383e-06,
+      "loss": 0.6423,
+      "step": 21945
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.779739155406581,
+      "learning_rate": 1.9532850439272576e-06,
+      "loss": 0.659,
+      "step": 21950
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 17.447892921304916,
+      "learning_rate": 1.950959974447083e-06,
+      "loss": 0.6411,
+      "step": 21955
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.38037099062299,
+      "learning_rate": 1.948635954138668e-06,
+      "loss": 0.6598,
+      "step": 21960
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 6.985338264358882,
+      "learning_rate": 1.946312983801708e-06,
+      "loss": 0.6597,
+      "step": 21965
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.06480244294857,
+      "learning_rate": 1.943991064235532e-06,
+      "loss": 0.6527,
+      "step": 21970
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.539366017792932,
+      "learning_rate": 1.941670196239113e-06,
+      "loss": 0.6831,
+      "step": 21975
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.784330135145143,
+      "learning_rate": 1.939350380611058e-06,
+      "loss": 0.6861,
+      "step": 21980
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.570023660771087,
+      "learning_rate": 1.937031618149616e-06,
+      "loss": 0.6576,
+      "step": 21985
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.7423422973082845,
+      "learning_rate": 1.9347139096526662e-06,
+      "loss": 0.6765,
+      "step": 21990
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.434972787435978,
+      "learning_rate": 1.932397255917734e-06,
+      "loss": 0.6976,
+      "step": 21995
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.393669477772615,
+      "learning_rate": 1.930081657741974e-06,
+      "loss": 0.6386,
+      "step": 22000
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.324763834075317,
+      "learning_rate": 1.9277671159221858e-06,
+      "loss": 0.6698,
+      "step": 22005
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.172778216238288,
+      "learning_rate": 1.925453631254796e-06,
+      "loss": 0.668,
+      "step": 22010
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 10.775476470719338,
+      "learning_rate": 1.9231412045358794e-06,
+      "loss": 0.7001,
+      "step": 22015
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.772612247891434,
+      "learning_rate": 1.920829836561134e-06,
+      "loss": 0.6343,
+      "step": 22020
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.638808772116786,
+      "learning_rate": 1.9185195281258984e-06,
+      "loss": 0.6662,
+      "step": 22025
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.477423201454245,
+      "learning_rate": 1.9162102800251526e-06,
+      "loss": 0.691,
+      "step": 22030
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.001265825880683,
+      "learning_rate": 1.913902093053502e-06,
+      "loss": 0.7369,
+      "step": 22035
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.695603228654765,
+      "learning_rate": 1.911594968005195e-06,
+      "loss": 0.6901,
+      "step": 22040
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.699503683926651,
+      "learning_rate": 1.90928890567411e-06,
+      "loss": 0.6603,
+      "step": 22045
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.246464717025585,
+      "learning_rate": 1.9069839068537605e-06,
+      "loss": 0.6683,
+      "step": 22050
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.441835122688492,
+      "learning_rate": 1.9046799723372927e-06,
+      "loss": 0.6552,
+      "step": 22055
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.807563202328714,
+      "learning_rate": 1.902377102917492e-06,
+      "loss": 0.6635,
+      "step": 22060
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.53254206934949,
+      "learning_rate": 1.90007529938677e-06,
+      "loss": 0.6673,
+      "step": 22065
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.36613438278603,
+      "learning_rate": 1.897774562537178e-06,
+      "loss": 0.6644,
+      "step": 22070
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 8.838309800932239,
+      "learning_rate": 1.895474893160396e-06,
+      "loss": 0.7222,
+      "step": 22075
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 7.298006664524484,
+      "learning_rate": 1.8931762920477387e-06,
+      "loss": 0.6583,
+      "step": 22080
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.656911860726481,
+      "learning_rate": 1.89087875999015e-06,
+      "loss": 0.6445,
+      "step": 22085
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.649023226724349,
+      "learning_rate": 1.888582297778212e-06,
+      "loss": 0.6448,
+      "step": 22090
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.356781900163968,
+      "learning_rate": 1.8862869062021317e-06,
+      "loss": 0.6458,
+      "step": 22095
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.5443744334239895,
+      "learning_rate": 1.8839925860517549e-06,
+      "loss": 0.6573,
+      "step": 22100
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.735248749146872,
+      "learning_rate": 1.8816993381165533e-06,
+      "loss": 0.6369,
+      "step": 22105
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.68724647456985,
+      "learning_rate": 1.8794071631856315e-06,
+      "loss": 0.6606,
+      "step": 22110
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.519650423597918,
+      "learning_rate": 1.8771160620477219e-06,
+      "loss": 0.6397,
+      "step": 22115
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 7.213776592218962,
+      "learning_rate": 1.8748260354911945e-06,
+      "loss": 0.6731,
+      "step": 22120
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 8.309313240765013,
+      "learning_rate": 1.872537084304042e-06,
+      "loss": 0.6633,
+      "step": 22125
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.250550954950561,
+      "learning_rate": 1.8702492092738934e-06,
+      "loss": 0.6962,
+      "step": 22130
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.756675355745477,
+      "learning_rate": 1.8679624111880024e-06,
+      "loss": 0.6402,
+      "step": 22135
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.987116484431764,
+      "learning_rate": 1.8656766908332542e-06,
+      "loss": 0.669,
+      "step": 22140
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.662568308039815,
+      "learning_rate": 1.8633920489961615e-06,
+      "loss": 0.6795,
+      "step": 22145
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.9315334222705465,
+      "learning_rate": 1.8611084864628708e-06,
+      "loss": 0.63,
+      "step": 22150
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 7.919007732778563,
+      "learning_rate": 1.8588260040191518e-06,
+      "loss": 0.6346,
+      "step": 22155
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.209612486019074,
+      "learning_rate": 1.856544602450403e-06,
+      "loss": 0.6401,
+      "step": 22160
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.573632350427878,
+      "learning_rate": 1.8542642825416558e-06,
+      "loss": 0.6574,
+      "step": 22165
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.75801656606795,
+      "learning_rate": 1.8519850450775646e-06,
+      "loss": 0.681,
+      "step": 22170
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.473139584757566,
+      "learning_rate": 1.849706890842412e-06,
+      "loss": 0.64,
+      "step": 22175
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.356833240856356,
+      "learning_rate": 1.8474298206201086e-06,
+      "loss": 0.6876,
+      "step": 22180
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.005661901628179,
+      "learning_rate": 1.8451538351941938e-06,
+      "loss": 0.6857,
+      "step": 22185
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.55621385776115,
+      "learning_rate": 1.84287893534783e-06,
+      "loss": 0.6653,
+      "step": 22190
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.004544719422335,
+      "learning_rate": 1.8406051218638104e-06,
+      "loss": 0.6388,
+      "step": 22195
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.650527884490754,
+      "learning_rate": 1.8383323955245513e-06,
+      "loss": 0.666,
+      "step": 22200
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.510569408301339,
+      "learning_rate": 1.8360607571120948e-06,
+      "loss": 0.6684,
+      "step": 22205
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.61974386639768,
+      "learning_rate": 1.8337902074081082e-06,
+      "loss": 0.669,
+      "step": 22210
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.199501659759417,
+      "learning_rate": 1.831520747193889e-06,
+      "loss": 0.6624,
+      "step": 22215
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.019792770858063,
+      "learning_rate": 1.8292523772503524e-06,
+      "loss": 0.6581,
+      "step": 22220
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 8.958303016321787,
+      "learning_rate": 1.826985098358046e-06,
+      "loss": 0.6779,
+      "step": 22225
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.503162405669565,
+      "learning_rate": 1.8247189112971374e-06,
+      "loss": 0.6477,
+      "step": 22230
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.53591139782825,
+      "learning_rate": 1.8224538168474182e-06,
+      "loss": 0.686,
+      "step": 22235
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.93514558069354,
+      "learning_rate": 1.820189815788304e-06,
+      "loss": 0.6986,
+      "step": 22240
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.085521784062833,
+      "learning_rate": 1.8179269088988387e-06,
+      "loss": 0.6652,
+      "step": 22245
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.759343061989954,
+      "learning_rate": 1.8156650969576832e-06,
+      "loss": 0.6996,
+      "step": 22250
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.081130624986382,
+      "learning_rate": 1.8134043807431283e-06,
+      "loss": 0.6706,
+      "step": 22255
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.646210752535273,
+      "learning_rate": 1.811144761033083e-06,
+      "loss": 0.6515,
+      "step": 22260
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.462019258658007,
+      "learning_rate": 1.8088862386050786e-06,
+      "loss": 0.6332,
+      "step": 22265
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.5455720420032675,
+      "learning_rate": 1.8066288142362704e-06,
+      "loss": 0.6632,
+      "step": 22270
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.0054769315556555,
+      "learning_rate": 1.8043724887034392e-06,
+      "loss": 0.6543,
+      "step": 22275
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.001193032887485,
+      "learning_rate": 1.802117262782982e-06,
+      "loss": 0.6798,
+      "step": 22280
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.377042635357598,
+      "learning_rate": 1.799863137250919e-06,
+      "loss": 0.628,
+      "step": 22285
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.247020646583163,
+      "learning_rate": 1.7976101128828955e-06,
+      "loss": 0.6841,
+      "step": 22290
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.228469202329293,
+      "learning_rate": 1.7953581904541733e-06,
+      "loss": 0.6506,
+      "step": 22295
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.332671657525808,
+      "learning_rate": 1.7931073707396373e-06,
+      "loss": 0.6919,
+      "step": 22300
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 8.67677861390262,
+      "learning_rate": 1.7908576545137907e-06,
+      "loss": 0.6763,
+      "step": 22305
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.955850137989377,
+      "learning_rate": 1.7886090425507612e-06,
+      "loss": 0.6497,
+      "step": 22310
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.066982632939822,
+      "learning_rate": 1.7863615356242913e-06,
+      "loss": 0.6893,
+      "step": 22315
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.719905492192519,
+      "learning_rate": 1.7841151345077495e-06,
+      "loss": 0.6588,
+      "step": 22320
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.610428719794262,
+      "learning_rate": 1.7818698399741186e-06,
+      "loss": 0.6422,
+      "step": 22325
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 8.213106162955256,
+      "learning_rate": 1.7796256527960021e-06,
+      "loss": 0.6617,
+      "step": 22330
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.229355768945967,
+      "learning_rate": 1.7773825737456208e-06,
+      "loss": 0.6505,
+      "step": 22335
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.012731623542265,
+      "learning_rate": 1.7751406035948193e-06,
+      "loss": 0.6619,
+      "step": 22340
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.413531472969937,
+      "learning_rate": 1.7728997431150546e-06,
+      "loss": 0.6601,
+      "step": 22345
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.726110813450331,
+      "learning_rate": 1.7706599930774077e-06,
+      "loss": 0.6832,
+      "step": 22350
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.9672144231735045,
+      "learning_rate": 1.768421354252573e-06,
+      "loss": 0.654,
+      "step": 22355
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.407604286978984,
+      "learning_rate": 1.7661838274108633e-06,
+      "loss": 0.7097,
+      "step": 22360
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.588861929961504,
+      "learning_rate": 1.7639474133222085e-06,
+      "loss": 0.6319,
+      "step": 22365
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.068515096985442,
+      "learning_rate": 1.7617121127561598e-06,
+      "loss": 0.6928,
+      "step": 22370
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.565497603312442,
+      "learning_rate": 1.7594779264818783e-06,
+      "loss": 0.6595,
+      "step": 22375
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.856303691873805,
+      "learning_rate": 1.7572448552681493e-06,
+      "loss": 0.6576,
+      "step": 22380
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.032417280865937,
+      "learning_rate": 1.7550128998833681e-06,
+      "loss": 0.6665,
+      "step": 22385
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 11.027154143621729,
+      "learning_rate": 1.7527820610955487e-06,
+      "loss": 0.6895,
+      "step": 22390
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.220039667640416,
+      "learning_rate": 1.7505523396723189e-06,
+      "loss": 0.6352,
+      "step": 22395
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 9.297318647388199,
+      "learning_rate": 1.7483237363809274e-06,
+      "loss": 0.6873,
+      "step": 22400
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.900496581353949,
+      "learning_rate": 1.7460962519882324e-06,
+      "loss": 0.6739,
+      "step": 22405
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.656133617830663,
+      "learning_rate": 1.7438698872607073e-06,
+      "loss": 0.6538,
+      "step": 22410
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.577365486763903,
+      "learning_rate": 1.7416446429644462e-06,
+      "loss": 0.6357,
+      "step": 22415
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.4964458786704755,
+      "learning_rate": 1.739420519865151e-06,
+      "loss": 0.6212,
+      "step": 22420
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.718619521205935,
+      "learning_rate": 1.7371975187281408e-06,
+      "loss": 0.6586,
+      "step": 22425
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.828469851925843,
+      "learning_rate": 1.7349756403183466e-06,
+      "loss": 0.6864,
+      "step": 22430
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.822043565193145,
+      "learning_rate": 1.7327548854003174e-06,
+      "loss": 0.6553,
+      "step": 22435
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.742210440858266,
+      "learning_rate": 1.7305352547382104e-06,
+      "loss": 0.6594,
+      "step": 22440
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.268040093322078,
+      "learning_rate": 1.7283167490958008e-06,
+      "loss": 0.704,
+      "step": 22445
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.3592022150205665,
+      "learning_rate": 1.726099369236473e-06,
+      "loss": 0.655,
+      "step": 22450
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.995899113494076,
+      "learning_rate": 1.7238831159232257e-06,
+      "loss": 0.6753,
+      "step": 22455
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.883862383907867,
+      "learning_rate": 1.7216679899186672e-06,
+      "loss": 0.6574,
+      "step": 22460
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.639544085369703,
+      "learning_rate": 1.719453991985024e-06,
+      "loss": 0.6652,
+      "step": 22465
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.225580088013155,
+      "learning_rate": 1.7172411228841268e-06,
+      "loss": 0.6618,
+      "step": 22470
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.1915848336128985,
+      "learning_rate": 1.7150293833774257e-06,
+      "loss": 0.6785,
+      "step": 22475
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 7.4886162606891284,
+      "learning_rate": 1.7128187742259766e-06,
+      "loss": 0.6523,
+      "step": 22480
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.071153259933611,
+      "learning_rate": 1.7106092961904476e-06,
+      "loss": 0.6642,
+      "step": 22485
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.821964514546382,
+      "learning_rate": 1.708400950031116e-06,
+      "loss": 0.6571,
+      "step": 22490
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.567740592588277,
+      "learning_rate": 1.706193736507875e-06,
+      "loss": 0.6624,
+      "step": 22495
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.930879935653045,
+      "learning_rate": 1.7039876563802227e-06,
+      "loss": 0.6651,
+      "step": 22500
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.0767573323260935,
+      "learning_rate": 1.701782710407271e-06,
+      "loss": 0.7218,
+      "step": 22505
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.6670914829207115,
+      "learning_rate": 1.699578899347738e-06,
+      "loss": 0.6669,
+      "step": 22510
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.849989536360458,
+      "learning_rate": 1.697376223959954e-06,
+      "loss": 0.6287,
+      "step": 22515
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 8.789381725465098,
+      "learning_rate": 1.6951746850018553e-06,
+      "loss": 0.6648,
+      "step": 22520
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 7.480342356637824,
+      "learning_rate": 1.6929742832309926e-06,
+      "loss": 0.6871,
+      "step": 22525
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 8.677024202708074,
+      "learning_rate": 1.6907750194045187e-06,
+      "loss": 0.672,
+      "step": 22530
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.2619899057185044,
+      "learning_rate": 1.6885768942792018e-06,
+      "loss": 0.6492,
+      "step": 22535
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.64342507169857,
+      "learning_rate": 1.6863799086114123e-06,
+      "loss": 0.6228,
+      "step": 22540
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.816052603229862,
+      "learning_rate": 1.684184063157132e-06,
+      "loss": 0.6745,
+      "step": 22545
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.13577322884629,
+      "learning_rate": 1.6819893586719482e-06,
+      "loss": 0.6422,
+      "step": 22550
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.261141020345327,
+      "learning_rate": 1.6797957959110556e-06,
+      "loss": 0.6444,
+      "step": 22555
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.338569930570511,
+      "learning_rate": 1.6776033756292602e-06,
+      "loss": 0.6916,
+      "step": 22560
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.419978426971964,
+      "learning_rate": 1.675412098580968e-06,
+      "loss": 0.6775,
+      "step": 22565
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.16182641300474,
+      "learning_rate": 1.6732219655201992e-06,
+      "loss": 0.6604,
+      "step": 22570
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.597935147747176,
+      "learning_rate": 1.6710329772005745e-06,
+      "loss": 0.6568,
+      "step": 22575
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.515372501062043,
+      "learning_rate": 1.668845134375323e-06,
+      "loss": 0.6796,
+      "step": 22580
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.9916103673781,
+      "learning_rate": 1.6666584377972778e-06,
+      "loss": 0.6468,
+      "step": 22585
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.799949235202191,
+      "learning_rate": 1.6644728882188816e-06,
+      "loss": 0.6541,
+      "step": 22590
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 7.06369827344823,
+      "learning_rate": 1.6622884863921774e-06,
+      "loss": 0.645,
+      "step": 22595
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.824620430723345,
+      "learning_rate": 1.6601052330688183e-06,
+      "loss": 0.6786,
+      "step": 22600
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.431448980072174,
+      "learning_rate": 1.657923129000059e-06,
+      "loss": 0.6947,
+      "step": 22605
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.597594890904755,
+      "learning_rate": 1.6557421749367587e-06,
+      "loss": 0.7003,
+      "step": 22610
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.8169484999390715,
+      "learning_rate": 1.6535623716293803e-06,
+      "loss": 0.6554,
+      "step": 22615
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.649438638850983,
+      "learning_rate": 1.6513837198279954e-06,
+      "loss": 0.6925,
+      "step": 22620
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.363151737668417,
+      "learning_rate": 1.6492062202822724e-06,
+      "loss": 0.675,
+      "step": 22625
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.638961681603949,
+      "learning_rate": 1.6470298737414902e-06,
+      "loss": 0.6805,
+      "step": 22630
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.301544447481115,
+      "learning_rate": 1.6448546809545263e-06,
+      "loss": 0.6685,
+      "step": 22635
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.82592619935771,
+      "learning_rate": 1.6426806426698616e-06,
+      "loss": 0.6311,
+      "step": 22640
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.879463534993098,
+      "learning_rate": 1.6405077596355802e-06,
+      "loss": 0.676,
+      "step": 22645
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.285753076248275,
+      "learning_rate": 1.6383360325993714e-06,
+      "loss": 0.6604,
+      "step": 22650
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.806887277989003,
+      "learning_rate": 1.636165462308521e-06,
+      "loss": 0.7141,
+      "step": 22655
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.653980698418252,
+      "learning_rate": 1.6339960495099243e-06,
+      "loss": 0.6709,
+      "step": 22660
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.908926619764527,
+      "learning_rate": 1.6318277949500722e-06,
+      "loss": 0.6616,
+      "step": 22665
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.916726063911858,
+      "learning_rate": 1.6296606993750585e-06,
+      "loss": 0.6744,
+      "step": 22670
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.085829241125342,
+      "learning_rate": 1.6274947635305787e-06,
+      "loss": 0.6666,
+      "step": 22675
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.298049943306973,
+      "learning_rate": 1.6253299881619288e-06,
+      "loss": 0.6646,
+      "step": 22680
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.5751950869327045,
+      "learning_rate": 1.623166374014008e-06,
+      "loss": 0.6832,
+      "step": 22685
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.355871168560011,
+      "learning_rate": 1.6210039218313111e-06,
+      "loss": 0.6712,
+      "step": 22690
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.288403929339322,
+      "learning_rate": 1.6188426323579393e-06,
+      "loss": 0.6812,
+      "step": 22695
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.5701276572794125,
+      "learning_rate": 1.616682506337588e-06,
+      "loss": 0.6712,
+      "step": 22700
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.251012685634926,
+      "learning_rate": 1.6145235445135548e-06,
+      "loss": 0.7031,
+      "step": 22705
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.5135520460923795,
+      "learning_rate": 1.6123657476287351e-06,
+      "loss": 0.659,
+      "step": 22710
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.66950416575332,
+      "learning_rate": 1.6102091164256274e-06,
+      "loss": 0.6622,
+      "step": 22715
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.036823131575185,
+      "learning_rate": 1.6080536516463236e-06,
+      "loss": 0.6851,
+      "step": 22720
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.4755189801415325,
+      "learning_rate": 1.60589935403252e-06,
+      "loss": 0.6726,
+      "step": 22725
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.375815713966879,
+      "learning_rate": 1.6037462243255075e-06,
+      "loss": 0.6622,
+      "step": 22730
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.399869288391677,
+      "learning_rate": 1.6015942632661751e-06,
+      "loss": 0.6638,
+      "step": 22735
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.892863779789559,
+      "learning_rate": 1.5994434715950092e-06,
+      "loss": 0.6664,
+      "step": 22740
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 10.217792975059123,
+      "learning_rate": 1.5972938500520985e-06,
+      "loss": 0.6393,
+      "step": 22745
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.600183963459423,
+      "learning_rate": 1.5951453993771226e-06,
+      "loss": 0.6497,
+      "step": 22750
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.397161807213989,
+      "learning_rate": 1.5929981203093642e-06,
+      "loss": 0.6868,
+      "step": 22755
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.338230143136408,
+      "learning_rate": 1.5908520135876981e-06,
+      "loss": 0.631,
+      "step": 22760
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 9.48096451098916,
+      "learning_rate": 1.588707079950599e-06,
+      "loss": 0.6505,
+      "step": 22765
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.999293793939398,
+      "learning_rate": 1.586563320136133e-06,
+      "loss": 0.637,
+      "step": 22770
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.274154137351849,
+      "learning_rate": 1.5844207348819702e-06,
+      "loss": 0.6752,
+      "step": 22775
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 6.788575808926672,
+      "learning_rate": 1.5822793249253687e-06,
+      "loss": 0.6621,
+      "step": 22780
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.857226848457538,
+      "learning_rate": 1.5801390910031888e-06,
+      "loss": 0.6832,
+      "step": 22785
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.658009144634473,
+      "learning_rate": 1.5780000338518813e-06,
+      "loss": 0.6528,
+      "step": 22790
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 16.1108955367705,
+      "learning_rate": 1.5758621542074942e-06,
+      "loss": 0.6745,
+      "step": 22795
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 13.716842477092792,
+      "learning_rate": 1.573725452805669e-06,
+      "loss": 0.6798,
+      "step": 22800
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 12.84118164022468,
+      "learning_rate": 1.5715899303816413e-06,
+      "loss": 0.6628,
+      "step": 22805
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 13.752749266021626,
+      "learning_rate": 1.569455587670246e-06,
+      "loss": 0.689,
+      "step": 22810
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.108777183409205,
+      "learning_rate": 1.5673224254059045e-06,
+      "loss": 0.6336,
+      "step": 22815
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.070827902531056,
+      "learning_rate": 1.5651904443226396e-06,
+      "loss": 0.6531,
+      "step": 22820
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 6.386136834259939,
+      "learning_rate": 1.563059645154062e-06,
+      "loss": 0.6565,
+      "step": 22825
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.73848076802254,
+      "learning_rate": 1.5609300286333785e-06,
+      "loss": 0.637,
+      "step": 22830
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.383849619388004,
+      "learning_rate": 1.558801595493385e-06,
+      "loss": 0.6544,
+      "step": 22835
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.528671138607547,
+      "learning_rate": 1.556674346466478e-06,
+      "loss": 0.6391,
+      "step": 22840
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.863716125361647,
+      "learning_rate": 1.5545482822846375e-06,
+      "loss": 0.6687,
+      "step": 22845
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.416901948032352,
+      "learning_rate": 1.5524234036794443e-06,
+      "loss": 0.6686,
+      "step": 22850
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 12.080893725057612,
+      "learning_rate": 1.5502997113820651e-06,
+      "loss": 0.6806,
+      "step": 22855
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 12.492772940957037,
+      "learning_rate": 1.5481772061232602e-06,
+      "loss": 0.6668,
+      "step": 22860
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.6746567293411605,
+      "learning_rate": 1.5460558886333799e-06,
+      "loss": 0.6356,
+      "step": 22865
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.517073475954265,
+      "learning_rate": 1.5439357596423715e-06,
+      "loss": 0.6996,
+      "step": 22870
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.808254989184655,
+      "learning_rate": 1.5418168198797656e-06,
+      "loss": 0.664,
+      "step": 22875
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.489570597294558,
+      "learning_rate": 1.5396990700746906e-06,
+      "loss": 0.6744,
+      "step": 22880
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.180070947431376,
+      "learning_rate": 1.5375825109558606e-06,
+      "loss": 0.6738,
+      "step": 22885
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.799182335651492,
+      "learning_rate": 1.5354671432515816e-06,
+      "loss": 0.6175,
+      "step": 22890
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.500868500253037,
+      "learning_rate": 1.533352967689748e-06,
+      "loss": 0.6295,
+      "step": 22895
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.943229695166293,
+      "learning_rate": 1.5312399849978483e-06,
+      "loss": 0.6333,
+      "step": 22900
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 7.070154327116882,
+      "learning_rate": 1.5291281959029558e-06,
+      "loss": 0.6425,
+      "step": 22905
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.857751537719575,
+      "learning_rate": 1.5270176011317373e-06,
+      "loss": 0.6621,
+      "step": 22910
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.647239303953281,
+      "learning_rate": 1.5249082014104455e-06,
+      "loss": 0.646,
+      "step": 22915
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.693460297092233,
+      "learning_rate": 1.5227999974649221e-06,
+      "loss": 0.6826,
+      "step": 22920
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.809710792652713,
+      "learning_rate": 1.5206929900205987e-06,
+      "loss": 0.637,
+      "step": 22925
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.580585052715756,
+      "learning_rate": 1.5185871798024925e-06,
+      "loss": 0.6762,
+      "step": 22930
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.500498281272657,
+      "learning_rate": 1.516482567535214e-06,
+      "loss": 0.6469,
+      "step": 22935
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.953886299364294,
+      "learning_rate": 1.514379153942956e-06,
+      "loss": 0.6251,
+      "step": 22940
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.8641256553373,
+      "learning_rate": 1.5122769397495047e-06,
+      "loss": 0.6595,
+      "step": 22945
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.563681811327853,
+      "learning_rate": 1.510175925678224e-06,
+      "loss": 0.6188,
+      "step": 22950
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.174693521988725,
+      "learning_rate": 1.508076112452076e-06,
+      "loss": 0.6701,
+      "step": 22955
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.501701102219561,
+      "learning_rate": 1.5059775007936006e-06,
+      "loss": 0.682,
+      "step": 22960
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.772847198348246,
+      "learning_rate": 1.5038800914249319e-06,
+      "loss": 0.6474,
+      "step": 22965
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.9431492869509865,
+      "learning_rate": 1.501783885067783e-06,
+      "loss": 0.6448,
+      "step": 22970
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.472881197234141,
+      "learning_rate": 1.4996888824434613e-06,
+      "loss": 0.6552,
+      "step": 22975
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.316621141713757,
+      "learning_rate": 1.4975950842728482e-06,
+      "loss": 0.6278,
+      "step": 22980
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.813533380241714,
+      "learning_rate": 1.4955024912764239e-06,
+      "loss": 0.6656,
+      "step": 22985
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.427506467377018,
+      "learning_rate": 1.4934111041742427e-06,
+      "loss": 0.629,
+      "step": 22990
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.177308682078762,
+      "learning_rate": 1.4913209236859533e-06,
+      "loss": 0.6516,
+      "step": 22995
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.602772286736399,
+      "learning_rate": 1.4892319505307817e-06,
+      "loss": 0.6848,
+      "step": 23000
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.021566911989718,
+      "learning_rate": 1.4871441854275454e-06,
+      "loss": 0.6633,
+      "step": 23005
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.1205395441480155,
+      "learning_rate": 1.4850576290946368e-06,
+      "loss": 0.6336,
+      "step": 23010
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.050517972420797,
+      "learning_rate": 1.4829722822500425e-06,
+      "loss": 0.6554,
+      "step": 23015
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.489818487186654,
+      "learning_rate": 1.480888145611325e-06,
+      "loss": 0.6831,
+      "step": 23020
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.368923565018206,
+      "learning_rate": 1.478805219895637e-06,
+      "loss": 0.6764,
+      "step": 23025
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.585422569645293,
+      "learning_rate": 1.4767235058197083e-06,
+      "loss": 0.6559,
+      "step": 23030
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 7.324264673646722,
+      "learning_rate": 1.474643004099857e-06,
+      "loss": 0.6808,
+      "step": 23035
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.070237790985046,
+      "learning_rate": 1.4725637154519817e-06,
+      "loss": 0.6641,
+      "step": 23040
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.047696957897613,
+      "learning_rate": 1.4704856405915629e-06,
+      "loss": 0.6386,
+      "step": 23045
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.5509977864925,
+      "learning_rate": 1.4684087802336639e-06,
+      "loss": 0.6647,
+      "step": 23050
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.120281916761617,
+      "learning_rate": 1.466333135092929e-06,
+      "loss": 0.6543,
+      "step": 23055
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.583061331252208,
+      "learning_rate": 1.4642587058835883e-06,
+      "loss": 0.6635,
+      "step": 23060
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 7.757596000491188,
+      "learning_rate": 1.4621854933194491e-06,
+      "loss": 0.6824,
+      "step": 23065
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.383801982904171,
+      "learning_rate": 1.4601134981139054e-06,
+      "loss": 0.6578,
+      "step": 23070
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.3470348884541306,
+      "learning_rate": 1.4580427209799231e-06,
+      "loss": 0.6928,
+      "step": 23075
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.149969443706589,
+      "learning_rate": 1.4559731626300589e-06,
+      "loss": 0.6617,
+      "step": 23080
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.284858096905827,
+      "learning_rate": 1.4539048237764431e-06,
+      "loss": 0.6275,
+      "step": 23085
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.693178152027507,
+      "learning_rate": 1.451837705130792e-06,
+      "loss": 0.6713,
+      "step": 23090
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.203824484556633,
+      "learning_rate": 1.449771807404396e-06,
+      "loss": 0.6607,
+      "step": 23095
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.086803924453811,
+      "learning_rate": 1.447707131308133e-06,
+      "loss": 0.7089,
+      "step": 23100
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.049690406164953,
+      "learning_rate": 1.4456436775524507e-06,
+      "loss": 0.7033,
+      "step": 23105
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.761020015525577,
+      "learning_rate": 1.4435814468473858e-06,
+      "loss": 0.6751,
+      "step": 23110
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.406705510139943,
+      "learning_rate": 1.4415204399025462e-06,
+      "loss": 0.631,
+      "step": 23115
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.127818797007829,
+      "learning_rate": 1.4394606574271264e-06,
+      "loss": 0.6809,
+      "step": 23120
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 7.045815118104132,
+      "learning_rate": 1.437402100129892e-06,
+      "loss": 0.6879,
+      "step": 23125
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.939354889411304,
+      "learning_rate": 1.4353447687191958e-06,
+      "loss": 0.6896,
+      "step": 23130
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.292790763207718,
+      "learning_rate": 1.4332886639029564e-06,
+      "loss": 0.7029,
+      "step": 23135
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.154024926520374,
+      "learning_rate": 1.4312337863886832e-06,
+      "loss": 0.6413,
+      "step": 23140
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.647242944147119,
+      "learning_rate": 1.4291801368834534e-06,
+      "loss": 0.6296,
+      "step": 23145
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 6.41781643535116,
+      "learning_rate": 1.42712771609393e-06,
+      "loss": 0.6493,
+      "step": 23150
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 6.03442067665093,
+      "learning_rate": 1.4250765247263455e-06,
+      "loss": 0.6604,
+      "step": 23155
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.550340133803211,
+      "learning_rate": 1.423026563486517e-06,
+      "loss": 0.6379,
+      "step": 23160
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.03642193203217,
+      "learning_rate": 1.420977833079829e-06,
+      "loss": 0.6574,
+      "step": 23165
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.5802584153495705,
+      "learning_rate": 1.4189303342112515e-06,
+      "loss": 0.6574,
+      "step": 23170
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.571903557702888,
+      "learning_rate": 1.416884067585324e-06,
+      "loss": 0.6682,
+      "step": 23175
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.723195708364407,
+      "learning_rate": 1.4148390339061686e-06,
+      "loss": 0.6845,
+      "step": 23180
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.344770915417063,
+      "learning_rate": 1.412795233877477e-06,
+      "loss": 0.6471,
+      "step": 23185
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.694346065527504,
+      "learning_rate": 1.4107526682025196e-06,
+      "loss": 0.6362,
+      "step": 23190
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.6656465393345625,
+      "learning_rate": 1.4087113375841405e-06,
+      "loss": 0.626,
+      "step": 23195
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 7.125473741431489,
+      "learning_rate": 1.4066712427247585e-06,
+      "loss": 0.65,
+      "step": 23200
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.056846590080103,
+      "learning_rate": 1.404632384326372e-06,
+      "loss": 0.6943,
+      "step": 23205
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.925503705896461,
+      "learning_rate": 1.4025947630905462e-06,
+      "loss": 0.6557,
+      "step": 23210
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.065099671118211,
+      "learning_rate": 1.400558379718428e-06,
+      "loss": 0.7129,
+      "step": 23215
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.1811074220151765,
+      "learning_rate": 1.3985232349107337e-06,
+      "loss": 0.6796,
+      "step": 23220
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.636562868475947,
+      "learning_rate": 1.3964893293677544e-06,
+      "loss": 0.6606,
+      "step": 23225
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.413028304997516,
+      "learning_rate": 1.3944566637893532e-06,
+      "loss": 0.667,
+      "step": 23230
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.872599984362106,
+      "learning_rate": 1.392425238874972e-06,
+      "loss": 0.6637,
+      "step": 23235
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.7251617213033485,
+      "learning_rate": 1.3903950553236195e-06,
+      "loss": 0.6672,
+      "step": 23240
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.3003296095887995,
+      "learning_rate": 1.3883661138338816e-06,
+      "loss": 0.6466,
+      "step": 23245
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.494566292087462,
+      "learning_rate": 1.386338415103915e-06,
+      "loss": 0.6855,
+      "step": 23250
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.15293869738178,
+      "learning_rate": 1.384311959831448e-06,
+      "loss": 0.6595,
+      "step": 23255
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.260877143165857,
+      "learning_rate": 1.3822867487137808e-06,
+      "loss": 0.6825,
+      "step": 23260
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.584211338372017,
+      "learning_rate": 1.38026278244779e-06,
+      "loss": 0.6439,
+      "step": 23265
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 7.742745689669459,
+      "learning_rate": 1.3782400617299164e-06,
+      "loss": 0.6723,
+      "step": 23270
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.726444616139492,
+      "learning_rate": 1.37621858725618e-06,
+      "loss": 0.6351,
+      "step": 23275
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.850540870101814,
+      "learning_rate": 1.374198359722167e-06,
+      "loss": 0.6377,
+      "step": 23280
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.95117451315653,
+      "learning_rate": 1.3721793798230348e-06,
+      "loss": 0.6468,
+      "step": 23285
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.481108016903852,
+      "learning_rate": 1.370161648253512e-06,
+      "loss": 0.6457,
+      "step": 23290
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.867437599763508,
+      "learning_rate": 1.3681451657079003e-06,
+      "loss": 0.7072,
+      "step": 23295
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.348183075500717,
+      "learning_rate": 1.3661299328800671e-06,
+      "loss": 0.6474,
+      "step": 23300
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.566750840451449,
+      "learning_rate": 1.3641159504634543e-06,
+      "loss": 0.6517,
+      "step": 23305
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.349563812986563,
+      "learning_rate": 1.3621032191510708e-06,
+      "loss": 0.6582,
+      "step": 23310
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.230110035485306,
+      "learning_rate": 1.3600917396354946e-06,
+      "loss": 0.6464,
+      "step": 23315
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 10.85878257847084,
+      "learning_rate": 1.3580815126088742e-06,
+      "loss": 0.6586,
+      "step": 23320
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 11.345914802277472,
+      "learning_rate": 1.356072538762926e-06,
+      "loss": 0.643,
+      "step": 23325
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.8404459088455924,
+      "learning_rate": 1.3540648187889377e-06,
+      "loss": 0.6631,
+      "step": 23330
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.5794802817734865,
+      "learning_rate": 1.352058353377762e-06,
+      "loss": 0.6473,
+      "step": 23335
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.9389158480385,
+      "learning_rate": 1.3500531432198239e-06,
+      "loss": 0.6806,
+      "step": 23340
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 6.3870434258097735,
+      "learning_rate": 1.3480491890051128e-06,
+      "loss": 0.6266,
+      "step": 23345
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.826035233529999,
+      "learning_rate": 1.3460464914231886e-06,
+      "loss": 0.6739,
+      "step": 23350
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.647906550361867,
+      "learning_rate": 1.3440450511631748e-06,
+      "loss": 0.669,
+      "step": 23355
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.637883983239464,
+      "learning_rate": 1.3420448689137683e-06,
+      "loss": 0.6564,
+      "step": 23360
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.5577202461210335,
+      "learning_rate": 1.3400459453632274e-06,
+      "loss": 0.6824,
+      "step": 23365
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.9500063154271485,
+      "learning_rate": 1.3380482811993827e-06,
+      "loss": 0.6521,
+      "step": 23370
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.150985099315612,
+      "learning_rate": 1.3360518771096264e-06,
+      "loss": 0.6485,
+      "step": 23375
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.006946347892576,
+      "learning_rate": 1.3340567337809203e-06,
+      "loss": 0.6921,
+      "step": 23380
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.375026650050526,
+      "learning_rate": 1.332062851899789e-06,
+      "loss": 0.6219,
+      "step": 23385
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.841278654003525,
+      "learning_rate": 1.3300702321523285e-06,
+      "loss": 0.6747,
+      "step": 23390
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.856888018203015,
+      "learning_rate": 1.3280788752241945e-06,
+      "loss": 0.6781,
+      "step": 23395
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.087734511284156,
+      "learning_rate": 1.326088781800614e-06,
+      "loss": 0.7146,
+      "step": 23400
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.674249998453914,
+      "learning_rate": 1.3240999525663745e-06,
+      "loss": 0.6532,
+      "step": 23405
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.303622485507553,
+      "learning_rate": 1.3221123882058308e-06,
+      "loss": 0.6753,
+      "step": 23410
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.834287356147514,
+      "learning_rate": 1.3201260894028994e-06,
+      "loss": 0.6871,
+      "step": 23415
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.456355924902777,
+      "learning_rate": 1.318141056841068e-06,
+      "loss": 0.6214,
+      "step": 23420
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.193507840480104,
+      "learning_rate": 1.3161572912033804e-06,
+      "loss": 0.6692,
+      "step": 23425
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.216069177319419,
+      "learning_rate": 1.3141747931724525e-06,
+      "loss": 0.6571,
+      "step": 23430
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.731333771254167,
+      "learning_rate": 1.3121935634304579e-06,
+      "loss": 0.6906,
+      "step": 23435
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.787135408765722,
+      "learning_rate": 1.310213602659135e-06,
+      "loss": 0.6366,
+      "step": 23440
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.078790304953358,
+      "learning_rate": 1.308234911539788e-06,
+      "loss": 0.6694,
+      "step": 23445
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.6198193220233525,
+      "learning_rate": 1.3062574907532804e-06,
+      "loss": 0.6594,
+      "step": 23450
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.8832634578953735,
+      "learning_rate": 1.3042813409800436e-06,
+      "loss": 0.6642,
+      "step": 23455
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.567256783115041,
+      "learning_rate": 1.3023064629000658e-06,
+      "loss": 0.6852,
+      "step": 23460
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.4284491630095015,
+      "learning_rate": 1.3003328571929047e-06,
+      "loss": 0.6544,
+      "step": 23465
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.960176754434972,
+      "learning_rate": 1.2983605245376729e-06,
+      "loss": 0.6995,
+      "step": 23470
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.14495847600321,
+      "learning_rate": 1.2963894656130498e-06,
+      "loss": 0.6466,
+      "step": 23475
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.411771484130297,
+      "learning_rate": 1.2944196810972726e-06,
+      "loss": 0.6451,
+      "step": 23480
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.403061557609689,
+      "learning_rate": 1.2924511716681454e-06,
+      "loss": 0.6707,
+      "step": 23485
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 9.665497217526969,
+      "learning_rate": 1.290483938003027e-06,
+      "loss": 0.6474,
+      "step": 23490
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.353025034488214,
+      "learning_rate": 1.2885179807788445e-06,
+      "loss": 0.6332,
+      "step": 23495
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.23222585755923,
+      "learning_rate": 1.2865533006720799e-06,
+      "loss": 0.6703,
+      "step": 23500
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.189597357989905,
+      "learning_rate": 1.284589898358778e-06,
+      "loss": 0.6521,
+      "step": 23505
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.614627502483765,
+      "learning_rate": 1.2826277745145416e-06,
+      "loss": 0.6625,
+      "step": 23510
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.686704893298934,
+      "learning_rate": 1.2806669298145391e-06,
+      "loss": 0.6506,
+      "step": 23515
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.7921094655898955,
+      "learning_rate": 1.2787073649334919e-06,
+      "loss": 0.679,
+      "step": 23520
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.666328142877876,
+      "learning_rate": 1.2767490805456878e-06,
+      "loss": 0.6433,
+      "step": 23525
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.584125934385775,
+      "learning_rate": 1.2747920773249689e-06,
+      "loss": 0.6884,
+      "step": 23530
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 7.078045665988331,
+      "learning_rate": 1.2728363559447388e-06,
+      "loss": 0.6599,
+      "step": 23535
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.828946933390636,
+      "learning_rate": 1.2708819170779568e-06,
+      "loss": 0.6351,
+      "step": 23540
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.93210162599946,
+      "learning_rate": 1.2689287613971467e-06,
+      "loss": 0.6515,
+      "step": 23545
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.907369932396423,
+      "learning_rate": 1.266976889574385e-06,
+      "loss": 0.6519,
+      "step": 23550
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.496370017842469,
+      "learning_rate": 1.265026302281312e-06,
+      "loss": 0.6496,
+      "step": 23555
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.370476330349814,
+      "learning_rate": 1.2630770001891207e-06,
+      "loss": 0.6559,
+      "step": 23560
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.738527745570117,
+      "learning_rate": 1.2611289839685643e-06,
+      "loss": 0.652,
+      "step": 23565
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.5919750707047156,
+      "learning_rate": 1.2591822542899535e-06,
+      "loss": 0.6782,
+      "step": 23570
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.649660615172721,
+      "learning_rate": 1.257236811823155e-06,
+      "loss": 0.6538,
+      "step": 23575
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.520625159182393,
+      "learning_rate": 1.2552926572375957e-06,
+      "loss": 0.6689,
+      "step": 23580
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.9033900446678516,
+      "learning_rate": 1.2533497912022558e-06,
+      "loss": 0.6437,
+      "step": 23585
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.766767179662107,
+      "learning_rate": 1.2514082143856748e-06,
+      "loss": 0.6579,
+      "step": 23590
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.433224776492155,
+      "learning_rate": 1.2494679274559473e-06,
+      "loss": 0.6443,
+      "step": 23595
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.979552939552086,
+      "learning_rate": 1.2475289310807243e-06,
+      "loss": 0.6721,
+      "step": 23600
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.67763557490779,
+      "learning_rate": 1.2455912259272101e-06,
+      "loss": 0.6534,
+      "step": 23605
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.332743212638864,
+      "learning_rate": 1.2436548126621706e-06,
+      "loss": 0.6616,
+      "step": 23610
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 7.280536435536539,
+      "learning_rate": 1.241719691951921e-06,
+      "loss": 0.6559,
+      "step": 23615
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.5144024446355715,
+      "learning_rate": 1.2397858644623372e-06,
+      "loss": 0.6806,
+      "step": 23620
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.825569952036882,
+      "learning_rate": 1.2378533308588465e-06,
+      "loss": 0.6852,
+      "step": 23625
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.720916702390011,
+      "learning_rate": 1.2359220918064307e-06,
+      "loss": 0.642,
+      "step": 23630
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.798257914773307,
+      "learning_rate": 1.2339921479696272e-06,
+      "loss": 0.6712,
+      "step": 23635
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.585871975942351,
+      "learning_rate": 1.2320635000125297e-06,
+      "loss": 0.6744,
+      "step": 23640
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.093392040206783,
+      "learning_rate": 1.2301361485987817e-06,
+      "loss": 0.6983,
+      "step": 23645
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.9083491018896055,
+      "learning_rate": 1.2282100943915854e-06,
+      "loss": 0.7137,
+      "step": 23650
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.209742991514805,
+      "learning_rate": 1.2262853380536938e-06,
+      "loss": 0.66,
+      "step": 23655
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.855657115362898,
+      "learning_rate": 1.2243618802474127e-06,
+      "loss": 0.6402,
+      "step": 23660
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.611373859685703,
+      "learning_rate": 1.2224397216346012e-06,
+      "loss": 0.6706,
+      "step": 23665
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.840925762847756,
+      "learning_rate": 1.2205188628766745e-06,
+      "loss": 0.6681,
+      "step": 23670
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.042033497059596,
+      "learning_rate": 1.2185993046345961e-06,
+      "loss": 0.6468,
+      "step": 23675
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.79871160771968,
+      "learning_rate": 1.2166810475688867e-06,
+      "loss": 0.6741,
+      "step": 23680
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.105323448947047,
+      "learning_rate": 1.214764092339616e-06,
+      "loss": 0.6735,
+      "step": 23685
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.891121972055124,
+      "learning_rate": 1.212848439606405e-06,
+      "loss": 0.6561,
+      "step": 23690
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.4686279997552765,
+      "learning_rate": 1.2109340900284289e-06,
+      "loss": 0.6541,
+      "step": 23695
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.01156044355106,
+      "learning_rate": 1.2090210442644124e-06,
+      "loss": 0.668,
+      "step": 23700
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.618000263976089,
+      "learning_rate": 1.2071093029726343e-06,
+      "loss": 0.6481,
+      "step": 23705
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.163482808177356,
+      "learning_rate": 1.205198866810922e-06,
+      "loss": 0.6409,
+      "step": 23710
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.790103543725066,
+      "learning_rate": 1.2032897364366563e-06,
+      "loss": 0.6461,
+      "step": 23715
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.230244969450162,
+      "learning_rate": 1.2013819125067661e-06,
+      "loss": 0.6942,
+      "step": 23720
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.56863451913176,
+      "learning_rate": 1.199475395677731e-06,
+      "loss": 0.6497,
+      "step": 23725
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.490697737307678,
+      "learning_rate": 1.1975701866055816e-06,
+      "loss": 0.6548,
+      "step": 23730
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.674980329651185,
+      "learning_rate": 1.1956662859459006e-06,
+      "loss": 0.6524,
+      "step": 23735
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.130828897789995,
+      "learning_rate": 1.1937636943538155e-06,
+      "loss": 0.6814,
+      "step": 23740
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.096759628684305,
+      "learning_rate": 1.1918624124840094e-06,
+      "loss": 0.6951,
+      "step": 23745
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.732147767802345,
+      "learning_rate": 1.18996244099071e-06,
+      "loss": 0.6996,
+      "step": 23750
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.852807030794389,
+      "learning_rate": 1.1880637805276956e-06,
+      "loss": 0.6602,
+      "step": 23755
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.124165341584474,
+      "learning_rate": 1.1861664317482918e-06,
+      "loss": 0.6541,
+      "step": 23760
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.067019003192417,
+      "learning_rate": 1.184270395305378e-06,
+      "loss": 0.6576,
+      "step": 23765
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.195958907393889,
+      "learning_rate": 1.182375671851375e-06,
+      "loss": 0.6619,
+      "step": 23770
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.072963402580256,
+      "learning_rate": 1.1804822620382584e-06,
+      "loss": 0.6233,
+      "step": 23775
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.2949230601384825,
+      "learning_rate": 1.1785901665175474e-06,
+      "loss": 0.6862,
+      "step": 23780
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.679220448866176,
+      "learning_rate": 1.1766993859403109e-06,
+      "loss": 0.6397,
+      "step": 23785
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.529126893550042,
+      "learning_rate": 1.1748099209571618e-06,
+      "loss": 0.6602,
+      "step": 23790
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.105973681550202,
+      "learning_rate": 1.1729217722182673e-06,
+      "loss": 0.6719,
+      "step": 23795
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.280664552972269,
+      "learning_rate": 1.171034940373334e-06,
+      "loss": 0.6734,
+      "step": 23800
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.513459079235484,
+      "learning_rate": 1.1691494260716225e-06,
+      "loss": 0.6807,
+      "step": 23805
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.795157957580197,
+      "learning_rate": 1.1672652299619342e-06,
+      "loss": 0.6232,
+      "step": 23810
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 8.094072639673021,
+      "learning_rate": 1.1653823526926195e-06,
+      "loss": 0.6686,
+      "step": 23815
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.185770027064905,
+      "learning_rate": 1.163500794911575e-06,
+      "loss": 0.6832,
+      "step": 23820
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.688434083755331,
+      "learning_rate": 1.1616205572662415e-06,
+      "loss": 0.6613,
+      "step": 23825
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.044012018706139,
+      "learning_rate": 1.1597416404036095e-06,
+      "loss": 0.7053,
+      "step": 23830
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.448281320498033,
+      "learning_rate": 1.1578640449702095e-06,
+      "loss": 0.6427,
+      "step": 23835
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.846487787867139,
+      "learning_rate": 1.1559877716121232e-06,
+      "loss": 0.6528,
+      "step": 23840
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.610950244909516,
+      "learning_rate": 1.1541128209749735e-06,
+      "loss": 0.6585,
+      "step": 23845
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.877553176280549,
+      "learning_rate": 1.1522391937039284e-06,
+      "loss": 0.6696,
+      "step": 23850
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.477651738532338,
+      "learning_rate": 1.150366890443701e-06,
+      "loss": 0.677,
+      "step": 23855
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.863572635163428,
+      "learning_rate": 1.1484959118385507e-06,
+      "loss": 0.6449,
+      "step": 23860
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.7215699626198155,
+      "learning_rate": 1.1466262585322773e-06,
+      "loss": 0.6645,
+      "step": 23865
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 8.375682442605772,
+      "learning_rate": 1.1447579311682294e-06,
+      "loss": 0.7003,
+      "step": 23870
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.609454163336808,
+      "learning_rate": 1.1428909303892955e-06,
+      "loss": 0.6503,
+      "step": 23875
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.961901802189119,
+      "learning_rate": 1.1410252568379082e-06,
+      "loss": 0.6505,
+      "step": 23880
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.013798412490743,
+      "learning_rate": 1.1391609111560425e-06,
+      "loss": 0.6742,
+      "step": 23885
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.366997262349579,
+      "learning_rate": 1.137297893985222e-06,
+      "loss": 0.6679,
+      "step": 23890
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.604333797071156,
+      "learning_rate": 1.1354362059665047e-06,
+      "loss": 0.6788,
+      "step": 23895
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.409950786328532,
+      "learning_rate": 1.1335758477404995e-06,
+      "loss": 0.6651,
+      "step": 23900
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.340362687340669,
+      "learning_rate": 1.1317168199473522e-06,
+      "loss": 0.6428,
+      "step": 23905
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.850769904421837,
+      "learning_rate": 1.1298591232267524e-06,
+      "loss": 0.6648,
+      "step": 23910
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.969602368637574,
+      "learning_rate": 1.1280027582179309e-06,
+      "loss": 0.6632,
+      "step": 23915
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.396984680311234,
+      "learning_rate": 1.1261477255596632e-06,
+      "loss": 0.659,
+      "step": 23920
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.951125667931925,
+      "learning_rate": 1.124294025890262e-06,
+      "loss": 0.6385,
+      "step": 23925
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.376467123518012,
+      "learning_rate": 1.1224416598475857e-06,
+      "loss": 0.666,
+      "step": 23930
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.42513092423465,
+      "learning_rate": 1.1205906280690315e-06,
+      "loss": 0.6845,
+      "step": 23935
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.187259221881291,
+      "learning_rate": 1.1187409311915365e-06,
+      "loss": 0.6453,
+      "step": 23940
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.725902531276005,
+      "learning_rate": 1.1168925698515782e-06,
+      "loss": 0.688,
+      "step": 23945
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.6032004663512955,
+      "learning_rate": 1.11504554468518e-06,
+      "loss": 0.6641,
+      "step": 23950
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.473231294951365,
+      "learning_rate": 1.113199856327899e-06,
+      "loss": 0.6495,
+      "step": 23955
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.635391217197278,
+      "learning_rate": 1.1113555054148335e-06,
+      "loss": 0.6375,
+      "step": 23960
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.718512384709406,
+      "learning_rate": 1.1095124925806262e-06,
+      "loss": 0.6804,
+      "step": 23965
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.954687232537294,
+      "learning_rate": 1.1076708184594542e-06,
+      "loss": 0.6424,
+      "step": 23970
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.237282573011134,
+      "learning_rate": 1.1058304836850358e-06,
+      "loss": 0.6636,
+      "step": 23975
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.847006852076549,
+      "learning_rate": 1.1039914888906278e-06,
+      "loss": 0.6557,
+      "step": 23980
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.255883899216501,
+      "learning_rate": 1.1021538347090282e-06,
+      "loss": 0.6363,
+      "step": 23985
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.797507804413398,
+      "learning_rate": 1.1003175217725699e-06,
+      "loss": 0.6141,
+      "step": 23990
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.190261202725136,
+      "learning_rate": 1.098482550713129e-06,
+      "loss": 0.6663,
+      "step": 23995
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.210572111259796,
+      "learning_rate": 1.0966489221621162e-06,
+      "loss": 0.651,
+      "step": 24000
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.273773088540418,
+      "learning_rate": 1.0948166367504814e-06,
+      "loss": 0.6401,
+      "step": 24005
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.767623899763042,
+      "learning_rate": 1.0929856951087104e-06,
+      "loss": 0.6333,
+      "step": 24010
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.412046448881226,
+      "learning_rate": 1.0911560978668312e-06,
+      "loss": 0.6715,
+      "step": 24015
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.078105296172169,
+      "learning_rate": 1.0893278456544043e-06,
+      "loss": 0.6443,
+      "step": 24020
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.674597763088829,
+      "learning_rate": 1.087500939100532e-06,
+      "loss": 0.6508,
+      "step": 24025
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.110649682914834,
+      "learning_rate": 1.0856753788338498e-06,
+      "loss": 0.6628,
+      "step": 24030
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.889614638207886,
+      "learning_rate": 1.0838511654825306e-06,
+      "loss": 0.673,
+      "step": 24035
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.984844405841715,
+      "learning_rate": 1.0820282996742842e-06,
+      "loss": 0.6379,
+      "step": 24040
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.518256321449587,
+      "learning_rate": 1.0802067820363588e-06,
+      "loss": 0.6502,
+      "step": 24045
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.845221841902757,
+      "learning_rate": 1.0783866131955345e-06,
+      "loss": 0.6563,
+      "step": 24050
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.6040059079289595,
+      "learning_rate": 1.0765677937781327e-06,
+      "loss": 0.6574,
+      "step": 24055
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.081550821628119,
+      "learning_rate": 1.0747503244100055e-06,
+      "loss": 0.6582,
+      "step": 24060
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.16529242847246,
+      "learning_rate": 1.0729342057165427e-06,
+      "loss": 0.6523,
+      "step": 24065
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.544136274685921,
+      "learning_rate": 1.0711194383226675e-06,
+      "loss": 0.6813,
+      "step": 24070
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.539291616022181,
+      "learning_rate": 1.0693060228528424e-06,
+      "loss": 0.66,
+      "step": 24075
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.528425012110296,
+      "learning_rate": 1.0674939599310607e-06,
+      "loss": 0.6513,
+      "step": 24080
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.497645977169182,
+      "learning_rate": 1.06568325018085e-06,
+      "loss": 0.6785,
+      "step": 24085
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.452333913167858,
+      "learning_rate": 1.063873894225278e-06,
+      "loss": 0.6291,
+      "step": 24090
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 7.584321380477787,
+      "learning_rate": 1.062065892686937e-06,
+      "loss": 0.6179,
+      "step": 24095
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.582992398770889,
+      "learning_rate": 1.0602592461879623e-06,
+      "loss": 0.6542,
+      "step": 24100
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.267445725604313,
+      "learning_rate": 1.0584539553500167e-06,
+      "loss": 0.6883,
+      "step": 24105
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.106951719637822,
+      "learning_rate": 1.056650020794302e-06,
+      "loss": 0.6289,
+      "step": 24110
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 7.748623451954727,
+      "learning_rate": 1.0548474431415473e-06,
+      "loss": 0.6364,
+      "step": 24115
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.443000217909867,
+      "learning_rate": 1.0530462230120209e-06,
+      "loss": 0.6766,
+      "step": 24120
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.48550594637477,
+      "learning_rate": 1.0512463610255191e-06,
+      "loss": 0.6367,
+      "step": 24125
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 7.411209327002572,
+      "learning_rate": 1.049447857801373e-06,
+      "loss": 0.6673,
+      "step": 24130
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 24.461037766404242,
+      "learning_rate": 1.0476507139584447e-06,
+      "loss": 0.6432,
+      "step": 24135
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.983197491207409,
+      "learning_rate": 1.0458549301151317e-06,
+      "loss": 0.6388,
+      "step": 24140
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.714272084787362,
+      "learning_rate": 1.04406050688936e-06,
+      "loss": 0.6924,
+      "step": 24145
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.18344029670136,
+      "learning_rate": 1.0422674448985904e-06,
+      "loss": 0.6532,
+      "step": 24150
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.14167289300526,
+      "learning_rate": 1.040475744759813e-06,
+      "loss": 0.6907,
+      "step": 24155
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.647224608466981,
+      "learning_rate": 1.0386854070895496e-06,
+      "loss": 0.6835,
+      "step": 24160
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.015265097882171,
+      "learning_rate": 1.0368964325038533e-06,
+      "loss": 0.6691,
+      "step": 24165
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.5761073878121215,
+      "learning_rate": 1.035108821618311e-06,
+      "loss": 0.635,
+      "step": 24170
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.162102016764382,
+      "learning_rate": 1.0333225750480341e-06,
+      "loss": 0.6435,
+      "step": 24175
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.211445872418304,
+      "learning_rate": 1.0315376934076725e-06,
+      "loss": 0.6869,
+      "step": 24180
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.234258386467219,
+      "learning_rate": 1.0297541773114e-06,
+      "loss": 0.6333,
+      "step": 24185
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.055423005378473,
+      "learning_rate": 1.0279720273729232e-06,
+      "loss": 0.6543,
+      "step": 24190
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.32115384536653,
+      "learning_rate": 1.0261912442054767e-06,
+      "loss": 0.6433,
+      "step": 24195
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.849032712997024,
+      "learning_rate": 1.024411828421829e-06,
+      "loss": 0.66,
+      "step": 24200
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.072989418040996,
+      "learning_rate": 1.022633780634274e-06,
+      "loss": 0.68,
+      "step": 24205
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.483775285184719,
+      "learning_rate": 1.0208571014546347e-06,
+      "loss": 0.6556,
+      "step": 24210
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.745063361912836,
+      "learning_rate": 1.0190817914942696e-06,
+      "loss": 0.6549,
+      "step": 24215
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.906242377889271,
+      "learning_rate": 1.0173078513640543e-06,
+      "loss": 0.6336,
+      "step": 24220
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.337955797386548,
+      "learning_rate": 1.0155352816744058e-06,
+      "loss": 0.6436,
+      "step": 24225
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.599573279630294,
+      "learning_rate": 1.013764083035259e-06,
+      "loss": 0.6662,
+      "step": 24230
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.345380794791357,
+      "learning_rate": 1.0119942560560858e-06,
+      "loss": 0.6526,
+      "step": 24235
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.940158221488668,
+      "learning_rate": 1.0102258013458783e-06,
+      "loss": 0.6329,
+      "step": 24240
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.484670311151196,
+      "learning_rate": 1.0084587195131652e-06,
+      "loss": 0.6431,
+      "step": 24245
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.733477820239863,
+      "learning_rate": 1.0066930111659911e-06,
+      "loss": 0.66,
+      "step": 24250
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.399131133773677,
+      "learning_rate": 1.0049286769119393e-06,
+      "loss": 0.665,
+      "step": 24255
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.436299381570852,
+      "learning_rate": 1.0031657173581126e-06,
+      "loss": 0.6253,
+      "step": 24260
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.653731055188008,
+      "learning_rate": 1.0014041331111467e-06,
+      "loss": 0.6347,
+      "step": 24265
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.907820309286419,
+      "learning_rate": 9.996439247771983e-07,
+      "loss": 0.6569,
+      "step": 24270
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.544214480872975,
+      "learning_rate": 9.97885092961957e-07,
+      "loss": 0.6536,
+      "step": 24275
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.587307184447402,
+      "learning_rate": 9.961276382706297e-07,
+      "loss": 0.669,
+      "step": 24280
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.151800180599868,
+      "learning_rate": 9.943715613079602e-07,
+      "loss": 0.6471,
+      "step": 24285
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.647118728799816,
+      "learning_rate": 9.926168626782085e-07,
+      "loss": 0.6501,
+      "step": 24290
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.864846064764639,
+      "learning_rate": 9.908635429851688e-07,
+      "loss": 0.6601,
+      "step": 24295
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.941681400199244,
+      "learning_rate": 9.891116028321528e-07,
+      "loss": 0.657,
+      "step": 24300
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 7.459932839259485,
+      "learning_rate": 9.873610428220065e-07,
+      "loss": 0.6518,
+      "step": 24305
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.689099853989275,
+      "learning_rate": 9.856118635570905e-07,
+      "loss": 0.675,
+      "step": 24310
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.415278394684226,
+      "learning_rate": 9.83864065639299e-07,
+      "loss": 0.6749,
+      "step": 24315
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.166749219171328,
+      "learning_rate": 9.821176496700453e-07,
+      "loss": 0.6639,
+      "step": 24320
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.802799710730936,
+      "learning_rate": 9.803726162502726e-07,
+      "loss": 0.6423,
+      "step": 24325
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.814502203551193,
+      "learning_rate": 9.786289659804415e-07,
+      "loss": 0.6593,
+      "step": 24330
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 7.235234383606917,
+      "learning_rate": 9.768866994605452e-07,
+      "loss": 0.6661,
+      "step": 24335
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.895525185606499,
+      "learning_rate": 9.75145817290092e-07,
+      "loss": 0.6807,
+      "step": 24340
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.272616856347768,
+      "learning_rate": 9.73406320068117e-07,
+      "loss": 0.6332,
+      "step": 24345
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 7.1266478704623015,
+      "learning_rate": 9.716682083931834e-07,
+      "loss": 0.6698,
+      "step": 24350
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.686239580595646,
+      "learning_rate": 9.699314828633704e-07,
+      "loss": 0.6389,
+      "step": 24355
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 6.269954961352537,
+      "learning_rate": 9.68196144076286e-07,
+      "loss": 0.6258,
+      "step": 24360
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.586232695798104,
+      "learning_rate": 9.66462192629058e-07,
+      "loss": 0.6671,
+      "step": 24365
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.715433525036328,
+      "learning_rate": 9.647296291183367e-07,
+      "loss": 0.6313,
+      "step": 24370
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.60639124847477,
+      "learning_rate": 9.629984541402947e-07,
+      "loss": 0.6235,
+      "step": 24375
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.911709687109333,
+      "learning_rate": 9.612686682906302e-07,
+      "loss": 0.6797,
+      "step": 24380
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.824313647744851,
+      "learning_rate": 9.59540272164558e-07,
+      "loss": 0.6612,
+      "step": 24385
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.787688302758495,
+      "learning_rate": 9.578132663568207e-07,
+      "loss": 0.6451,
+      "step": 24390
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.836387498664002,
+      "learning_rate": 9.560876514616775e-07,
+      "loss": 0.6509,
+      "step": 24395
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.868549786245656,
+      "learning_rate": 9.54363428072911e-07,
+      "loss": 0.6472,
+      "step": 24400
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 6.530863320667234,
+      "learning_rate": 9.526405967838237e-07,
+      "loss": 0.6555,
+      "step": 24405
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.391324501270228,
+      "learning_rate": 9.509191581872424e-07,
+      "loss": 0.6357,
+      "step": 24410
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.6127302095667,
+      "learning_rate": 9.491991128755107e-07,
+      "loss": 0.6515,
+      "step": 24415
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.712671864213921,
+      "learning_rate": 9.474804614404959e-07,
+      "loss": 0.6574,
+      "step": 24420
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.847014489297407,
+      "learning_rate": 9.457632044735832e-07,
+      "loss": 0.6373,
+      "step": 24425
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 6.670228672865111,
+      "learning_rate": 9.440473425656815e-07,
+      "loss": 0.6506,
+      "step": 24430
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.431668423806074,
+      "learning_rate": 9.423328763072131e-07,
+      "loss": 0.6459,
+      "step": 24435
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.034693170576436,
+      "learning_rate": 9.406198062881278e-07,
+      "loss": 0.6449,
+      "step": 24440
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.678252267266481,
+      "learning_rate": 9.389081330978883e-07,
+      "loss": 0.6589,
+      "step": 24445
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.156093510328251,
+      "learning_rate": 9.371978573254831e-07,
+      "loss": 0.6928,
+      "step": 24450
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 8.083625790363284,
+      "learning_rate": 9.354889795594135e-07,
+      "loss": 0.6822,
+      "step": 24455
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.270059897769438,
+      "learning_rate": 9.33781500387706e-07,
+      "loss": 0.6475,
+      "step": 24460
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.445776137690954,
+      "learning_rate": 9.320754203978999e-07,
+      "loss": 0.6577,
+      "step": 24465
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.724132080897397,
+      "learning_rate": 9.303707401770545e-07,
+      "loss": 0.6127,
+      "step": 24470
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.683063811483321,
+      "learning_rate": 9.286674603117524e-07,
+      "loss": 0.6603,
+      "step": 24475
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.257766289785195,
+      "learning_rate": 9.269655813880868e-07,
+      "loss": 0.6006,
+      "step": 24480
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.429484324997625,
+      "learning_rate": 9.25265103991676e-07,
+      "loss": 0.6603,
+      "step": 24485
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.8748023788537385,
+      "learning_rate": 9.235660287076509e-07,
+      "loss": 0.6937,
+      "step": 24490
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.573077935697372,
+      "learning_rate": 9.218683561206621e-07,
+      "loss": 0.6657,
+      "step": 24495
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.0768056663720245,
+      "learning_rate": 9.201720868148756e-07,
+      "loss": 0.6426,
+      "step": 24500
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.132452067770202,
+      "learning_rate": 9.184772213739784e-07,
+      "loss": 0.6173,
+      "step": 24505
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.8371733786325946,
+      "learning_rate": 9.167837603811702e-07,
+      "loss": 0.6421,
+      "step": 24510
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.925534485149385,
+      "learning_rate": 9.15091704419171e-07,
+      "loss": 0.6504,
+      "step": 24515
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.885909463985392,
+      "learning_rate": 9.134010540702148e-07,
+      "loss": 0.6602,
+      "step": 24520
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.120338319418244,
+      "learning_rate": 9.117118099160526e-07,
+      "loss": 0.6821,
+      "step": 24525
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.798152916059508,
+      "learning_rate": 9.100239725379512e-07,
+      "loss": 0.6748,
+      "step": 24530
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.905722956211278,
+      "learning_rate": 9.083375425166952e-07,
+      "loss": 0.6423,
+      "step": 24535
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.674599790896468,
+      "learning_rate": 9.066525204325816e-07,
+      "loss": 0.6314,
+      "step": 24540
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.787652908868984,
+      "learning_rate": 9.049689068654277e-07,
+      "loss": 0.6461,
+      "step": 24545
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.648534073284891,
+      "learning_rate": 9.032867023945613e-07,
+      "loss": 0.6802,
+      "step": 24550
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.963734604380744,
+      "learning_rate": 9.016059075988282e-07,
+      "loss": 0.6637,
+      "step": 24555
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.815717399319649,
+      "learning_rate": 8.999265230565863e-07,
+      "loss": 0.666,
+      "step": 24560
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.534025588775262,
+      "learning_rate": 8.982485493457133e-07,
+      "loss": 0.6598,
+      "step": 24565
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.243752686300016,
+      "learning_rate": 8.965719870435951e-07,
+      "loss": 0.679,
+      "step": 24570
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.169525277173541,
+      "learning_rate": 8.948968367271388e-07,
+      "loss": 0.6735,
+      "step": 24575
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.372567152329839,
+      "learning_rate": 8.932230989727598e-07,
+      "loss": 0.6407,
+      "step": 24580
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.503524583394472,
+      "learning_rate": 8.915507743563901e-07,
+      "loss": 0.6268,
+      "step": 24585
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.582464553338467,
+      "learning_rate": 8.89879863453475e-07,
+      "loss": 0.6439,
+      "step": 24590
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.4729040322625515,
+      "learning_rate": 8.882103668389719e-07,
+      "loss": 0.6668,
+      "step": 24595
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.280305513971691,
+      "learning_rate": 8.865422850873556e-07,
+      "loss": 0.6826,
+      "step": 24600
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.931605865332295,
+      "learning_rate": 8.848756187726093e-07,
+      "loss": 0.6705,
+      "step": 24605
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.2975829385507325,
+      "learning_rate": 8.83210368468233e-07,
+      "loss": 0.6587,
+      "step": 24610
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.032825354671978,
+      "learning_rate": 8.815465347472374e-07,
+      "loss": 0.657,
+      "step": 24615
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.7918494728805,
+      "learning_rate": 8.798841181821449e-07,
+      "loss": 0.6433,
+      "step": 24620
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.6352119324579295,
+      "learning_rate": 8.782231193449914e-07,
+      "loss": 0.6052,
+      "step": 24625
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.468768095821291,
+      "learning_rate": 8.765635388073274e-07,
+      "loss": 0.6506,
+      "step": 24630
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.1714584235928,
+      "learning_rate": 8.749053771402094e-07,
+      "loss": 0.6813,
+      "step": 24635
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.405067333949699,
+      "learning_rate": 8.732486349142127e-07,
+      "loss": 0.6597,
+      "step": 24640
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.641166645937143,
+      "learning_rate": 8.715933126994192e-07,
+      "loss": 0.6686,
+      "step": 24645
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.540422303816012,
+      "learning_rate": 8.699394110654241e-07,
+      "loss": 0.6277,
+      "step": 24650
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 8.926265663207731,
+      "learning_rate": 8.682869305813312e-07,
+      "loss": 0.6619,
+      "step": 24655
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.522604459042804,
+      "learning_rate": 8.666358718157608e-07,
+      "loss": 0.6753,
+      "step": 24660
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.8097535901666495,
+      "learning_rate": 8.649862353368377e-07,
+      "loss": 0.6776,
+      "step": 24665
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.158174973798557,
+      "learning_rate": 8.633380217122028e-07,
+      "loss": 0.6765,
+      "step": 24670
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.610447525325233,
+      "learning_rate": 8.616912315090037e-07,
+      "loss": 0.6734,
+      "step": 24675
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.612845668220532,
+      "learning_rate": 8.600458652938997e-07,
+      "loss": 0.6521,
+      "step": 24680
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.578334838267179,
+      "learning_rate": 8.584019236330576e-07,
+      "loss": 0.6442,
+      "step": 24685
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.757414987867542,
+      "learning_rate": 8.567594070921592e-07,
+      "loss": 0.6377,
+      "step": 24690
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.545040668554131,
+      "learning_rate": 8.551183162363907e-07,
+      "loss": 0.6203,
+      "step": 24695
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.18825691136152,
+      "learning_rate": 8.534786516304516e-07,
+      "loss": 0.6589,
+      "step": 24700
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.63240360899482,
+      "learning_rate": 8.518404138385483e-07,
+      "loss": 0.6737,
+      "step": 24705
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.375068745954727,
+      "learning_rate": 8.502036034243965e-07,
+      "loss": 0.6233,
+      "step": 24710
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.711918972811405,
+      "learning_rate": 8.485682209512203e-07,
+      "loss": 0.6563,
+      "step": 24715
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.740392950733569,
+      "learning_rate": 8.469342669817548e-07,
+      "loss": 0.6779,
+      "step": 24720
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.782537893732716,
+      "learning_rate": 8.453017420782422e-07,
+      "loss": 0.6634,
+      "step": 24725
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.0966574448446185,
+      "learning_rate": 8.436706468024303e-07,
+      "loss": 0.67,
+      "step": 24730
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.884923814880293,
+      "learning_rate": 8.420409817155806e-07,
+      "loss": 0.6687,
+      "step": 24735
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.659544018108539,
+      "learning_rate": 8.404127473784574e-07,
+      "loss": 0.6462,
+      "step": 24740
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.709118306156166,
+      "learning_rate": 8.387859443513353e-07,
+      "loss": 0.679,
+      "step": 24745
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.928574888944737,
+      "learning_rate": 8.371605731939935e-07,
+      "loss": 0.6381,
+      "step": 24750
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.063956339138812,
+      "learning_rate": 8.355366344657245e-07,
+      "loss": 0.6671,
+      "step": 24755
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.831064742072634,
+      "learning_rate": 8.339141287253205e-07,
+      "loss": 0.6376,
+      "step": 24760
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.496507983904041,
+      "learning_rate": 8.322930565310867e-07,
+      "loss": 0.6523,
+      "step": 24765
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.897423559897414,
+      "learning_rate": 8.306734184408316e-07,
+      "loss": 0.654,
+      "step": 24770
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.895179239345184,
+      "learning_rate": 8.290552150118714e-07,
+      "loss": 0.6587,
+      "step": 24775
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.01025021028483,
+      "learning_rate": 8.274384468010266e-07,
+      "loss": 0.6599,
+      "step": 24780
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.621504495377497,
+      "learning_rate": 8.258231143646284e-07,
+      "loss": 0.6536,
+      "step": 24785
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.015773076364035,
+      "learning_rate": 8.242092182585082e-07,
+      "loss": 0.6886,
+      "step": 24790
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.427442287654225,
+      "learning_rate": 8.225967590380091e-07,
+      "loss": 0.6637,
+      "step": 24795
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 6.137015803961235,
+      "learning_rate": 8.209857372579749e-07,
+      "loss": 0.6189,
+      "step": 24800
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.886094258233801,
+      "learning_rate": 8.193761534727579e-07,
+      "loss": 0.6334,
+      "step": 24805
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.027983452333498,
+      "learning_rate": 8.177680082362116e-07,
+      "loss": 0.6403,
+      "step": 24810
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.498959374655465,
+      "learning_rate": 8.161613021017006e-07,
+      "loss": 0.665,
+      "step": 24815
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.442887102085162,
+      "learning_rate": 8.145560356220883e-07,
+      "loss": 0.6512,
+      "step": 24820
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 6.059385265435554,
+      "learning_rate": 8.129522093497483e-07,
+      "loss": 0.661,
+      "step": 24825
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.89352893079638,
+      "learning_rate": 8.11349823836553e-07,
+      "loss": 0.6473,
+      "step": 24830
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.557086942989607,
+      "learning_rate": 8.097488796338837e-07,
+      "loss": 0.6246,
+      "step": 24835
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.471587700245278,
+      "learning_rate": 8.081493772926208e-07,
+      "loss": 0.6219,
+      "step": 24840
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 6.8867025140941545,
+      "learning_rate": 8.06551317363155e-07,
+      "loss": 0.6267,
+      "step": 24845
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.849642771433342,
+      "learning_rate": 8.049547003953756e-07,
+      "loss": 0.6704,
+      "step": 24850
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.900048710600587,
+      "learning_rate": 8.033595269386762e-07,
+      "loss": 0.7352,
+      "step": 24855
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.458521029193161,
+      "learning_rate": 8.017657975419563e-07,
+      "loss": 0.6711,
+      "step": 24860
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.234623918704893,
+      "learning_rate": 8.001735127536153e-07,
+      "loss": 0.6682,
+      "step": 24865
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.8119753199973125,
+      "learning_rate": 7.985826731215579e-07,
+      "loss": 0.6526,
+      "step": 24870
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.786871359170267,
+      "learning_rate": 7.969932791931883e-07,
+      "loss": 0.6588,
+      "step": 24875
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.888432868770594,
+      "learning_rate": 7.954053315154181e-07,
+      "loss": 0.6491,
+      "step": 24880
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.7430510735248745,
+      "learning_rate": 7.938188306346567e-07,
+      "loss": 0.6593,
+      "step": 24885
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 6.1545479704772434,
+      "learning_rate": 7.922337770968192e-07,
+      "loss": 0.6435,
+      "step": 24890
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.849948977490544,
+      "learning_rate": 7.906501714473203e-07,
+      "loss": 0.6504,
+      "step": 24895
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 9.081958242724612,
+      "learning_rate": 7.89068014231077e-07,
+      "loss": 0.6467,
+      "step": 24900
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.470923465659294,
+      "learning_rate": 7.874873059925064e-07,
+      "loss": 0.6437,
+      "step": 24905
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.654415047763868,
+      "learning_rate": 7.85908047275532e-07,
+      "loss": 0.6388,
+      "step": 24910
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 6.073545061463334,
+      "learning_rate": 7.843302386235724e-07,
+      "loss": 0.6478,
+      "step": 24915
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.715609234994534,
+      "learning_rate": 7.827538805795526e-07,
+      "loss": 0.6644,
+      "step": 24920
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.964046487206026,
+      "learning_rate": 7.811789736858943e-07,
+      "loss": 0.6411,
+      "step": 24925
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 6.154862219711928,
+      "learning_rate": 7.796055184845219e-07,
+      "loss": 0.6658,
+      "step": 24930
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.905050013439889,
+      "learning_rate": 7.780335155168584e-07,
+      "loss": 0.6531,
+      "step": 24935
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.581144513340311,
+      "learning_rate": 7.764629653238309e-07,
+      "loss": 0.661,
+      "step": 24940
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.231113583542581,
+      "learning_rate": 7.748938684458623e-07,
+      "loss": 0.7024,
+      "step": 24945
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.860678825782125,
+      "learning_rate": 7.733262254228791e-07,
+      "loss": 0.6368,
+      "step": 24950
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.975033156518935,
+      "learning_rate": 7.717600367943046e-07,
+      "loss": 0.6649,
+      "step": 24955
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.01741938313855,
+      "learning_rate": 7.701953030990628e-07,
+      "loss": 0.658,
+      "step": 24960
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.3034097032783265,
+      "learning_rate": 7.686320248755763e-07,
+      "loss": 0.6758,
+      "step": 24965
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 7.27663529194628,
+      "learning_rate": 7.670702026617699e-07,
+      "loss": 0.6453,
+      "step": 24970
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.33105884995202,
+      "learning_rate": 7.655098369950643e-07,
+      "loss": 0.6452,
+      "step": 24975
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.392173457773588,
+      "learning_rate": 7.639509284123776e-07,
+      "loss": 0.6375,
+      "step": 24980
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.3268626627599165,
+      "learning_rate": 7.623934774501318e-07,
+      "loss": 0.6545,
+      "step": 24985
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.913510886945881,
+      "learning_rate": 7.608374846442429e-07,
+      "loss": 0.6389,
+      "step": 24990
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.420185217342942,
+      "learning_rate": 7.592829505301269e-07,
+      "loss": 0.6462,
+      "step": 24995
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.50249743438501,
+      "learning_rate": 7.577298756426959e-07,
+      "loss": 0.6746,
+      "step": 25000
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.406885599331131,
+      "learning_rate": 7.561782605163642e-07,
+      "loss": 0.6584,
+      "step": 25005
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.776451399414106,
+      "learning_rate": 7.546281056850386e-07,
+      "loss": 0.647,
+      "step": 25010
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.647823902193932,
+      "learning_rate": 7.530794116821282e-07,
+      "loss": 0.6189,
+      "step": 25015
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.680773430815062,
+      "learning_rate": 7.515321790405356e-07,
+      "loss": 0.6375,
+      "step": 25020
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.969723876599489,
+      "learning_rate": 7.499864082926627e-07,
+      "loss": 0.6222,
+      "step": 25025
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.758278915875785,
+      "learning_rate": 7.484420999704067e-07,
+      "loss": 0.6381,
+      "step": 25030
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.498832604608828,
+      "learning_rate": 7.46899254605164e-07,
+      "loss": 0.6355,
+      "step": 25035
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 6.443240026602288,
+      "learning_rate": 7.453578727278249e-07,
+      "loss": 0.6575,
+      "step": 25040
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.819968215010338,
+      "learning_rate": 7.43817954868779e-07,
+      "loss": 0.6548,
+      "step": 25045
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.28827657407379,
+      "learning_rate": 7.422795015579098e-07,
+      "loss": 0.619,
+      "step": 25050
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.686959750334854,
+      "learning_rate": 7.407425133245971e-07,
+      "loss": 0.6313,
+      "step": 25055
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 8.081679427774791,
+      "learning_rate": 7.392069906977167e-07,
+      "loss": 0.651,
+      "step": 25060
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.242750449479009,
+      "learning_rate": 7.376729342056427e-07,
+      "loss": 0.6706,
+      "step": 25065
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.727060169319375,
+      "learning_rate": 7.361403443762399e-07,
+      "loss": 0.664,
+      "step": 25070
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.85291155153888,
+      "learning_rate": 7.346092217368727e-07,
+      "loss": 0.6669,
+      "step": 25075
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.782691470831053,
+      "learning_rate": 7.330795668143992e-07,
+      "loss": 0.6496,
+      "step": 25080
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.031602082846597,
+      "learning_rate": 7.315513801351709e-07,
+      "loss": 0.6241,
+      "step": 25085
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.13748043404483,
+      "learning_rate": 7.300246622250345e-07,
+      "loss": 0.6558,
+      "step": 25090
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.022918393131529,
+      "learning_rate": 7.284994136093348e-07,
+      "loss": 0.6203,
+      "step": 25095
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.577534653944751,
+      "learning_rate": 7.269756348129064e-07,
+      "loss": 0.6342,
+      "step": 25100
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.3508017293347105,
+      "learning_rate": 7.254533263600816e-07,
+      "loss": 0.6571,
+      "step": 25105
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.6107215829957715,
+      "learning_rate": 7.239324887746846e-07,
+      "loss": 0.6478,
+      "step": 25110
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.9246172109137705,
+      "learning_rate": 7.22413122580034e-07,
+      "loss": 0.6753,
+      "step": 25115
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.752569415648657,
+      "learning_rate": 7.208952282989423e-07,
+      "loss": 0.6719,
+      "step": 25120
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.074928665972307,
+      "learning_rate": 7.193788064537149e-07,
+      "loss": 0.6584,
+      "step": 25125
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.459360160030664,
+      "learning_rate": 7.178638575661523e-07,
+      "loss": 0.6757,
+      "step": 25130
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.9071075389801475,
+      "learning_rate": 7.163503821575457e-07,
+      "loss": 0.6371,
+      "step": 25135
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.469886611538309,
+      "learning_rate": 7.14838380748682e-07,
+      "loss": 0.651,
+      "step": 25140
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.255804557272553,
+      "learning_rate": 7.133278538598388e-07,
+      "loss": 0.6691,
+      "step": 25145
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.76982918005449,
+      "learning_rate": 7.118188020107869e-07,
+      "loss": 0.6618,
+      "step": 25150
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.35076015863343,
+      "learning_rate": 7.103112257207884e-07,
+      "loss": 0.617,
+      "step": 25155
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.487834302585936,
+      "learning_rate": 7.088051255086015e-07,
+      "loss": 0.6986,
+      "step": 25160
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.413840768505342,
+      "learning_rate": 7.073005018924706e-07,
+      "loss": 0.694,
+      "step": 25165
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.55955416421179,
+      "learning_rate": 7.057973553901387e-07,
+      "loss": 0.6594,
+      "step": 25170
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.613757922631881,
+      "learning_rate": 7.04295686518835e-07,
+      "loss": 0.6504,
+      "step": 25175
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.900243725454976,
+      "learning_rate": 7.027954957952827e-07,
+      "loss": 0.6405,
+      "step": 25180
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.027603532032837,
+      "learning_rate": 7.012967837356949e-07,
+      "loss": 0.6339,
+      "step": 25185
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.712147105387408,
+      "learning_rate": 6.997995508557792e-07,
+      "loss": 0.6172,
+      "step": 25190
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.581397598376956,
+      "learning_rate": 6.983037976707297e-07,
+      "loss": 0.6563,
+      "step": 25195
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.936731592965554,
+      "learning_rate": 6.968095246952361e-07,
+      "loss": 0.6421,
+      "step": 25200
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.692962308108754,
+      "learning_rate": 6.953167324434751e-07,
+      "loss": 0.6399,
+      "step": 25205
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 11.914079098706027,
+      "learning_rate": 6.938254214291152e-07,
+      "loss": 0.6798,
+      "step": 25210
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.763857783003577,
+      "learning_rate": 6.92335592165313e-07,
+      "loss": 0.6922,
+      "step": 25215
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.452148035606174,
+      "learning_rate": 6.908472451647213e-07,
+      "loss": 0.6676,
+      "step": 25220
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.178984053236935,
+      "learning_rate": 6.89360380939475e-07,
+      "loss": 0.6449,
+      "step": 25225
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.261025469317658,
+      "learning_rate": 6.878750000012058e-07,
+      "loss": 0.6287,
+      "step": 25230
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.163820142436162,
+      "learning_rate": 6.863911028610309e-07,
+      "loss": 0.6342,
+      "step": 25235
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.186245405153009,
+      "learning_rate": 6.849086900295571e-07,
+      "loss": 0.6386,
+      "step": 25240
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.625364394762081,
+      "learning_rate": 6.834277620168817e-07,
+      "loss": 0.628,
+      "step": 25245
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.309413384246584,
+      "learning_rate": 6.819483193325898e-07,
+      "loss": 0.6031,
+      "step": 25250
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.5437712430174235,
+      "learning_rate": 6.804703624857578e-07,
+      "loss": 0.6801,
+      "step": 25255
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.719041204609119,
+      "learning_rate": 6.789938919849475e-07,
+      "loss": 0.6371,
+      "step": 25260
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.4311190616552825,
+      "learning_rate": 6.775189083382128e-07,
+      "loss": 0.6481,
+      "step": 25265
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.49817950043537,
+      "learning_rate": 6.760454120530935e-07,
+      "loss": 0.6764,
+      "step": 25270
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.075615059252561,
+      "learning_rate": 6.74573403636618e-07,
+      "loss": 0.6722,
+      "step": 25275
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.756692975556284,
+      "learning_rate": 6.731028835953024e-07,
+      "loss": 0.6153,
+      "step": 25280
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.111532811336681,
+      "learning_rate": 6.716338524351523e-07,
+      "loss": 0.6435,
+      "step": 25285
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.235658015353194,
+      "learning_rate": 6.701663106616591e-07,
+      "loss": 0.6366,
+      "step": 25290
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 8.998525285202195,
+      "learning_rate": 6.687002587798036e-07,
+      "loss": 0.617,
+      "step": 25295
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.636753103897285,
+      "learning_rate": 6.672356972940519e-07,
+      "loss": 0.6578,
+      "step": 25300
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.007653059949343,
+      "learning_rate": 6.657726267083592e-07,
+      "loss": 0.6836,
+      "step": 25305
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 7.4049423574361946,
+      "learning_rate": 6.643110475261644e-07,
+      "loss": 0.6093,
+      "step": 25310
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.575892380855971,
+      "learning_rate": 6.62850960250398e-07,
+      "loss": 0.6632,
+      "step": 25315
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.247916695481189,
+      "learning_rate": 6.613923653834731e-07,
+      "loss": 0.6536,
+      "step": 25320
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.71368785157455,
+      "learning_rate": 6.599352634272921e-07,
+      "loss": 0.6091,
+      "step": 25325
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.105286131354021,
+      "learning_rate": 6.584796548832422e-07,
+      "loss": 0.6343,
+      "step": 25330
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.713066779444538,
+      "learning_rate": 6.570255402521964e-07,
+      "loss": 0.6458,
+      "step": 25335
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.530370525092025,
+      "learning_rate": 6.555729200345123e-07,
+      "loss": 0.6326,
+      "step": 25340
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.959540168806156,
+      "learning_rate": 6.541217947300388e-07,
+      "loss": 0.6574,
+      "step": 25345
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.631356069842528,
+      "learning_rate": 6.52672164838104e-07,
+      "loss": 0.6537,
+      "step": 25350
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.655849920407781,
+      "learning_rate": 6.512240308575257e-07,
+      "loss": 0.6484,
+      "step": 25355
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.14779961172738,
+      "learning_rate": 6.497773932866064e-07,
+      "loss": 0.634,
+      "step": 25360
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 6.745994577184643,
+      "learning_rate": 6.483322526231284e-07,
+      "loss": 0.651,
+      "step": 25365
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.284058608425205,
+      "learning_rate": 6.468886093643673e-07,
+      "loss": 0.6718,
+      "step": 25370
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.781063229895192,
+      "learning_rate": 6.454464640070767e-07,
+      "loss": 0.6623,
+      "step": 25375
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.43704953047697,
+      "learning_rate": 6.440058170475006e-07,
+      "loss": 0.6134,
+      "step": 25380
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.953656567496981,
+      "learning_rate": 6.425666689813603e-07,
+      "loss": 0.7096,
+      "step": 25385
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.7737790556705,
+      "learning_rate": 6.411290203038705e-07,
+      "loss": 0.643,
+      "step": 25390
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.713787102142757,
+      "learning_rate": 6.396928715097189e-07,
+      "loss": 0.6776,
+      "step": 25395
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.8241511109663975,
+      "learning_rate": 6.382582230930868e-07,
+      "loss": 0.6387,
+      "step": 25400
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.215920059201571,
+      "learning_rate": 6.368250755476329e-07,
+      "loss": 0.6091,
+      "step": 25405
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 6.24655780189847,
+      "learning_rate": 6.353934293665049e-07,
+      "loss": 0.6465,
+      "step": 25410
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.51722974692802,
+      "learning_rate": 6.339632850423272e-07,
+      "loss": 0.6697,
+      "step": 25415
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.942691073595876,
+      "learning_rate": 6.325346430672158e-07,
+      "loss": 0.6949,
+      "step": 25420
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.600010750073224,
+      "learning_rate": 6.311075039327602e-07,
+      "loss": 0.6358,
+      "step": 25425
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.194071088902065,
+      "learning_rate": 6.296818681300409e-07,
+      "loss": 0.633,
+      "step": 25430
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.352558847451886,
+      "learning_rate": 6.28257736149615e-07,
+      "loss": 0.6241,
+      "step": 25435
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.656036452931519,
+      "learning_rate": 6.268351084815283e-07,
+      "loss": 0.6695,
+      "step": 25440
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.692429799378556,
+      "learning_rate": 6.254139856153024e-07,
+      "loss": 0.6111,
+      "step": 25445
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.833729995327521,
+      "learning_rate": 6.23994368039948e-07,
+      "loss": 0.6356,
+      "step": 25450
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.541053282950694,
+      "learning_rate": 6.225762562439503e-07,
+      "loss": 0.6815,
+      "step": 25455
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.63160893567276,
+      "learning_rate": 6.211596507152823e-07,
+      "loss": 0.6209,
+      "step": 25460
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.7630558895319615,
+      "learning_rate": 6.197445519413958e-07,
+      "loss": 0.6881,
+      "step": 25465
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.555410308682889,
+      "learning_rate": 6.183309604092258e-07,
+      "loss": 0.632,
+      "step": 25470
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.86849347035945,
+      "learning_rate": 6.16918876605187e-07,
+      "loss": 0.6498,
+      "step": 25475
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.873602369763425,
+      "learning_rate": 6.155083010151769e-07,
+      "loss": 0.6766,
+      "step": 25480
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.367303692125798,
+      "learning_rate": 6.140992341245728e-07,
+      "loss": 0.6312,
+      "step": 25485
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.862852100253972,
+      "learning_rate": 6.126916764182334e-07,
+      "loss": 0.6281,
+      "step": 25490
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.515087993078539,
+      "learning_rate": 6.11285628380498e-07,
+      "loss": 0.6453,
+      "step": 25495
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.2591560119771845,
+      "learning_rate": 6.098810904951847e-07,
+      "loss": 0.6283,
+      "step": 25500
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.87669698541844,
+      "learning_rate": 6.084780632455967e-07,
+      "loss": 0.6248,
+      "step": 25505
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.576326511788119,
+      "learning_rate": 6.070765471145113e-07,
+      "loss": 0.6219,
+      "step": 25510
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.332526356057729,
+      "learning_rate": 6.056765425841921e-07,
+      "loss": 0.667,
+      "step": 25515
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.774005536488563,
+      "learning_rate": 6.04278050136376e-07,
+      "loss": 0.633,
+      "step": 25520
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.115837864694559,
+      "learning_rate": 6.028810702522853e-07,
+      "loss": 0.6713,
+      "step": 25525
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.133344507166991,
+      "learning_rate": 6.014856034126177e-07,
+      "loss": 0.6668,
+      "step": 25530
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.432610346496629,
+      "learning_rate": 6.000916500975534e-07,
+      "loss": 0.6277,
+      "step": 25535
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.360887456370527,
+      "learning_rate": 5.986992107867495e-07,
+      "loss": 0.6635,
+      "step": 25540
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.656144876230373,
+      "learning_rate": 5.973082859593448e-07,
+      "loss": 0.6342,
+      "step": 25545
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.42985086767463,
+      "learning_rate": 5.959188760939527e-07,
+      "loss": 0.6248,
+      "step": 25550
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.887388152336578,
+      "learning_rate": 5.945309816686695e-07,
+      "loss": 0.588,
+      "step": 25555
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.9624664483204315,
+      "learning_rate": 5.931446031610666e-07,
+      "loss": 0.6716,
+      "step": 25560
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.040677312192779,
+      "learning_rate": 5.917597410481979e-07,
+      "loss": 0.6604,
+      "step": 25565
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.935228644214927,
+      "learning_rate": 5.903763958065912e-07,
+      "loss": 0.6344,
+      "step": 25570
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.952775601502335,
+      "learning_rate": 5.889945679122566e-07,
+      "loss": 0.6559,
+      "step": 25575
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.7423239151175185,
+      "learning_rate": 5.876142578406763e-07,
+      "loss": 0.6531,
+      "step": 25580
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.254082493619603,
+      "learning_rate": 5.862354660668168e-07,
+      "loss": 0.6331,
+      "step": 25585
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.633593931063916,
+      "learning_rate": 5.848581930651165e-07,
+      "loss": 0.6595,
+      "step": 25590
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.51468653861724,
+      "learning_rate": 5.834824393094962e-07,
+      "loss": 0.668,
+      "step": 25595
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 6.2054497461118,
+      "learning_rate": 5.821082052733495e-07,
+      "loss": 0.6203,
+      "step": 25600
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 6.41402199262695,
+      "learning_rate": 5.807354914295516e-07,
+      "loss": 0.6508,
+      "step": 25605
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.13511331457535,
+      "learning_rate": 5.793642982504477e-07,
+      "loss": 0.6391,
+      "step": 25610
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.619860066420089,
+      "learning_rate": 5.779946262078683e-07,
+      "loss": 0.6209,
+      "step": 25615
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.702009815902476,
+      "learning_rate": 5.766264757731144e-07,
+      "loss": 0.6334,
+      "step": 25620
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 7.854644332692116,
+      "learning_rate": 5.752598474169641e-07,
+      "loss": 0.6912,
+      "step": 25625
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.9694105036745775,
+      "learning_rate": 5.738947416096752e-07,
+      "loss": 0.6628,
+      "step": 25630
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.500559211637084,
+      "learning_rate": 5.725311588209786e-07,
+      "loss": 0.6253,
+      "step": 25635
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.848868708725839,
+      "learning_rate": 5.711690995200814e-07,
+      "loss": 0.6545,
+      "step": 25640
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.647854890983991,
+      "learning_rate": 5.698085641756657e-07,
+      "loss": 0.6283,
+      "step": 25645
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.390714116328605,
+      "learning_rate": 5.684495532558931e-07,
+      "loss": 0.6334,
+      "step": 25650
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.771148030368524,
+      "learning_rate": 5.670920672283958e-07,
+      "loss": 0.6491,
+      "step": 25655
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.782194409574925,
+      "learning_rate": 5.657361065602846e-07,
+      "loss": 0.6233,
+      "step": 25660
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.897256882115723,
+      "learning_rate": 5.643816717181444e-07,
+      "loss": 0.6263,
+      "step": 25665
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.424184580177064,
+      "learning_rate": 5.630287631680343e-07,
+      "loss": 0.5929,
+      "step": 25670
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.881655967229124,
+      "learning_rate": 5.616773813754883e-07,
+      "loss": 0.6481,
+      "step": 25675
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.506211542493089,
+      "learning_rate": 5.60327526805517e-07,
+      "loss": 0.6363,
+      "step": 25680
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 7.092793877990308,
+      "learning_rate": 5.589791999226024e-07,
+      "loss": 0.6262,
+      "step": 25685
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.5625815137116215,
+      "learning_rate": 5.576324011907041e-07,
+      "loss": 0.6374,
+      "step": 25690
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.210948120306893,
+      "learning_rate": 5.562871310732543e-07,
+      "loss": 0.6637,
+      "step": 25695
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.441147446353072,
+      "learning_rate": 5.54943390033158e-07,
+      "loss": 0.6427,
+      "step": 25700
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.73805515667014,
+      "learning_rate": 5.53601178532795e-07,
+      "loss": 0.6232,
+      "step": 25705
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 9.367015312850405,
+      "learning_rate": 5.522604970340201e-07,
+      "loss": 0.6698,
+      "step": 25710
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 6.430392634994808,
+      "learning_rate": 5.509213459981594e-07,
+      "loss": 0.6715,
+      "step": 25715
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 6.593853208661611,
+      "learning_rate": 5.495837258860154e-07,
+      "loss": 0.6764,
+      "step": 25720
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.917194336277024,
+      "learning_rate": 5.482476371578605e-07,
+      "loss": 0.6576,
+      "step": 25725
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.599302216384735,
+      "learning_rate": 5.469130802734418e-07,
+      "loss": 0.6801,
+      "step": 25730
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.5781171029935255,
+      "learning_rate": 5.455800556919777e-07,
+      "loss": 0.628,
+      "step": 25735
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.24417630576127,
+      "learning_rate": 5.442485638721635e-07,
+      "loss": 0.6454,
+      "step": 25740
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 6.1990419917474675,
+      "learning_rate": 5.429186052721613e-07,
+      "loss": 0.6697,
+      "step": 25745
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.870940237800007,
+      "learning_rate": 5.415901803496109e-07,
+      "loss": 0.6187,
+      "step": 25750
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.990545890958486,
+      "learning_rate": 5.402632895616217e-07,
+      "loss": 0.6298,
+      "step": 25755
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 6.069670700384858,
+      "learning_rate": 5.389379333647748e-07,
+      "loss": 0.6339,
+      "step": 25760
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.133013881271903,
+      "learning_rate": 5.37614112215124e-07,
+      "loss": 0.6353,
+      "step": 25765
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.550175486105931,
+      "learning_rate": 5.362918265681943e-07,
+      "loss": 0.5891,
+      "step": 25770
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.150306726699754,
+      "learning_rate": 5.349710768789851e-07,
+      "loss": 0.6505,
+      "step": 25775
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.704186001901201,
+      "learning_rate": 5.336518636019622e-07,
+      "loss": 0.6493,
+      "step": 25780
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.631385357677084,
+      "learning_rate": 5.323341871910687e-07,
+      "loss": 0.6723,
+      "step": 25785
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.732305304742764,
+      "learning_rate": 5.310180480997135e-07,
+      "loss": 0.6353,
+      "step": 25790
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.81983535599789,
+      "learning_rate": 5.297034467807805e-07,
+      "loss": 0.6561,
+      "step": 25795
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.4578611668626955,
+      "learning_rate": 5.2839038368662e-07,
+      "loss": 0.6382,
+      "step": 25800
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.919902980430912,
+      "learning_rate": 5.27078859269059e-07,
+      "loss": 0.6479,
+      "step": 25805
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.126133325799623,
+      "learning_rate": 5.257688739793892e-07,
+      "loss": 0.678,
+      "step": 25810
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 6.436875044498097,
+      "learning_rate": 5.24460428268378e-07,
+      "loss": 0.6517,
+      "step": 25815
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.790793229197409,
+      "learning_rate": 5.231535225862583e-07,
+      "loss": 0.6788,
+      "step": 25820
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.505001576559171,
+      "learning_rate": 5.218481573827356e-07,
+      "loss": 0.6755,
+      "step": 25825
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.509741158174838,
+      "learning_rate": 5.205443331069843e-07,
+      "loss": 0.6237,
+      "step": 25830
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.540920818207695,
+      "learning_rate": 5.1924205020765e-07,
+      "loss": 0.6581,
+      "step": 25835
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.4711223646163125,
+      "learning_rate": 5.179413091328461e-07,
+      "loss": 0.6207,
+      "step": 25840
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 6.107980371352623,
+      "learning_rate": 5.166421103301572e-07,
+      "loss": 0.657,
+      "step": 25845
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.71289938635372,
+      "learning_rate": 5.153444542466368e-07,
+      "loss": 0.6399,
+      "step": 25850
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.43401069724985,
+      "learning_rate": 5.140483413288061e-07,
+      "loss": 0.6474,
+      "step": 25855
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.692708943604419,
+      "learning_rate": 5.127537720226555e-07,
+      "loss": 0.6567,
+      "step": 25860
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.929392791908258,
+      "learning_rate": 5.114607467736471e-07,
+      "loss": 0.6174,
+      "step": 25865
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.493020788611299,
+      "learning_rate": 5.101692660267077e-07,
+      "loss": 0.6331,
+      "step": 25870
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 7.050134354806047,
+      "learning_rate": 5.088793302262362e-07,
+      "loss": 0.6717,
+      "step": 25875
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.760627520220874,
+      "learning_rate": 5.075909398160983e-07,
+      "loss": 0.6459,
+      "step": 25880
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.579875926689988,
+      "learning_rate": 5.06304095239627e-07,
+      "loss": 0.6128,
+      "step": 25885
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.695488079068494,
+      "learning_rate": 5.050187969396248e-07,
+      "loss": 0.6409,
+      "step": 25890
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.584920910978887,
+      "learning_rate": 5.037350453583601e-07,
+      "loss": 0.6127,
+      "step": 25895
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.733672284793962,
+      "learning_rate": 5.024528409375728e-07,
+      "loss": 0.6261,
+      "step": 25900
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.334173651853007,
+      "learning_rate": 5.011721841184663e-07,
+      "loss": 0.6602,
+      "step": 25905
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.659865730931912,
+      "learning_rate": 4.998930753417153e-07,
+      "loss": 0.6366,
+      "step": 25910
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 7.149491262060584,
+      "learning_rate": 4.986155150474592e-07,
+      "loss": 0.6125,
+      "step": 25915
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.661274683439253,
+      "learning_rate": 4.973395036753054e-07,
+      "loss": 0.6447,
+      "step": 25920
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.369340660498729,
+      "learning_rate": 4.960650416643259e-07,
+      "loss": 0.6536,
+      "step": 25925
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.4477401440865405,
+      "learning_rate": 4.947921294530656e-07,
+      "loss": 0.6509,
+      "step": 25930
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.025337057836171,
+      "learning_rate": 4.935207674795295e-07,
+      "loss": 0.6458,
+      "step": 25935
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.64075589139611,
+      "learning_rate": 4.922509561811939e-07,
+      "loss": 0.6374,
+      "step": 25940
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.465561320407794,
+      "learning_rate": 4.909826959949988e-07,
+      "loss": 0.6422,
+      "step": 25945
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.385235783038693,
+      "learning_rate": 4.897159873573515e-07,
+      "loss": 0.6421,
+      "step": 25950
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.594086031534952,
+      "learning_rate": 4.884508307041241e-07,
+      "loss": 0.6374,
+      "step": 25955
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.258903294857566,
+      "learning_rate": 4.871872264706579e-07,
+      "loss": 0.6678,
+      "step": 25960
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.863058821194196,
+      "learning_rate": 4.859251750917559e-07,
+      "loss": 0.6825,
+      "step": 25965
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.665283377330753,
+      "learning_rate": 4.846646770016905e-07,
+      "loss": 0.6431,
+      "step": 25970
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.468217346387384,
+      "learning_rate": 4.834057326341973e-07,
+      "loss": 0.6687,
+      "step": 25975
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.913611862161178,
+      "learning_rate": 4.821483424224776e-07,
+      "loss": 0.6269,
+      "step": 25980
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.830245506259714,
+      "learning_rate": 4.808925067991977e-07,
+      "loss": 0.642,
+      "step": 25985
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.179273168502301,
+      "learning_rate": 4.796382261964905e-07,
+      "loss": 0.647,
+      "step": 25990
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.474430609458084,
+      "learning_rate": 4.78385501045951e-07,
+      "loss": 0.6752,
+      "step": 25995
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.195319875581904,
+      "learning_rate": 4.771343317786431e-07,
+      "loss": 0.6482,
+      "step": 26000
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.632248905230632,
+      "learning_rate": 4.75884718825092e-07,
+      "loss": 0.6448,
+      "step": 26005
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.697644503611095,
+      "learning_rate": 4.7463666261528816e-07,
+      "loss": 0.6415,
+      "step": 26010
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.47786641642741,
+      "learning_rate": 4.7339016357868586e-07,
+      "loss": 0.6067,
+      "step": 26015
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.537017336696449,
+      "learning_rate": 4.7214522214420464e-07,
+      "loss": 0.6452,
+      "step": 26020
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.8881464897238285,
+      "learning_rate": 4.7090183874022867e-07,
+      "loss": 0.6293,
+      "step": 26025
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 7.987028504731956,
+      "learning_rate": 4.696600137946028e-07,
+      "loss": 0.6627,
+      "step": 26030
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.634628482096437,
+      "learning_rate": 4.684197477346408e-07,
+      "loss": 0.6449,
+      "step": 26035
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.980914909416849,
+      "learning_rate": 4.6718104098711525e-07,
+      "loss": 0.6715,
+      "step": 26040
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.477828631881524,
+      "learning_rate": 4.65943893978264e-07,
+      "loss": 0.6841,
+      "step": 26045
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.962689410012361,
+      "learning_rate": 4.6470830713378714e-07,
+      "loss": 0.6577,
+      "step": 26050
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.410764725035844,
+      "learning_rate": 4.634742808788517e-07,
+      "loss": 0.6437,
+      "step": 26055
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.205768291835934,
+      "learning_rate": 4.62241815638082e-07,
+      "loss": 0.6492,
+      "step": 26060
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.529713580489901,
+      "learning_rate": 4.610109118355699e-07,
+      "loss": 0.6483,
+      "step": 26065
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.108805923799643,
+      "learning_rate": 4.597815698948688e-07,
+      "loss": 0.6324,
+      "step": 26070
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.6314997168050525,
+      "learning_rate": 4.585537902389925e-07,
+      "loss": 0.6332,
+      "step": 26075
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 7.184210025979383,
+      "learning_rate": 4.5732757329041866e-07,
+      "loss": 0.6728,
+      "step": 26080
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.714161295756465,
+      "learning_rate": 4.5610291947108866e-07,
+      "loss": 0.6529,
+      "step": 26085
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.667064450386645,
+      "learning_rate": 4.548798292024037e-07,
+      "loss": 0.6738,
+      "step": 26090
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 7.745480185868553,
+      "learning_rate": 4.536583029052294e-07,
+      "loss": 0.6351,
+      "step": 26095
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.460932447787701,
+      "learning_rate": 4.5243834099989006e-07,
+      "loss": 0.6804,
+      "step": 26100
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.212632081219248,
+      "learning_rate": 4.5121994390617484e-07,
+      "loss": 0.6317,
+      "step": 26105
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 8.231950682387714,
+      "learning_rate": 4.5000311204333123e-07,
+      "loss": 0.6345,
+      "step": 26110
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 6.870402774643485,
+      "learning_rate": 4.4878784583007207e-07,
+      "loss": 0.6527,
+      "step": 26115
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.89773131939057,
+      "learning_rate": 4.4757414568456724e-07,
+      "loss": 0.6368,
+      "step": 26120
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.53326405211866,
+      "learning_rate": 4.4636201202445164e-07,
+      "loss": 0.6441,
+      "step": 26125
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.791953087780344,
+      "learning_rate": 4.4515144526681875e-07,
+      "loss": 0.6438,
+      "step": 26130
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.825566433607306,
+      "learning_rate": 4.4394244582822264e-07,
+      "loss": 0.6512,
+      "step": 26135
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.235658704216288,
+      "learning_rate": 4.4273501412468e-07,
+      "loss": 0.6547,
+      "step": 26140
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.890768808912151,
+      "learning_rate": 4.4152915057166513e-07,
+      "loss": 0.6593,
+      "step": 26145
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.795158812209058,
+      "learning_rate": 4.403248555841166e-07,
+      "loss": 0.6529,
+      "step": 26150
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.131070693260541,
+      "learning_rate": 4.391221295764292e-07,
+      "loss": 0.6528,
+      "step": 26155
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.710697756328377,
+      "learning_rate": 4.379209729624617e-07,
+      "loss": 0.6503,
+      "step": 26160
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.6781511899293236,
+      "learning_rate": 4.367213861555308e-07,
+      "loss": 0.6672,
+      "step": 26165
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.517593966636261,
+      "learning_rate": 4.355233695684119e-07,
+      "loss": 0.6221,
+      "step": 26170
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.373524161316528,
+      "learning_rate": 4.343269236133413e-07,
+      "loss": 0.6879,
+      "step": 26175
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.720161138755838,
+      "learning_rate": 4.331320487020163e-07,
+      "loss": 0.6572,
+      "step": 26180
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.763630065649678,
+      "learning_rate": 4.319387452455903e-07,
+      "loss": 0.6046,
+      "step": 26185
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.422303645853079,
+      "learning_rate": 4.3074701365468043e-07,
+      "loss": 0.6208,
+      "step": 26190
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.433871515640049,
+      "learning_rate": 4.295568543393591e-07,
+      "loss": 0.6487,
+      "step": 26195
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.931653971987222,
+      "learning_rate": 4.283682677091583e-07,
+      "loss": 0.6361,
+      "step": 26200
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.890247256439932,
+      "learning_rate": 4.271812541730697e-07,
+      "loss": 0.7039,
+      "step": 26205
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.698518626266278,
+      "learning_rate": 4.2599581413954485e-07,
+      "loss": 0.6728,
+      "step": 26210
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.8057777852945,
+      "learning_rate": 4.2481194801649086e-07,
+      "loss": 0.6493,
+      "step": 26215
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.169528262794991,
+      "learning_rate": 4.236296562112768e-07,
+      "loss": 0.6523,
+      "step": 26220
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.521880829571063,
+      "learning_rate": 4.224489391307268e-07,
+      "loss": 0.6345,
+      "step": 26225
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.596943705221121,
+      "learning_rate": 4.212697971811247e-07,
+      "loss": 0.6282,
+      "step": 26230
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.40358921987457,
+      "learning_rate": 4.200922307682115e-07,
+      "loss": 0.6475,
+      "step": 26235
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.424249521185416,
+      "learning_rate": 4.1891624029718856e-07,
+      "loss": 0.6432,
+      "step": 26240
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.796678850595712,
+      "learning_rate": 4.1774182617271064e-07,
+      "loss": 0.6428,
+      "step": 26245
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.6661345801770056,
+      "learning_rate": 4.165689887988944e-07,
+      "loss": 0.6626,
+      "step": 26250
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.5806291067324105,
+      "learning_rate": 4.1539772857931203e-07,
+      "loss": 0.6459,
+      "step": 26255
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.5650564431500085,
+      "learning_rate": 4.142280459169923e-07,
+      "loss": 0.6337,
+      "step": 26260
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.73604227877278,
+      "learning_rate": 4.130599412144215e-07,
+      "loss": 0.6437,
+      "step": 26265
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.586531802102478,
+      "learning_rate": 4.118934148735437e-07,
+      "loss": 0.653,
+      "step": 26270
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.758452148639693,
+      "learning_rate": 4.107284672957601e-07,
+      "loss": 0.6437,
+      "step": 26275
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.571298067242407,
+      "learning_rate": 4.095650988819272e-07,
+      "loss": 0.6772,
+      "step": 26280
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.556138379592204,
+      "learning_rate": 4.084033100323598e-07,
+      "loss": 0.6234,
+      "step": 26285
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.604219528719242,
+      "learning_rate": 4.072431011468286e-07,
+      "loss": 0.6023,
+      "step": 26290
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.610685075191588,
+      "learning_rate": 4.0608447262455886e-07,
+      "loss": 0.6254,
+      "step": 26295
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.604971165395091,
+      "learning_rate": 4.049274248642343e-07,
+      "loss": 0.6382,
+      "step": 26300
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.805269762146627,
+      "learning_rate": 4.037719582639943e-07,
+      "loss": 0.6471,
+      "step": 26305
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.928338049264352,
+      "learning_rate": 4.02618073221433e-07,
+      "loss": 0.6322,
+      "step": 26310
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.635905096298693,
+      "learning_rate": 4.014657701336028e-07,
+      "loss": 0.6773,
+      "step": 26315
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.460898503134962,
+      "learning_rate": 4.003150493970087e-07,
+      "loss": 0.6585,
+      "step": 26320
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.856638872130659,
+      "learning_rate": 3.9916591140761294e-07,
+      "loss": 0.6428,
+      "step": 26325
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.54106911842793,
+      "learning_rate": 3.98018356560832e-07,
+      "loss": 0.6197,
+      "step": 26330
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.777343218581925,
+      "learning_rate": 3.9687238525153994e-07,
+      "loss": 0.6599,
+      "step": 26335
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.6748165213887125,
+      "learning_rate": 3.9572799787406245e-07,
+      "loss": 0.6462,
+      "step": 26340
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.5916775692590726,
+      "learning_rate": 3.945851948221846e-07,
+      "loss": 0.605,
+      "step": 26345
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.306976299073246,
+      "learning_rate": 3.9344397648914234e-07,
+      "loss": 0.6697,
+      "step": 26350
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.195639071441672,
+      "learning_rate": 3.9230434326762823e-07,
+      "loss": 0.6419,
+      "step": 26355
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.581093545879182,
+      "learning_rate": 3.9116629554978747e-07,
+      "loss": 0.6413,
+      "step": 26360
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.052458529720565,
+      "learning_rate": 3.9002983372722345e-07,
+      "loss": 0.6623,
+      "step": 26365
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.697125648228998,
+      "learning_rate": 3.888949581909901e-07,
+      "loss": 0.6843,
+      "step": 26370
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.660423087556206,
+      "learning_rate": 3.877616693315978e-07,
+      "loss": 0.6547,
+      "step": 26375
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.394527486808878,
+      "learning_rate": 3.8662996753901063e-07,
+      "loss": 0.6739,
+      "step": 26380
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.587506714746173,
+      "learning_rate": 3.8549985320264496e-07,
+      "loss": 0.6577,
+      "step": 26385
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.525985421444638,
+      "learning_rate": 3.8437132671137245e-07,
+      "loss": 0.6433,
+      "step": 26390
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.308533008713368,
+      "learning_rate": 3.8324438845351697e-07,
+      "loss": 0.6337,
+      "step": 26395
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.600283096400497,
+      "learning_rate": 3.8211903881685887e-07,
+      "loss": 0.6838,
+      "step": 26400
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.629575518720822,
+      "learning_rate": 3.8099527818862837e-07,
+      "loss": 0.6116,
+      "step": 26405
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.946398314146315,
+      "learning_rate": 3.7987310695551115e-07,
+      "loss": 0.6687,
+      "step": 26410
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.699259761441962,
+      "learning_rate": 3.787525255036456e-07,
+      "loss": 0.6748,
+      "step": 26415
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.4768450040424606,
+      "learning_rate": 3.7763353421862215e-07,
+      "loss": 0.6404,
+      "step": 26420
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.489871861176078,
+      "learning_rate": 3.7651613348548386e-07,
+      "loss": 0.6138,
+      "step": 26425
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.413401018566667,
+      "learning_rate": 3.7540032368872937e-07,
+      "loss": 0.6692,
+      "step": 26430
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.8938714822113205,
+      "learning_rate": 3.742861052123048e-07,
+      "loss": 0.5991,
+      "step": 26435
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.467851521742816,
+      "learning_rate": 3.7317347843961514e-07,
+      "loss": 0.7053,
+      "step": 26440
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.564001385017399,
+      "learning_rate": 3.7206244375351197e-07,
+      "loss": 0.6543,
+      "step": 26445
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.099093322914692,
+      "learning_rate": 3.7095300153630167e-07,
+      "loss": 0.6714,
+      "step": 26450
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.413616723694407,
+      "learning_rate": 3.6984515216974104e-07,
+      "loss": 0.6774,
+      "step": 26455
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.746017427039399,
+      "learning_rate": 3.687388960350424e-07,
+      "loss": 0.6702,
+      "step": 26460
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.302096269813245,
+      "learning_rate": 3.6763423351286576e-07,
+      "loss": 0.6302,
+      "step": 26465
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.413759822215279,
+      "learning_rate": 3.6653116498332587e-07,
+      "loss": 0.6451,
+      "step": 26470
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.924375370439851,
+      "learning_rate": 3.6542969082598576e-07,
+      "loss": 0.6791,
+      "step": 26475
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.551790902336037,
+      "learning_rate": 3.6432981141986347e-07,
+      "loss": 0.6193,
+      "step": 26480
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.447720030507796,
+      "learning_rate": 3.632315271434239e-07,
+      "loss": 0.6189,
+      "step": 26485
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.80993642009209,
+      "learning_rate": 3.6213483837458873e-07,
+      "loss": 0.6334,
+      "step": 26490
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.8623738488613215,
+      "learning_rate": 3.61039745490725e-07,
+      "loss": 0.6147,
+      "step": 26495
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.022644888678823,
+      "learning_rate": 3.5994624886865445e-07,
+      "loss": 0.697,
+      "step": 26500
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 12.240058964168647,
+      "learning_rate": 3.588543488846485e-07,
+      "loss": 0.6848,
+      "step": 26505
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.456351689321203,
+      "learning_rate": 3.5776404591442824e-07,
+      "loss": 0.633,
+      "step": 26510
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.244832953200636,
+      "learning_rate": 3.5667534033316466e-07,
+      "loss": 0.6951,
+      "step": 26515
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.539647528153177,
+      "learning_rate": 3.5558823251548304e-07,
+      "loss": 0.6083,
+      "step": 26520
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.573046024457291,
+      "learning_rate": 3.545027228354542e-07,
+      "loss": 0.6366,
+      "step": 26525
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.955392603809859,
+      "learning_rate": 3.534188116666004e-07,
+      "loss": 0.681,
+      "step": 26530
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.170833130307659,
+      "learning_rate": 3.523364993818978e-07,
+      "loss": 0.6342,
+      "step": 26535
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.891291415673024,
+      "learning_rate": 3.512557863537647e-07,
+      "loss": 0.6489,
+      "step": 26540
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.888406857967937,
+      "learning_rate": 3.5017667295407676e-07,
+      "loss": 0.6544,
+      "step": 26545
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.449293637678464,
+      "learning_rate": 3.490991595541532e-07,
+      "loss": 0.6436,
+      "step": 26550
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.652288939825329,
+      "learning_rate": 3.480232465247679e-07,
+      "loss": 0.6586,
+      "step": 26555
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 6.060881608415734,
+      "learning_rate": 3.469489342361393e-07,
+      "loss": 0.6442,
+      "step": 26560
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.798399988610016,
+      "learning_rate": 3.458762230579388e-07,
+      "loss": 0.6235,
+      "step": 26565
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.838544020817497,
+      "learning_rate": 3.44805113359285e-07,
+      "loss": 0.652,
+      "step": 26570
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.5975411064486185,
+      "learning_rate": 3.4373560550874543e-07,
+      "loss": 0.6477,
+      "step": 26575
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.549271557696654,
+      "learning_rate": 3.426676998743361e-07,
+      "loss": 0.6937,
+      "step": 26580
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.760921452630401,
+      "learning_rate": 3.416013968235238e-07,
+      "loss": 0.6432,
+      "step": 26585
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.4813525599711355,
+      "learning_rate": 3.4053669672322096e-07,
+      "loss": 0.6855,
+      "step": 26590
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.671748925862381,
+      "learning_rate": 3.3947359993979077e-07,
+      "loss": 0.6466,
+      "step": 26595
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.543481218279123,
+      "learning_rate": 3.3841210683904393e-07,
+      "loss": 0.6411,
+      "step": 26600
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.72674022959049,
+      "learning_rate": 3.3735221778623815e-07,
+      "loss": 0.6554,
+      "step": 26605
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.098608945175818,
+      "learning_rate": 3.362939331460807e-07,
+      "loss": 0.6626,
+      "step": 26610
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.787862398042563,
+      "learning_rate": 3.352372532827275e-07,
+      "loss": 0.6539,
+      "step": 26615
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.632119559373765,
+      "learning_rate": 3.341821785597787e-07,
+      "loss": 0.6774,
+      "step": 26620
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.783547536689469,
+      "learning_rate": 3.3312870934028685e-07,
+      "loss": 0.6352,
+      "step": 26625
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.297647085341283,
+      "learning_rate": 3.3207684598674906e-07,
+      "loss": 0.674,
+      "step": 26630
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.711494616548537,
+      "learning_rate": 3.310265888611097e-07,
+      "loss": 0.6873,
+      "step": 26635
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.4983153795010775,
+      "learning_rate": 3.2997793832476146e-07,
+      "loss": 0.6039,
+      "step": 26640
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.69938871439085,
+      "learning_rate": 3.2893089473854447e-07,
+      "loss": 0.6504,
+      "step": 26645
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.5622571590941225,
+      "learning_rate": 3.2788545846274553e-07,
+      "loss": 0.6531,
+      "step": 26650
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.100048240343806,
+      "learning_rate": 3.2684162985709757e-07,
+      "loss": 0.6577,
+      "step": 26655
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.270853011278117,
+      "learning_rate": 3.2579940928078204e-07,
+      "loss": 0.6192,
+      "step": 26660
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.928062910367746,
+      "learning_rate": 3.247587970924243e-07,
+      "loss": 0.6612,
+      "step": 26665
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 8.925526303808986,
+      "learning_rate": 3.237197936501002e-07,
+      "loss": 0.6387,
+      "step": 26670
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.50319252883846,
+      "learning_rate": 3.226823993113276e-07,
+      "loss": 0.6456,
+      "step": 26675
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.515467389106821,
+      "learning_rate": 3.216466144330749e-07,
+      "loss": 0.6674,
+      "step": 26680
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.683024789220738,
+      "learning_rate": 3.2061243937175304e-07,
+      "loss": 0.6754,
+      "step": 26685
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.525353924162903,
+      "learning_rate": 3.195798744832235e-07,
+      "loss": 0.648,
+      "step": 26690
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.236326022567566,
+      "learning_rate": 3.185489201227865e-07,
+      "loss": 0.6259,
+      "step": 26695
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.172427585267906,
+      "learning_rate": 3.1751957664519606e-07,
+      "loss": 0.6693,
+      "step": 26700
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.397112012371753,
+      "learning_rate": 3.164918444046461e-07,
+      "loss": 0.6263,
+      "step": 26705
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.769481924908539,
+      "learning_rate": 3.154657237547798e-07,
+      "loss": 0.6951,
+      "step": 26710
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.80030276412543,
+      "learning_rate": 3.144412150486831e-07,
+      "loss": 0.6363,
+      "step": 26715
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.923919053517498,
+      "learning_rate": 3.134183186388906e-07,
+      "loss": 0.7015,
+      "step": 26720
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.634432469691726,
+      "learning_rate": 3.1239703487737696e-07,
+      "loss": 0.6231,
+      "step": 26725
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.6519209114755435,
+      "learning_rate": 3.1137736411556705e-07,
+      "loss": 0.6219,
+      "step": 26730
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.644362142363348,
+      "learning_rate": 3.103593067043276e-07,
+      "loss": 0.6476,
+      "step": 26735
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.0411318413340815,
+      "learning_rate": 3.093428629939721e-07,
+      "loss": 0.6551,
+      "step": 26740
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.615899235277585,
+      "learning_rate": 3.0832803333425643e-07,
+      "loss": 0.6275,
+      "step": 26745
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.031842335224812,
+      "learning_rate": 3.0731481807438513e-07,
+      "loss": 0.652,
+      "step": 26750
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.7918819848872864,
+      "learning_rate": 3.063032175630015e-07,
+      "loss": 0.6294,
+      "step": 26755
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.49577226983595,
+      "learning_rate": 3.052932321481983e-07,
+      "loss": 0.6443,
+      "step": 26760
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.230227096478425,
+      "learning_rate": 3.0428486217750907e-07,
+      "loss": 0.6361,
+      "step": 26765
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.67422220177396,
+      "learning_rate": 3.032781079979147e-07,
+      "loss": 0.6537,
+      "step": 26770
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.363855352539699,
+      "learning_rate": 3.0227296995583797e-07,
+      "loss": 0.6151,
+      "step": 26775
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 7.584201273966142,
+      "learning_rate": 3.012694483971446e-07,
+      "loss": 0.6498,
+      "step": 26780
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.896712764016411,
+      "learning_rate": 3.0026754366714703e-07,
+      "loss": 0.6589,
+      "step": 26785
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.134313710305708,
+      "learning_rate": 2.9926725611059747e-07,
+      "loss": 0.6656,
+      "step": 26790
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.870412747383471,
+      "learning_rate": 2.982685860716966e-07,
+      "loss": 0.6446,
+      "step": 26795
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.89087219092221,
+      "learning_rate": 2.9727153389408347e-07,
+      "loss": 0.6327,
+      "step": 26800
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.952435309248969,
+      "learning_rate": 2.962760999208453e-07,
+      "loss": 0.665,
+      "step": 26805
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.448454332529152,
+      "learning_rate": 2.95282284494508e-07,
+      "loss": 0.6392,
+      "step": 26810
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.802846152453651,
+      "learning_rate": 2.9429008795704317e-07,
+      "loss": 0.6402,
+      "step": 26815
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.889305219959239,
+      "learning_rate": 2.932995106498637e-07,
+      "loss": 0.6156,
+      "step": 26820
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.515546156661981,
+      "learning_rate": 2.9231055291382816e-07,
+      "loss": 0.6196,
+      "step": 26825
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.409786442384293,
+      "learning_rate": 2.9132321508923424e-07,
+      "loss": 0.615,
+      "step": 26830
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.82496443011332,
+      "learning_rate": 2.903374975158257e-07,
+      "loss": 0.666,
+      "step": 26835
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.321518340014227,
+      "learning_rate": 2.893534005327858e-07,
+      "loss": 0.6391,
+      "step": 26840
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.6860386188295795,
+      "learning_rate": 2.883709244787414e-07,
+      "loss": 0.6422,
+      "step": 26845
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.54933372316617,
+      "learning_rate": 2.8739006969176087e-07,
+      "loss": 0.6227,
+      "step": 26850
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.550003769397981,
+      "learning_rate": 2.8641083650935765e-07,
+      "loss": 0.6237,
+      "step": 26855
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.100015574627647,
+      "learning_rate": 2.854332252684827e-07,
+      "loss": 0.6396,
+      "step": 26860
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.526484394100703,
+      "learning_rate": 2.844572363055326e-07,
+      "loss": 0.6314,
+      "step": 26865
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.262718390081325,
+      "learning_rate": 2.8348286995634247e-07,
+      "loss": 0.6888,
+      "step": 26870
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.466091994999804,
+      "learning_rate": 2.8251012655619417e-07,
+      "loss": 0.6439,
+      "step": 26875
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.722071904748594,
+      "learning_rate": 2.815390064398038e-07,
+      "loss": 0.6373,
+      "step": 26880
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.302036926701708,
+      "learning_rate": 2.8056950994133524e-07,
+      "loss": 0.6513,
+      "step": 26885
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.580234389013792,
+      "learning_rate": 2.796016373943905e-07,
+      "loss": 0.6607,
+      "step": 26890
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.854842515682165,
+      "learning_rate": 2.786353891320143e-07,
+      "loss": 0.6205,
+      "step": 26895
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.375572061695649,
+      "learning_rate": 2.7767076548669004e-07,
+      "loss": 0.6762,
+      "step": 26900
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.422668495277672,
+      "learning_rate": 2.7670776679034727e-07,
+      "loss": 0.6533,
+      "step": 26905
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 6.482983874201208,
+      "learning_rate": 2.757463933743493e-07,
+      "loss": 0.6156,
+      "step": 26910
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.274977240043379,
+      "learning_rate": 2.747866455695053e-07,
+      "loss": 0.6541,
+      "step": 26915
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.630084332091941,
+      "learning_rate": 2.7382852370606337e-07,
+      "loss": 0.6466,
+      "step": 26920
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.469241103432817,
+      "learning_rate": 2.7287202811371203e-07,
+      "loss": 0.6348,
+      "step": 26925
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.203696119549017,
+      "learning_rate": 2.7191715912158187e-07,
+      "loss": 0.6029,
+      "step": 26930
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.2311056726551985,
+      "learning_rate": 2.7096391705824067e-07,
+      "loss": 0.6521,
+      "step": 26935
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.831648020213548,
+      "learning_rate": 2.700123022516998e-07,
+      "loss": 0.6684,
+      "step": 26940
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.001125628731334,
+      "learning_rate": 2.6906231502940637e-07,
+      "loss": 0.6459,
+      "step": 26945
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.581463486268398,
+      "learning_rate": 2.6811395571825327e-07,
+      "loss": 0.6374,
+      "step": 26950
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.827504175309236,
+      "learning_rate": 2.671672246445678e-07,
+      "loss": 0.6118,
+      "step": 26955
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.577015648069526,
+      "learning_rate": 2.662221221341199e-07,
+      "loss": 0.6472,
+      "step": 26960
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.8216784748881185,
+      "learning_rate": 2.652786485121195e-07,
+      "loss": 0.6551,
+      "step": 26965
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.206714745623948,
+      "learning_rate": 2.643368041032135e-07,
+      "loss": 0.6614,
+      "step": 26970
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.9204799966670425,
+      "learning_rate": 2.6339658923148934e-07,
+      "loss": 0.65,
+      "step": 26975
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.502378422522532,
+      "learning_rate": 2.624580042204755e-07,
+      "loss": 0.5966,
+      "step": 26980
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.841075778399853,
+      "learning_rate": 2.615210493931369e-07,
+      "loss": 0.6508,
+      "step": 26985
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.457993744591424,
+      "learning_rate": 2.605857250718802e-07,
+      "loss": 0.6604,
+      "step": 26990
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.407642117421727,
+      "learning_rate": 2.596520315785489e-07,
+      "loss": 0.6132,
+      "step": 26995
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.586783087603529,
+      "learning_rate": 2.5871996923442555e-07,
+      "loss": 0.6758,
+      "step": 27000
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.3373244259396495,
+      "learning_rate": 2.5778953836023247e-07,
+      "loss": 0.6464,
+      "step": 27005
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.6943349890765145,
+      "learning_rate": 2.5686073927612964e-07,
+      "loss": 0.6582,
+      "step": 27010
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.550383608738202,
+      "learning_rate": 2.5593357230171645e-07,
+      "loss": 0.6317,
+      "step": 27015
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.585911485678205,
+      "learning_rate": 2.550080377560299e-07,
+      "loss": 0.6652,
+      "step": 27020
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.611901149160534,
+      "learning_rate": 2.540841359575458e-07,
+      "loss": 0.6434,
+      "step": 27025
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 6.226867509992674,
+      "learning_rate": 2.5316186722417756e-07,
+      "loss": 0.6384,
+      "step": 27030
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.889591104976683,
+      "learning_rate": 2.5224123187327753e-07,
+      "loss": 0.6282,
+      "step": 27035
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.138212027862173,
+      "learning_rate": 2.5132223022163395e-07,
+      "loss": 0.6599,
+      "step": 27040
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.678815642637734,
+      "learning_rate": 2.5040486258547603e-07,
+      "loss": 0.6327,
+      "step": 27045
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.569598167796238,
+      "learning_rate": 2.4948912928046797e-07,
+      "loss": 0.6298,
+      "step": 27050
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.7706974750619535,
+      "learning_rate": 2.4857503062171386e-07,
+      "loss": 0.6319,
+      "step": 27055
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 6.3280113977072965,
+      "learning_rate": 2.4766256692375367e-07,
+      "loss": 0.6304,
+      "step": 27060
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.076679797067638,
+      "learning_rate": 2.467517385005652e-07,
+      "loss": 0.6282,
+      "step": 27065
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.1371346360423,
+      "learning_rate": 2.4584254566556276e-07,
+      "loss": 0.618,
+      "step": 27070
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.86654090029525,
+      "learning_rate": 2.4493498873160046e-07,
+      "loss": 0.6152,
+      "step": 27075
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.641826973860905,
+      "learning_rate": 2.440290680109664e-07,
+      "loss": 0.6607,
+      "step": 27080
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.154592587557687,
+      "learning_rate": 2.4312478381538775e-07,
+      "loss": 0.6139,
+      "step": 27085
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.593874514370757,
+      "learning_rate": 2.422221364560279e-07,
+      "loss": 0.6748,
+      "step": 27090
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.655599640633073,
+      "learning_rate": 2.413211262434867e-07,
+      "loss": 0.6744,
+      "step": 27095
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.777183373107945,
+      "learning_rate": 2.404217534878006e-07,
+      "loss": 0.6518,
+      "step": 27100
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.5609775207126,
+      "learning_rate": 2.395240184984432e-07,
+      "loss": 0.6391,
+      "step": 27105
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.5354817359318975,
+      "learning_rate": 2.3862792158432403e-07,
+      "loss": 0.6396,
+      "step": 27110
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.840120970779168,
+      "learning_rate": 2.377334630537903e-07,
+      "loss": 0.6534,
+      "step": 27115
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.754107720175892,
+      "learning_rate": 2.3684064321462309e-07,
+      "loss": 0.6446,
+      "step": 27120
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.446069918589517,
+      "learning_rate": 2.3594946237404104e-07,
+      "loss": 0.6295,
+      "step": 27125
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.89894032512749,
+      "learning_rate": 2.3505992083869832e-07,
+      "loss": 0.641,
+      "step": 27130
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.866888053961888,
+      "learning_rate": 2.3417201891468677e-07,
+      "loss": 0.6797,
+      "step": 27135
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.704806075752119,
+      "learning_rate": 2.332857569075303e-07,
+      "loss": 0.6402,
+      "step": 27140
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.795521106834384,
+      "learning_rate": 2.3240113512219332e-07,
+      "loss": 0.6492,
+      "step": 27145
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.505870048445661,
+      "learning_rate": 2.3151815386307175e-07,
+      "loss": 0.6641,
+      "step": 27150
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.294647387657272,
+      "learning_rate": 2.3063681343399923e-07,
+      "loss": 0.6221,
+      "step": 27155
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.379296908401169,
+      "learning_rate": 2.297571141382443e-07,
+      "loss": 0.6102,
+      "step": 27160
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 6.542098819508065,
+      "learning_rate": 2.2887905627850926e-07,
+      "loss": 0.6615,
+      "step": 27165
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.604680790205558,
+      "learning_rate": 2.2800264015693464e-07,
+      "loss": 0.6596,
+      "step": 27170
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.17649037933092,
+      "learning_rate": 2.2712786607509308e-07,
+      "loss": 0.624,
+      "step": 27175
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.553300967953392,
+      "learning_rate": 2.262547343339949e-07,
+      "loss": 0.6386,
+      "step": 27180
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.575406320081005,
+      "learning_rate": 2.2538324523408318e-07,
+      "loss": 0.6476,
+      "step": 27185
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.569759907083859,
+      "learning_rate": 2.2451339907523684e-07,
+      "loss": 0.6641,
+      "step": 27190
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.901676613919331,
+      "learning_rate": 2.236451961567676e-07,
+      "loss": 0.6235,
+      "step": 27195
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.735981562262451,
+      "learning_rate": 2.2277863677742538e-07,
+      "loss": 0.635,
+      "step": 27200
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.4045401013505705,
+      "learning_rate": 2.2191372123539057e-07,
+      "loss": 0.622,
+      "step": 27205
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.5875845615749595,
+      "learning_rate": 2.2105044982828173e-07,
+      "loss": 0.6534,
+      "step": 27210
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.110692783266679,
+      "learning_rate": 2.2018882285314848e-07,
+      "loss": 0.6655,
+      "step": 27215
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.163729086764986,
+      "learning_rate": 2.1932884060647585e-07,
+      "loss": 0.6263,
+      "step": 27220
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.04769979373338,
+      "learning_rate": 2.1847050338418275e-07,
+      "loss": 0.6169,
+      "step": 27225
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.0579975466535645,
+      "learning_rate": 2.1761381148162286e-07,
+      "loss": 0.6224,
+      "step": 27230
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.477945345762049,
+      "learning_rate": 2.1675876519358264e-07,
+      "loss": 0.631,
+      "step": 27235
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.502067958512376,
+      "learning_rate": 2.159053648142828e-07,
+      "loss": 0.6526,
+      "step": 27240
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.645772831828346,
+      "learning_rate": 2.1505361063737795e-07,
+      "loss": 0.6892,
+      "step": 27245
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.7562112410395585,
+      "learning_rate": 2.1420350295595527e-07,
+      "loss": 0.6234,
+      "step": 27250
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.715893602665484,
+      "learning_rate": 2.1335504206253577e-07,
+      "loss": 0.6438,
+      "step": 27255
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.788625339216968,
+      "learning_rate": 2.1250822824907535e-07,
+      "loss": 0.6887,
+      "step": 27260
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.742321681343637,
+      "learning_rate": 2.116630618069604e-07,
+      "loss": 0.6219,
+      "step": 27265
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.879362925821528,
+      "learning_rate": 2.108195430270127e-07,
+      "loss": 0.6526,
+      "step": 27270
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.045335088378256,
+      "learning_rate": 2.0997767219948616e-07,
+      "loss": 0.6327,
+      "step": 27275
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.009295957562433,
+      "learning_rate": 2.09137449614068e-07,
+      "loss": 0.6294,
+      "step": 27280
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.434427769144761,
+      "learning_rate": 2.082988755598764e-07,
+      "loss": 0.6671,
+      "step": 27285
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.990377036828536,
+      "learning_rate": 2.0746195032546657e-07,
+      "loss": 0.6371,
+      "step": 27290
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.797263478041765,
+      "learning_rate": 2.0662667419882155e-07,
+      "loss": 0.6445,
+      "step": 27295
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.001224407816903,
+      "learning_rate": 2.0579304746735917e-07,
+      "loss": 0.6829,
+      "step": 27300
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.193277092738359,
+      "learning_rate": 2.049610704179311e-07,
+      "loss": 0.6176,
+      "step": 27305
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.041487298048559,
+      "learning_rate": 2.0413074333681893e-07,
+      "loss": 0.6126,
+      "step": 27310
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 7.129523974749876,
+      "learning_rate": 2.0330206650973683e-07,
+      "loss": 0.6541,
+      "step": 27315
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.825415696213447,
+      "learning_rate": 2.024750402218323e-07,
+      "loss": 0.6706,
+      "step": 27320
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 6.6591871746914695,
+      "learning_rate": 2.0164966475768432e-07,
+      "loss": 0.6622,
+      "step": 27325
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 6.145263719644995,
+      "learning_rate": 2.0082594040130353e-07,
+      "loss": 0.635,
+      "step": 27330
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.636647280013002,
+      "learning_rate": 2.000038674361332e-07,
+      "loss": 0.6217,
+      "step": 27335
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.4739024748437695,
+      "learning_rate": 1.9918344614504814e-07,
+      "loss": 0.6514,
+      "step": 27340
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.0127000776778585,
+      "learning_rate": 1.9836467681035365e-07,
+      "loss": 0.6368,
+      "step": 27345
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.906114484345132,
+      "learning_rate": 1.975475597137866e-07,
+      "loss": 0.6439,
+      "step": 27350
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.916876018127332,
+      "learning_rate": 1.967320951365187e-07,
+      "loss": 0.6158,
+      "step": 27355
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.487110739372675,
+      "learning_rate": 1.959182833591483e-07,
+      "loss": 0.6318,
+      "step": 27360
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.636937516626221,
+      "learning_rate": 1.951061246617092e-07,
+      "loss": 0.6577,
+      "step": 27365
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.932916527673636,
+      "learning_rate": 1.942956193236628e-07,
+      "loss": 0.6169,
+      "step": 27370
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.89725898238156,
+      "learning_rate": 1.934867676239044e-07,
+      "loss": 0.6313,
+      "step": 27375
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.572865175199781,
+      "learning_rate": 1.926795698407574e-07,
+      "loss": 0.6603,
+      "step": 27380
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.9158419371051005,
+      "learning_rate": 1.918740262519797e-07,
+      "loss": 0.625,
+      "step": 27385
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.900828600336452,
+      "learning_rate": 1.9107013713475674e-07,
+      "loss": 0.6687,
+      "step": 27390
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.230901880339584,
+      "learning_rate": 1.9026790276570728e-07,
+      "loss": 0.6119,
+      "step": 27395
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.609157007664113,
+      "learning_rate": 1.8946732342087825e-07,
+      "loss": 0.6388,
+      "step": 27400
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.811633485820632,
+      "learning_rate": 1.8866839937574876e-07,
+      "loss": 0.6433,
+      "step": 27405
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.5620610799291015,
+      "learning_rate": 1.8787113090522723e-07,
+      "loss": 0.611,
+      "step": 27410
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.112562968547693,
+      "learning_rate": 1.8707551828365368e-07,
+      "loss": 0.6381,
+      "step": 27415
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.52468498356504,
+      "learning_rate": 1.8628156178479794e-07,
+      "loss": 0.6312,
+      "step": 27420
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.258889515886162,
+      "learning_rate": 1.854892616818582e-07,
+      "loss": 0.6233,
+      "step": 27425
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 6.628910817758399,
+      "learning_rate": 1.8469861824746583e-07,
+      "loss": 0.651,
+      "step": 27430
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.163731183357664,
+      "learning_rate": 1.8390963175367983e-07,
+      "loss": 0.6219,
+      "step": 27435
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.51936940513308,
+      "learning_rate": 1.831223024719897e-07,
+      "loss": 0.6593,
+      "step": 27440
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.8868413077096315,
+      "learning_rate": 1.823366306733143e-07,
+      "loss": 0.6509,
+      "step": 27445
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.137439186382453,
+      "learning_rate": 1.8155261662800295e-07,
+      "loss": 0.6079,
+      "step": 27450
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 6.193024837189629,
+      "learning_rate": 1.8077026060583369e-07,
+      "loss": 0.6733,
+      "step": 27455
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.470445592225861,
+      "learning_rate": 1.7998956287601565e-07,
+      "loss": 0.6343,
+      "step": 27460
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.9839207959687695,
+      "learning_rate": 1.7921052370718505e-07,
+      "loss": 0.6557,
+      "step": 27465
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.966319709791478,
+      "learning_rate": 1.7843314336740913e-07,
+      "loss": 0.6608,
+      "step": 27470
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.612224889221658,
+      "learning_rate": 1.776574221241828e-07,
+      "loss": 0.6375,
+      "step": 27475
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.1061979664003685,
+      "learning_rate": 1.7688336024443197e-07,
+      "loss": 0.6297,
+      "step": 27480
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.620897017042376,
+      "learning_rate": 1.7611095799450973e-07,
+      "loss": 0.6672,
+      "step": 27485
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.325307617224155,
+      "learning_rate": 1.753402156402001e-07,
+      "loss": 0.6339,
+      "step": 27490
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.107182911004926,
+      "learning_rate": 1.7457113344671484e-07,
+      "loss": 0.6398,
+      "step": 27495
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 6.10444693123381,
+      "learning_rate": 1.7380371167869282e-07,
+      "loss": 0.6638,
+      "step": 27500
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.388963100605161,
+      "learning_rate": 1.730379506002039e-07,
+      "loss": 0.6182,
+      "step": 27505
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.527852561965197,
+      "learning_rate": 1.7227385047474676e-07,
+      "loss": 0.6453,
+      "step": 27510
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.947131925278648,
+      "learning_rate": 1.7151141156524554e-07,
+      "loss": 0.6488,
+      "step": 27515
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.888298137702055,
+      "learning_rate": 1.70750634134057e-07,
+      "loss": 0.6201,
+      "step": 27520
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.068734448012588,
+      "learning_rate": 1.6999151844296237e-07,
+      "loss": 0.6648,
+      "step": 27525
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.719138525994361,
+      "learning_rate": 1.6923406475317316e-07,
+      "loss": 0.6649,
+      "step": 27530
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.014641777763894,
+      "learning_rate": 1.684782733253276e-07,
+      "loss": 0.645,
+      "step": 27535
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.405259394411325,
+      "learning_rate": 1.6772414441949436e-07,
+      "loss": 0.5895,
+      "step": 27540
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.434647710812157,
+      "learning_rate": 1.6697167829516747e-07,
+      "loss": 0.6287,
+      "step": 27545
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.570729319226043,
+      "learning_rate": 1.6622087521126983e-07,
+      "loss": 0.6165,
+      "step": 27550
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.708876473272213,
+      "learning_rate": 1.6547173542615258e-07,
+      "loss": 0.6634,
+      "step": 27555
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.443803641000985,
+      "learning_rate": 1.6472425919759338e-07,
+      "loss": 0.6735,
+      "step": 27560
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.503813881852303,
+      "learning_rate": 1.6397844678279872e-07,
+      "loss": 0.6524,
+      "step": 27565
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.262393504748039,
+      "learning_rate": 1.632342984384011e-07,
+      "loss": 0.6522,
+      "step": 27570
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.533398944217278,
+      "learning_rate": 1.6249181442046235e-07,
+      "loss": 0.6556,
+      "step": 27575
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.584461282149826,
+      "learning_rate": 1.6175099498446866e-07,
+      "loss": 0.6376,
+      "step": 27580
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.740322548336197,
+      "learning_rate": 1.610118403853378e-07,
+      "loss": 0.659,
+      "step": 27585
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.585743047346483,
+      "learning_rate": 1.6027435087741072e-07,
+      "loss": 0.6057,
+      "step": 27590
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 6.2909688966884225,
+      "learning_rate": 1.5953852671445668e-07,
+      "loss": 0.654,
+      "step": 27595
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.97205188031535,
+      "learning_rate": 1.5880436814967258e-07,
+      "loss": 0.627,
+      "step": 27600
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.519227048096504,
+      "learning_rate": 1.5807187543568136e-07,
+      "loss": 0.6219,
+      "step": 27605
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.361661472488166,
+      "learning_rate": 1.5734104882453305e-07,
+      "loss": 0.6861,
+      "step": 27610
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.8666409971276074,
+      "learning_rate": 1.5661188856770537e-07,
+      "loss": 0.6247,
+      "step": 27615
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.717116455709111,
+      "learning_rate": 1.558843949161004e-07,
+      "loss": 0.6532,
+      "step": 27620
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.630714814407945,
+      "learning_rate": 1.5515856812004849e-07,
+      "loss": 0.6157,
+      "step": 27625
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.28299970158232,
+      "learning_rate": 1.544344084293059e-07,
+      "loss": 0.6272,
+      "step": 27630
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 6.239301778257726,
+      "learning_rate": 1.5371191609305502e-07,
+      "loss": 0.6476,
+      "step": 27635
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.132571940847922,
+      "learning_rate": 1.5299109135990531e-07,
+      "loss": 0.6757,
+      "step": 27640
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 6.163828640077864,
+      "learning_rate": 1.5227193447789167e-07,
+      "loss": 0.646,
+      "step": 27645
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.623353690086022,
+      "learning_rate": 1.5155444569447565e-07,
+      "loss": 0.6297,
+      "step": 27650
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.618225671423743,
+      "learning_rate": 1.5083862525654413e-07,
+      "loss": 0.6644,
+      "step": 27655
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.8193840512997435,
+      "learning_rate": 1.5012447341040903e-07,
+      "loss": 0.6378,
+      "step": 27660
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.872251860166495,
+      "learning_rate": 1.4941199040181152e-07,
+      "loss": 0.6605,
+      "step": 27665
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.521733108340862,
+      "learning_rate": 1.4870117647591441e-07,
+      "loss": 0.6715,
+      "step": 27670
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.034286716147694,
+      "learning_rate": 1.4799203187730982e-07,
+      "loss": 0.6315,
+      "step": 27675
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.99830347398244,
+      "learning_rate": 1.4728455685001253e-07,
+      "loss": 0.6192,
+      "step": 27680
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.923759561043896,
+      "learning_rate": 1.4657875163746448e-07,
+      "loss": 0.6361,
+      "step": 27685
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.4891986428328545,
+      "learning_rate": 1.4587461648253253e-07,
+      "loss": 0.631,
+      "step": 27690
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.620859160067768,
+      "learning_rate": 1.4517215162750841e-07,
+      "loss": 0.626,
+      "step": 27695
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.350124688452219,
+      "learning_rate": 1.4447135731411044e-07,
+      "loss": 0.6229,
+      "step": 27700
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.222093872008338,
+      "learning_rate": 1.4377223378348014e-07,
+      "loss": 0.6814,
+      "step": 27705
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.804943687592125,
+      "learning_rate": 1.430747812761868e-07,
+      "loss": 0.6751,
+      "step": 27710
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.434005038612656,
+      "learning_rate": 1.4237900003222172e-07,
+      "loss": 0.6487,
+      "step": 27715
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.983682819390938,
+      "learning_rate": 1.4168489029100395e-07,
+      "loss": 0.6641,
+      "step": 27720
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.826412766166672,
+      "learning_rate": 1.4099245229137414e-07,
+      "loss": 0.6253,
+      "step": 27725
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.98851863031606,
+      "learning_rate": 1.4030168627160112e-07,
+      "loss": 0.6159,
+      "step": 27730
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.966095621552294,
+      "learning_rate": 1.3961259246937642e-07,
+      "loss": 0.6586,
+      "step": 27735
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.077498218523122,
+      "learning_rate": 1.3892517112181646e-07,
+      "loss": 0.6666,
+      "step": 27740
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.4972011043076145,
+      "learning_rate": 1.3823942246546207e-07,
+      "loss": 0.6735,
+      "step": 27745
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.6509709640658645,
+      "learning_rate": 1.3755534673627947e-07,
+      "loss": 0.6801,
+      "step": 27750
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.454353433000559,
+      "learning_rate": 1.36872944169657e-07,
+      "loss": 0.6128,
+      "step": 27755
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.563784938049496,
+      "learning_rate": 1.361922150004097e-07,
+      "loss": 0.6344,
+      "step": 27760
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.688473347214066,
+      "learning_rate": 1.355131594627762e-07,
+      "loss": 0.6378,
+      "step": 27765
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.675427454072095,
+      "learning_rate": 1.3483577779041802e-07,
+      "loss": 0.6397,
+      "step": 27770
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.560500501070449,
+      "learning_rate": 1.3416007021642252e-07,
+      "loss": 0.6303,
+      "step": 27775
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.279496533965027,
+      "learning_rate": 1.334860369732993e-07,
+      "loss": 0.6404,
+      "step": 27780
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.514573270455693,
+      "learning_rate": 1.328136782929823e-07,
+      "loss": 0.6456,
+      "step": 27785
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.712080775356427,
+      "learning_rate": 1.3214299440683032e-07,
+      "loss": 0.6441,
+      "step": 27790
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.380594002720372,
+      "learning_rate": 1.3147398554562374e-07,
+      "loss": 0.6455,
+      "step": 27795
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.474777394755265,
+      "learning_rate": 1.3080665193956954e-07,
+      "loss": 0.6088,
+      "step": 27800
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.393544527385068,
+      "learning_rate": 1.301409938182968e-07,
+      "loss": 0.6282,
+      "step": 27805
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.875236489949824,
+      "learning_rate": 1.294770114108551e-07,
+      "loss": 0.6368,
+      "step": 27810
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 6.864909001772833,
+      "learning_rate": 1.2881470494572278e-07,
+      "loss": 0.6636,
+      "step": 27815
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.021612888517245,
+      "learning_rate": 1.2815407465079754e-07,
+      "loss": 0.6444,
+      "step": 27820
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 7.051741591293166,
+      "learning_rate": 1.2749512075340198e-07,
+      "loss": 0.6464,
+      "step": 27825
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.710481817137641,
+      "learning_rate": 1.268378434802814e-07,
+      "loss": 0.6608,
+      "step": 27830
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.617873854790824,
+      "learning_rate": 1.2618224305760596e-07,
+      "loss": 0.6828,
+      "step": 27835
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.886618090636221,
+      "learning_rate": 1.2552831971096412e-07,
+      "loss": 0.6615,
+      "step": 27840
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.696371454416478,
+      "learning_rate": 1.2487607366537258e-07,
+      "loss": 0.6118,
+      "step": 27845
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.019821337329065,
+      "learning_rate": 1.2422550514526678e-07,
+      "loss": 0.6458,
+      "step": 27850
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.665430937369177,
+      "learning_rate": 1.2357661437450873e-07,
+      "loss": 0.6386,
+      "step": 27855
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.63196248195062,
+      "learning_rate": 1.229294015763799e-07,
+      "loss": 0.6203,
+      "step": 27860
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.628889587290443,
+      "learning_rate": 1.2228386697358653e-07,
+      "loss": 0.6239,
+      "step": 27865
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.365112314017564,
+      "learning_rate": 1.2164001078825482e-07,
+      "loss": 0.623,
+      "step": 27870
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.258002898744488,
+      "learning_rate": 1.2099783324193647e-07,
+      "loss": 0.6571,
+      "step": 27875
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.669424922639722,
+      "learning_rate": 1.2035733455560305e-07,
+      "loss": 0.6891,
+      "step": 27880
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.869084130597134,
+      "learning_rate": 1.1971851494965104e-07,
+      "loss": 0.6252,
+      "step": 27885
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 14.23946930823488,
+      "learning_rate": 1.1908137464389625e-07,
+      "loss": 0.651,
+      "step": 27890
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.59501403796409,
+      "learning_rate": 1.1844591385757942e-07,
+      "loss": 0.6481,
+      "step": 27895
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.9948682409578025,
+      "learning_rate": 1.178121328093601e-07,
+      "loss": 0.6465,
+      "step": 27900
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.395043259501831,
+      "learning_rate": 1.1718003171732384e-07,
+      "loss": 0.6535,
+      "step": 27905
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.313318406915275,
+      "learning_rate": 1.1654961079897442e-07,
+      "loss": 0.6322,
+      "step": 27910
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.613602971142326,
+      "learning_rate": 1.1592087027123999e-07,
+      "loss": 0.6677,
+      "step": 27915
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.073924506007503,
+      "learning_rate": 1.152938103504686e-07,
+      "loss": 0.6293,
+      "step": 27920
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.616028361075712,
+      "learning_rate": 1.1466843125243322e-07,
+      "loss": 0.638,
+      "step": 27925
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.381443254499906,
+      "learning_rate": 1.140447331923239e-07,
+      "loss": 0.6057,
+      "step": 27930
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.644498220335952,
+      "learning_rate": 1.1342271638475455e-07,
+      "loss": 0.6175,
+      "step": 27935
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.249951639838429,
+      "learning_rate": 1.1280238104376173e-07,
+      "loss": 0.6687,
+      "step": 27940
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.78029838029608,
+      "learning_rate": 1.1218372738280137e-07,
+      "loss": 0.6233,
+      "step": 27945
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.657499611806018,
+      "learning_rate": 1.115667556147526e-07,
+      "loss": 0.652,
+      "step": 27950
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.236393726741324,
+      "learning_rate": 1.1095146595191397e-07,
+      "loss": 0.6533,
+      "step": 27955
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.643644972620957,
+      "learning_rate": 1.1033785860600666e-07,
+      "loss": 0.6154,
+      "step": 27960
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.558633798199325,
+      "learning_rate": 1.0972593378817176e-07,
+      "loss": 0.6651,
+      "step": 27965
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.432431102445073,
+      "learning_rate": 1.0911569170897252e-07,
+      "loss": 0.6423,
+      "step": 27970
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.460147093023334,
+      "learning_rate": 1.0850713257839152e-07,
+      "loss": 0.6485,
+      "step": 27975
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.0266464235297015,
+      "learning_rate": 1.0790025660583514e-07,
+      "loss": 0.6477,
+      "step": 27980
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.702953698317971,
+      "learning_rate": 1.0729506400012801e-07,
+      "loss": 0.6596,
+      "step": 27985
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.1933909947174675,
+      "learning_rate": 1.0669155496951633e-07,
+      "loss": 0.5747,
+      "step": 27990
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.7820636291726055,
+      "learning_rate": 1.0608972972166675e-07,
+      "loss": 0.6255,
+      "step": 27995
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.026350823898908,
+      "learning_rate": 1.0548958846366697e-07,
+      "loss": 0.6381,
+      "step": 28000
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.826278656901001,
+      "learning_rate": 1.0489113140202456e-07,
+      "loss": 0.6491,
+      "step": 28005
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.047374508540337,
+      "learning_rate": 1.0429435874266869e-07,
+      "loss": 0.6206,
+      "step": 28010
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.188749544431611,
+      "learning_rate": 1.0369927069094788e-07,
+      "loss": 0.6763,
+      "step": 28015
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.4120748291853085,
+      "learning_rate": 1.0310586745163276e-07,
+      "loss": 0.6866,
+      "step": 28020
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.1115261193344494,
+      "learning_rate": 1.0251414922891e-07,
+      "loss": 0.6488,
+      "step": 28025
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.17494029759684,
+      "learning_rate": 1.0192411622639175e-07,
+      "loss": 0.596,
+      "step": 28030
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.61738215353897,
+      "learning_rate": 1.0133576864710671e-07,
+      "loss": 0.6573,
+      "step": 28035
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.420087106860671,
+      "learning_rate": 1.0074910669350457e-07,
+      "loss": 0.6262,
+      "step": 28040
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.711847300503816,
+      "learning_rate": 1.0016413056745555e-07,
+      "loss": 0.6349,
+      "step": 28045
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.972521433716798,
+      "learning_rate": 9.95808404702503e-08,
+      "loss": 0.6547,
+      "step": 28050
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.537633311092984,
+      "learning_rate": 9.899923660259658e-08,
+      "loss": 0.6912,
+      "step": 28055
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.051293484949038,
+      "learning_rate": 9.841931916462433e-08,
+      "loss": 0.6236,
+      "step": 28060
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.493774218309113,
+      "learning_rate": 9.784108835588335e-08,
+      "loss": 0.6791,
+      "step": 28065
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.576787830353115,
+      "learning_rate": 9.726454437534116e-08,
+      "loss": 0.6131,
+      "step": 28070
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.630404316217037,
+      "learning_rate": 9.668968742138741e-08,
+      "loss": 0.6616,
+      "step": 28075
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.891299088255928,
+      "learning_rate": 9.611651769182828e-08,
+      "loss": 0.6426,
+      "step": 28080
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.5114788096308756,
+      "learning_rate": 9.554503538389215e-08,
+      "loss": 0.6562,
+      "step": 28085
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 6.711102801133464,
+      "learning_rate": 9.497524069422448e-08,
+      "loss": 0.6548,
+      "step": 28090
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.066223190114222,
+      "learning_rate": 9.440713381889233e-08,
+      "loss": 0.6809,
+      "step": 28095
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.979158338870474,
+      "learning_rate": 9.384071495337932e-08,
+      "loss": 0.6261,
+      "step": 28100
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.64668267133173,
+      "learning_rate": 9.327598429259122e-08,
+      "loss": 0.6389,
+      "step": 28105
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 6.083600469147071,
+      "learning_rate": 9.271294203085035e-08,
+      "loss": 0.6395,
+      "step": 28110
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.465458441123022,
+      "learning_rate": 9.215158836189897e-08,
+      "loss": 0.6326,
+      "step": 28115
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 6.267730702068969,
+      "learning_rate": 9.159192347889755e-08,
+      "loss": 0.6629,
+      "step": 28120
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.391159402020457,
+      "learning_rate": 9.103394757442818e-08,
+      "loss": 0.6537,
+      "step": 28125
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.377867762700203,
+      "learning_rate": 9.047766084048837e-08,
+      "loss": 0.6667,
+      "step": 28130
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.430428637787642,
+      "learning_rate": 8.992306346849721e-08,
+      "loss": 0.6451,
+      "step": 28135
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.971622244790864,
+      "learning_rate": 8.93701556492893e-08,
+      "loss": 0.6494,
+      "step": 28140
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.812306458518884,
+      "learning_rate": 8.881893757312132e-08,
+      "loss": 0.6946,
+      "step": 28145
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.785301173240359,
+      "learning_rate": 8.826940942966544e-08,
+      "loss": 0.6429,
+      "step": 28150
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.8853191729513386,
+      "learning_rate": 8.77215714080154e-08,
+      "loss": 0.6692,
+      "step": 28155
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.361968654665912,
+      "learning_rate": 8.717542369667986e-08,
+      "loss": 0.6632,
+      "step": 28160
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.786993702451831,
+      "learning_rate": 8.66309664835896e-08,
+      "loss": 0.6265,
+      "step": 28165
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.65748116787266,
+      "learning_rate": 8.608819995609085e-08,
+      "loss": 0.6647,
+      "step": 28170
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.762331116389162,
+      "learning_rate": 8.554712430094925e-08,
+      "loss": 0.672,
+      "step": 28175
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.6782388279918115,
+      "learning_rate": 8.50077397043475e-08,
+      "loss": 0.6628,
+      "step": 28180
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.888639645100182,
+      "learning_rate": 8.447004635188882e-08,
+      "loss": 0.638,
+      "step": 28185
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.782015532131001,
+      "learning_rate": 8.393404442859243e-08,
+      "loss": 0.6418,
+      "step": 28190
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.625853443548703,
+      "learning_rate": 8.33997341188958e-08,
+      "loss": 0.6081,
+      "step": 28195
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.898338864474646,
+      "learning_rate": 8.286711560665461e-08,
+      "loss": 0.6188,
+      "step": 28200
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.4494813409120475,
+      "learning_rate": 8.233618907514285e-08,
+      "loss": 0.6067,
+      "step": 28205
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.97554379414742,
+      "learning_rate": 8.180695470705157e-08,
+      "loss": 0.6264,
+      "step": 28210
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.384806368994531,
+      "learning_rate": 8.127941268448903e-08,
+      "loss": 0.6668,
+      "step": 28215
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.543114379222686,
+      "learning_rate": 8.075356318898331e-08,
+      "loss": 0.6087,
+      "step": 28220
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.960937328784922,
+      "learning_rate": 8.022940640147803e-08,
+      "loss": 0.657,
+      "step": 28225
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.8604583820643255,
+      "learning_rate": 7.970694250233502e-08,
+      "loss": 0.6145,
+      "step": 28230
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.0096396603393805,
+      "learning_rate": 7.91861716713338e-08,
+      "loss": 0.6769,
+      "step": 28235
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.488894932987289,
+      "learning_rate": 7.866709408767104e-08,
+      "loss": 0.6487,
+      "step": 28240
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.000269938677299,
+      "learning_rate": 7.814970992995996e-08,
+      "loss": 0.6416,
+      "step": 28245
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.439658402094098,
+      "learning_rate": 7.763401937623371e-08,
+      "loss": 0.66,
+      "step": 28250
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.211970429252045,
+      "learning_rate": 7.712002260393925e-08,
+      "loss": 0.5953,
+      "step": 28255
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.0036940064815605,
+      "learning_rate": 7.660771978994397e-08,
+      "loss": 0.6751,
+      "step": 28260
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.004269407828965,
+      "learning_rate": 7.609711111052964e-08,
+      "loss": 0.6117,
+      "step": 28265
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.73802167825799,
+      "learning_rate": 7.558819674139628e-08,
+      "loss": 0.625,
+      "step": 28270
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.513531142950939,
+      "learning_rate": 7.508097685766103e-08,
+      "loss": 0.6253,
+      "step": 28275
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.782433144091685,
+      "learning_rate": 7.457545163385815e-08,
+      "loss": 0.6081,
+      "step": 28280
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.122275711517098,
+      "learning_rate": 7.407162124393741e-08,
+      "loss": 0.6561,
+      "step": 28285
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.99388262921669,
+      "learning_rate": 7.356948586126789e-08,
+      "loss": 0.6557,
+      "step": 28290
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.40740378136391,
+      "learning_rate": 7.306904565863249e-08,
+      "loss": 0.6308,
+      "step": 28295
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.894685108940628,
+      "learning_rate": 7.25703008082329e-08,
+      "loss": 0.6651,
+      "step": 28300
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.512397581048909,
+      "learning_rate": 7.207325148168631e-08,
+      "loss": 0.6553,
+      "step": 28305
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.426837513909502,
+      "learning_rate": 7.1577897850027e-08,
+      "loss": 0.6654,
+      "step": 28310
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.644005315701208,
+      "learning_rate": 7.10842400837064e-08,
+      "loss": 0.6559,
+      "step": 28315
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.608825459085924,
+      "learning_rate": 7.059227835259086e-08,
+      "loss": 0.6582,
+      "step": 28320
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.95551288211116,
+      "learning_rate": 7.010201282596385e-08,
+      "loss": 0.6505,
+      "step": 28325
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 6.002840384245793,
+      "learning_rate": 6.961344367252654e-08,
+      "loss": 0.6185,
+      "step": 28330
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 6.739543685579368,
+      "learning_rate": 6.912657106039389e-08,
+      "loss": 0.6302,
+      "step": 28335
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.368697153210344,
+      "learning_rate": 6.8641395157098e-08,
+      "loss": 0.656,
+      "step": 28340
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.085692854459775,
+      "learning_rate": 6.815791612958866e-08,
+      "loss": 0.6457,
+      "step": 28345
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.445170781602102,
+      "learning_rate": 6.767613414422946e-08,
+      "loss": 0.6764,
+      "step": 28350
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.554420320419269,
+      "learning_rate": 6.719604936680168e-08,
+      "loss": 0.6405,
+      "step": 28355
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.608501096926762,
+      "learning_rate": 6.671766196250207e-08,
+      "loss": 0.6301,
+      "step": 28360
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.375838707436465,
+      "learning_rate": 6.624097209594338e-08,
+      "loss": 0.6133,
+      "step": 28365
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.5510601164594044,
+      "learning_rate": 6.57659799311533e-08,
+      "loss": 0.648,
+      "step": 28370
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.507057003173995,
+      "learning_rate": 6.529268563157775e-08,
+      "loss": 0.6167,
+      "step": 28375
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.530000484593963,
+      "learning_rate": 6.48210893600748e-08,
+      "loss": 0.6149,
+      "step": 28380
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.561958919883844,
+      "learning_rate": 6.435119127892187e-08,
+      "loss": 0.6024,
+      "step": 28385
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.857815718909283,
+      "learning_rate": 6.388299154981014e-08,
+      "loss": 0.6474,
+      "step": 28390
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.530181675162402,
+      "learning_rate": 6.341649033384633e-08,
+      "loss": 0.6614,
+      "step": 28395
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.168044989441414,
+      "learning_rate": 6.295168779155315e-08,
+      "loss": 0.6458,
+      "step": 28400
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.6901341662647695,
+      "learning_rate": 6.248858408286872e-08,
+      "loss": 0.6399,
+      "step": 28405
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.4105278572727125,
+      "learning_rate": 6.202717936714675e-08,
+      "loss": 0.6443,
+      "step": 28410
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.419717470241641,
+      "learning_rate": 6.15674738031563e-08,
+      "loss": 0.6177,
+      "step": 28415
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.984848875763577,
+      "learning_rate": 6.110946754908087e-08,
+      "loss": 0.6572,
+      "step": 28420
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.196230562299611,
+      "learning_rate": 6.065316076252103e-08,
+      "loss": 0.6326,
+      "step": 28425
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.6588927974984395,
+      "learning_rate": 6.01985536004901e-08,
+      "loss": 0.6516,
+      "step": 28430
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.515798517057653,
+      "learning_rate": 5.974564621941958e-08,
+      "loss": 0.6212,
+      "step": 28435
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.920040257974786,
+      "learning_rate": 5.929443877515317e-08,
+      "loss": 0.6606,
+      "step": 28440
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.288818752985311,
+      "learning_rate": 5.884493142295222e-08,
+      "loss": 0.6362,
+      "step": 28445
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.6413709322516885,
+      "learning_rate": 5.839712431749078e-08,
+      "loss": 0.5756,
+      "step": 28450
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.351117395647321,
+      "learning_rate": 5.7951017612858926e-08,
+      "loss": 0.6283,
+      "step": 28455
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.561415313136726,
+      "learning_rate": 5.750661146256165e-08,
+      "loss": 0.6549,
+      "step": 28460
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.008280245016394,
+      "learning_rate": 5.7063906019518277e-08,
+      "loss": 0.6482,
+      "step": 28465
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.273730171002147,
+      "learning_rate": 5.6622901436064746e-08,
+      "loss": 0.5978,
+      "step": 28470
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.5930350497445405,
+      "learning_rate": 5.6183597863948555e-08,
+      "loss": 0.6405,
+      "step": 28475
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.114048875812452,
+      "learning_rate": 5.57459954543349e-08,
+      "loss": 0.6446,
+      "step": 28480
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.072325169261744,
+      "learning_rate": 5.531009435780166e-08,
+      "loss": 0.6248,
+      "step": 28485
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.572562477929448,
+      "learning_rate": 5.487589472434274e-08,
+      "loss": 0.6274,
+      "step": 28490
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 6.4104757417857945,
+      "learning_rate": 5.444339670336474e-08,
+      "loss": 0.6602,
+      "step": 28495
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.555281937287776,
+      "learning_rate": 5.401260044369028e-08,
+      "loss": 0.6843,
+      "step": 28500
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.332120683708192,
+      "learning_rate": 5.358350609355634e-08,
+      "loss": 0.6547,
+      "step": 28505
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.392595167719035,
+      "learning_rate": 5.31561138006137e-08,
+      "loss": 0.6257,
+      "step": 28510
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.111162902561142,
+      "learning_rate": 5.2730423711927495e-08,
+      "loss": 0.6605,
+      "step": 28515
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.485432804363226,
+      "learning_rate": 5.230643597397722e-08,
+      "loss": 0.6624,
+      "step": 28520
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.467263676798901,
+      "learning_rate": 5.1884150732656737e-08,
+      "loss": 0.6322,
+      "step": 28525
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.378733829067483,
+      "learning_rate": 5.146356813327369e-08,
+      "loss": 0.6203,
+      "step": 28530
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.93470561841416,
+      "learning_rate": 5.104468832055065e-08,
+      "loss": 0.6239,
+      "step": 28535
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.605602015042057,
+      "learning_rate": 5.062751143862399e-08,
+      "loss": 0.6922,
+      "step": 28540
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.825177520864215,
+      "learning_rate": 5.021203763104332e-08,
+      "loss": 0.6379,
+      "step": 28545
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 6.085782351837812,
+      "learning_rate": 4.979826704077262e-08,
+      "loss": 0.6138,
+      "step": 28550
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 6.327960087836945,
+      "learning_rate": 4.9386199810190214e-08,
+      "loss": 0.6185,
+      "step": 28555
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.725564634623508,
+      "learning_rate": 4.8975836081088246e-08,
+      "loss": 0.6457,
+      "step": 28560
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.797051546102331,
+      "learning_rate": 4.856717599467209e-08,
+      "loss": 0.6204,
+      "step": 28565
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.555268789277485,
+      "learning_rate": 4.816021969156148e-08,
+      "loss": 0.6455,
+      "step": 28570
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.6302031379590485,
+      "learning_rate": 4.775496731178997e-08,
+      "loss": 0.6175,
+      "step": 28575
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.398912320781518,
+      "learning_rate": 4.735141899480433e-08,
+      "loss": 0.6231,
+      "step": 28580
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.553316084687461,
+      "learning_rate": 4.694957487946517e-08,
+      "loss": 0.6552,
+      "step": 28585
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.666877163122385,
+      "learning_rate": 4.65494351040463e-08,
+      "loss": 0.6706,
+      "step": 28590
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.5545756201040595,
+      "learning_rate": 4.615099980623594e-08,
+      "loss": 0.6599,
+      "step": 28595
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.254769283834533,
+      "learning_rate": 4.575426912313441e-08,
+      "loss": 0.6596,
+      "step": 28600
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.205245943354018,
+      "learning_rate": 4.5359243191258065e-08,
+      "loss": 0.6479,
+      "step": 28605
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.379569371062987,
+      "learning_rate": 4.496592214653317e-08,
+      "loss": 0.6546,
+      "step": 28610
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.586698231517767,
+      "learning_rate": 4.457430612430202e-08,
+      "loss": 0.6606,
+      "step": 28615
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.551421074840228,
+      "learning_rate": 4.4184395259319055e-08,
+      "loss": 0.6152,
+      "step": 28620
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.2776623277684696,
+      "learning_rate": 4.379618968575305e-08,
+      "loss": 0.6255,
+      "step": 28625
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.500571494344823,
+      "learning_rate": 4.340968953718327e-08,
+      "loss": 0.6573,
+      "step": 28630
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.57162691956062,
+      "learning_rate": 4.302489494660611e-08,
+      "loss": 0.6461,
+      "step": 28635
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.088584468734806,
+      "learning_rate": 4.264180604642787e-08,
+      "loss": 0.6568,
+      "step": 28640
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.835489483073978,
+      "learning_rate": 4.226042296846977e-08,
+      "loss": 0.6373,
+      "step": 28645
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.995947787844424,
+      "learning_rate": 4.1880745843964065e-08,
+      "loss": 0.6473,
+      "step": 28650
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 6.055346263069701,
+      "learning_rate": 4.150277480355902e-08,
+      "loss": 0.6614,
+      "step": 28655
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.192967250079167,
+      "learning_rate": 4.1126509977312266e-08,
+      "loss": 0.5951,
+      "step": 28660
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.1990557125341,
+      "learning_rate": 4.075195149469802e-08,
+      "loss": 0.6582,
+      "step": 28665
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.760477686001706,
+      "learning_rate": 4.0379099484600396e-08,
+      "loss": 0.6702,
+      "step": 28670
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.6232514064036,
+      "learning_rate": 4.000795407531788e-08,
+      "loss": 0.6625,
+      "step": 28675
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.6543014868891825,
+      "learning_rate": 3.963851539456054e-08,
+      "loss": 0.6184,
+      "step": 28680
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.579939773683967,
+      "learning_rate": 3.9270783569452794e-08,
+      "loss": 0.6706,
+      "step": 28685
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.9146514121656235,
+      "learning_rate": 3.89047587265301e-08,
+      "loss": 0.6489,
+      "step": 28690
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.571272333715706,
+      "learning_rate": 3.85404409917417e-08,
+      "loss": 0.6581,
+      "step": 28695
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.444044399986175,
+      "learning_rate": 3.817783049044899e-08,
+      "loss": 0.6373,
+      "step": 28700
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.679007030962427,
+      "learning_rate": 3.781692734742548e-08,
+      "loss": 0.6619,
+      "step": 28705
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.484198835518726,
+      "learning_rate": 3.745773168685796e-08,
+      "loss": 0.654,
+      "step": 28710
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.51583489447258,
+      "learning_rate": 3.7100243632344767e-08,
+      "loss": 0.627,
+      "step": 28715
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.491214712554003,
+      "learning_rate": 3.674446330689807e-08,
+      "loss": 0.6442,
+      "step": 28720
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.871538819942215,
+      "learning_rate": 3.639039083294105e-08,
+      "loss": 0.6495,
+      "step": 28725
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.532573252897236,
+      "learning_rate": 3.603802633230957e-08,
+      "loss": 0.6503,
+      "step": 28730
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.605378611351734,
+      "learning_rate": 3.5687369926252216e-08,
+      "loss": 0.6521,
+      "step": 28735
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.854656221628359,
+      "learning_rate": 3.533842173542967e-08,
+      "loss": 0.6543,
+      "step": 28740
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.566033551316717,
+      "learning_rate": 3.499118187991368e-08,
+      "loss": 0.6348,
+      "step": 28745
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.536806510621569,
+      "learning_rate": 3.464565047919033e-08,
+      "loss": 0.6349,
+      "step": 28750
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.573169257564416,
+      "learning_rate": 3.430182765215617e-08,
+      "loss": 0.6367,
+      "step": 28755
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.124899593147868,
+      "learning_rate": 3.395971351712046e-08,
+      "loss": 0.622,
+      "step": 28760
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.411106375879562,
+      "learning_rate": 3.361930819180404e-08,
+      "loss": 0.6182,
+      "step": 28765
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.701918083291764,
+      "learning_rate": 3.328061179334041e-08,
+      "loss": 0.6617,
+      "step": 28770
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.451394066867098,
+      "learning_rate": 3.294362443827415e-08,
+      "loss": 0.6506,
+      "step": 28775
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.966870524830984,
+      "learning_rate": 3.260834624256304e-08,
+      "loss": 0.6466,
+      "step": 28780
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.952005053396158,
+      "learning_rate": 3.227477732157536e-08,
+      "loss": 0.6261,
+      "step": 28785
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.530753020996644,
+      "learning_rate": 3.1942917790092064e-08,
+      "loss": 0.6375,
+      "step": 28790
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.760861224352346,
+      "learning_rate": 3.1612767762305706e-08,
+      "loss": 0.6499,
+      "step": 28795
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.295427145204775,
+      "learning_rate": 3.1284327351820964e-08,
+      "loss": 0.6451,
+      "step": 28800
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.941272981204676,
+      "learning_rate": 3.095759667165299e-08,
+      "loss": 0.6474,
+      "step": 28805
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.0352796812221525,
+      "learning_rate": 3.0632575834230184e-08,
+      "loss": 0.6505,
+      "step": 28810
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.187736867890282,
+      "learning_rate": 3.030926495139142e-08,
+      "loss": 0.6327,
+      "step": 28815
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.441092784267658,
+      "learning_rate": 2.998766413438881e-08,
+      "loss": 0.6631,
+      "step": 28820
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.575506974996182,
+      "learning_rate": 2.9667773493883833e-08,
+      "loss": 0.6408,
+      "step": 28825
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.9177492504304015,
+      "learning_rate": 2.9349593139950116e-08,
+      "loss": 0.6361,
+      "step": 28830
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.001552291691809,
+      "learning_rate": 2.903312318207452e-08,
+      "loss": 0.6109,
+      "step": 28835
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.813831831113115,
+      "learning_rate": 2.8718363729153264e-08,
+      "loss": 0.5963,
+      "step": 28840
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.465535623922916,
+      "learning_rate": 2.8405314889494718e-08,
+      "loss": 0.6458,
+      "step": 28845
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.728693209432457,
+      "learning_rate": 2.8093976770819377e-08,
+      "loss": 0.6618,
+      "step": 28850
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.665203123879959,
+      "learning_rate": 2.7784349480257654e-08,
+      "loss": 0.6137,
+      "step": 28855
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.069120938493329,
+      "learning_rate": 2.747643312435211e-08,
+      "loss": 0.6607,
+      "step": 28860
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.394533717027899,
+      "learning_rate": 2.717022780905687e-08,
+      "loss": 0.6679,
+      "step": 28865
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.495963076629805,
+      "learning_rate": 2.686573363973599e-08,
+      "loss": 0.6757,
+      "step": 28870
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 7.649924820960543,
+      "learning_rate": 2.6562950721167324e-08,
+      "loss": 0.6546,
+      "step": 28875
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.541609873383282,
+      "learning_rate": 2.6261879157536417e-08,
+      "loss": 0.6147,
+      "step": 28880
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.56383200283721,
+      "learning_rate": 2.5962519052442627e-08,
+      "loss": 0.6423,
+      "step": 28885
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.566909873140656,
+      "learning_rate": 2.566487050889521e-08,
+      "loss": 0.6306,
+      "step": 28890
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.792659340433287,
+      "learning_rate": 2.536893362931503e-08,
+      "loss": 0.6826,
+      "step": 28895
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.966064284959946,
+      "learning_rate": 2.5074708515532843e-08,
+      "loss": 0.625,
+      "step": 28900
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.277741643462846,
+      "learning_rate": 2.4782195268792663e-08,
+      "loss": 0.662,
+      "step": 28905
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.4850580438407865,
+      "learning_rate": 2.449139398974676e-08,
+      "loss": 0.6188,
+      "step": 28910
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.677554062649509,
+      "learning_rate": 2.4202304778460083e-08,
+      "loss": 0.6213,
+      "step": 28915
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.664419136392676,
+      "learning_rate": 2.3914927734408068e-08,
+      "loss": 0.7018,
+      "step": 28920
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.547867417991105,
+      "learning_rate": 2.3629262956476605e-08,
+      "loss": 0.6345,
+      "step": 28925
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.245861658306631,
+      "learning_rate": 2.334531054296263e-08,
+      "loss": 0.6428,
+      "step": 28930
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.527071288715775,
+      "learning_rate": 2.306307059157409e-08,
+      "loss": 0.6533,
+      "step": 28935
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.01586437245096,
+      "learning_rate": 2.2782543199429407e-08,
+      "loss": 0.6258,
+      "step": 28940
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.136739091567763,
+      "learning_rate": 2.250372846305804e-08,
+      "loss": 0.6246,
+      "step": 28945
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.407009944313833,
+      "learning_rate": 2.222662647839935e-08,
+      "loss": 0.6838,
+      "step": 28950
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.477006030237568,
+      "learning_rate": 2.1951237340804287e-08,
+      "loss": 0.6309,
+      "step": 28955
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.92534364353323,
+      "learning_rate": 2.1677561145034275e-08,
+      "loss": 0.6635,
+      "step": 28960
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.0919001765583545,
+      "learning_rate": 2.1405597985260096e-08,
+      "loss": 0.656,
+      "step": 28965
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.620281001158985,
+      "learning_rate": 2.1135347955064666e-08,
+      "loss": 0.6142,
+      "step": 28970
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.52863045661348,
+      "learning_rate": 2.0866811147440825e-08,
+      "loss": 0.6213,
+      "step": 28975
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.328487665223623,
+      "learning_rate": 2.0599987654791877e-08,
+      "loss": 0.6211,
+      "step": 28980
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.585527272224497,
+      "learning_rate": 2.0334877568930488e-08,
+      "loss": 0.6704,
+      "step": 28985
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.881744284413331,
+      "learning_rate": 2.0071480981082025e-08,
+      "loss": 0.6598,
+      "step": 28990
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.006635771364114,
+      "learning_rate": 1.980979798188065e-08,
+      "loss": 0.5994,
+      "step": 28995
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.228311640387504,
+      "learning_rate": 1.9549828661371562e-08,
+      "loss": 0.6527,
+      "step": 29000
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.266788354535082,
+      "learning_rate": 1.9291573109008754e-08,
+      "loss": 0.6247,
+      "step": 29005
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.670303037452094,
+      "learning_rate": 1.903503141365892e-08,
+      "loss": 0.6328,
+      "step": 29010
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.461366194656828,
+      "learning_rate": 1.8780203663597562e-08,
+      "loss": 0.6534,
+      "step": 29015
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.666554491042867,
+      "learning_rate": 1.852708994651009e-08,
+      "loss": 0.6293,
+      "step": 29020
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.600295377045515,
+      "learning_rate": 1.82756903494935e-08,
+      "loss": 0.6302,
+      "step": 29025
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.060027863989442,
+      "learning_rate": 1.80260049590536e-08,
+      "loss": 0.6439,
+      "step": 29030
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.7525193226534554,
+      "learning_rate": 1.7778033861106657e-08,
+      "loss": 0.6362,
+      "step": 29035
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.346645034935819,
+      "learning_rate": 1.7531777140980534e-08,
+      "loss": 0.6455,
+      "step": 29040
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.530727983211217,
+      "learning_rate": 1.7287234883410775e-08,
+      "loss": 0.6448,
+      "step": 29045
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.409577499950719,
+      "learning_rate": 1.7044407172543964e-08,
+      "loss": 0.6295,
+      "step": 29050
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.738767443957085,
+      "learning_rate": 1.6803294091937707e-08,
+      "loss": 0.6209,
+      "step": 29055
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.795117614097678,
+      "learning_rate": 1.6563895724558966e-08,
+      "loss": 0.6724,
+      "step": 29060
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.948788391002489,
+      "learning_rate": 1.632621215278296e-08,
+      "loss": 0.6555,
+      "step": 29065
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.1906257152362905,
+      "learning_rate": 1.609024345839816e-08,
+      "loss": 0.6347,
+      "step": 29070
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.519593142681205,
+      "learning_rate": 1.5855989722600164e-08,
+      "loss": 0.6477,
+      "step": 29075
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 7.336439870075627,
+      "learning_rate": 1.5623451025995605e-08,
+      "loss": 0.6656,
+      "step": 29080
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.152827841994058,
+      "learning_rate": 1.5392627448601038e-08,
+      "loss": 0.6836,
+      "step": 29085
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.919461978854965,
+      "learning_rate": 1.5163519069842924e-08,
+      "loss": 0.641,
+      "step": 29090
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.28278762714753,
+      "learning_rate": 1.4936125968555982e-08,
+      "loss": 0.6402,
+      "step": 29095
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.664234099410667,
+      "learning_rate": 1.4710448222987617e-08,
+      "loss": 0.6693,
+      "step": 29100
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.5214278499396485,
+      "learning_rate": 1.448648591079238e-08,
+      "loss": 0.6272,
+      "step": 29105
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.322543422565335,
+      "learning_rate": 1.4264239109035293e-08,
+      "loss": 0.6445,
+      "step": 29110
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.001090059520327,
+      "learning_rate": 1.4043707894191294e-08,
+      "loss": 0.6431,
+      "step": 29115
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.858119709583894,
+      "learning_rate": 1.382489234214579e-08,
+      "loss": 0.6427,
+      "step": 29120
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.12636538178491,
+      "learning_rate": 1.3607792528192443e-08,
+      "loss": 0.664,
+      "step": 29125
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.381062848659886,
+      "learning_rate": 1.3392408527034828e-08,
+      "loss": 0.6266,
+      "step": 29130
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.636254099359382,
+      "learning_rate": 1.3178740412786995e-08,
+      "loss": 0.6124,
+      "step": 29135
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.89566771162359,
+      "learning_rate": 1.2966788258971797e-08,
+      "loss": 0.6728,
+      "step": 29140
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.962272202172932,
+      "learning_rate": 1.275655213852145e-08,
+      "loss": 0.6557,
+      "step": 29145
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.58680564099952,
+      "learning_rate": 1.2548032123777531e-08,
+      "loss": 0.6365,
+      "step": 29150
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.482177696174969,
+      "learning_rate": 1.23412282864932e-08,
+      "loss": 0.6259,
+      "step": 29155
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.556075353290765,
+      "learning_rate": 1.2136140697827647e-08,
+      "loss": 0.6113,
+      "step": 29160
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.9236493920375715,
+      "learning_rate": 1.1932769428352753e-08,
+      "loss": 0.6693,
+      "step": 29165
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.378265065373981,
+      "learning_rate": 1.1731114548046984e-08,
+      "loss": 0.6324,
+      "step": 29170
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.684460127715717,
+      "learning_rate": 1.1531176126300948e-08,
+      "loss": 0.6659,
+      "step": 29175
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.441044263152972,
+      "learning_rate": 1.1332954231912385e-08,
+      "loss": 0.656,
+      "step": 29180
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.058056139113183,
+      "learning_rate": 1.1136448933090072e-08,
+      "loss": 0.6624,
+      "step": 29185
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 6.786763547952368,
+      "learning_rate": 1.0941660297449919e-08,
+      "loss": 0.64,
+      "step": 29190
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.766601665819574,
+      "learning_rate": 1.0748588392019976e-08,
+      "loss": 0.655,
+      "step": 29195
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.627987229644329,
+      "learning_rate": 1.0557233283235436e-08,
+      "loss": 0.6978,
+      "step": 29200
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.066294768413905,
+      "learning_rate": 1.0367595036941402e-08,
+      "loss": 0.6804,
+      "step": 29205
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.36893000113714,
+      "learning_rate": 1.0179673718392347e-08,
+      "loss": 0.6507,
+      "step": 29210
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.440729392593793,
+      "learning_rate": 9.993469392251542e-09,
+      "loss": 0.621,
+      "step": 29215
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.157361517178861,
+      "learning_rate": 9.808982122591626e-09,
+      "loss": 0.6287,
+      "step": 29220
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.203885568995115,
+      "learning_rate": 9.62621197289515e-09,
+      "loss": 0.6518,
+      "step": 29225
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.6360921751758095,
+      "learning_rate": 9.445159006052918e-09,
+      "loss": 0.6565,
+      "step": 29230
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.43492102869121,
+      "learning_rate": 9.26582328436454e-09,
+      "loss": 0.6023,
+      "step": 29235
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.347866074081188,
+      "learning_rate": 9.088204869540096e-09,
+      "loss": 0.6087,
+      "step": 29240
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.382927803473537,
+      "learning_rate": 8.912303822697366e-09,
+      "loss": 0.6657,
+      "step": 29245
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.906534231293559,
+      "learning_rate": 8.738120204363487e-09,
+      "loss": 0.6746,
+      "step": 29250
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.626001112094817,
+      "learning_rate": 8.565654074476071e-09,
+      "loss": 0.6432,
+      "step": 29255
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.613228889261147,
+      "learning_rate": 8.394905492378758e-09,
+      "loss": 0.6669,
+      "step": 29260
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.950407492252555,
+      "learning_rate": 8.225874516827325e-09,
+      "loss": 0.6557,
+      "step": 29265
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.8617479981023735,
+      "learning_rate": 8.05856120598525e-09,
+      "loss": 0.64,
+      "step": 29270
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.431349888918077,
+      "learning_rate": 7.892965617423698e-09,
+      "loss": 0.6653,
+      "step": 29275
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.570237562146757,
+      "learning_rate": 7.729087808124868e-09,
+      "loss": 0.6714,
+      "step": 29280
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.504146337990504,
+      "learning_rate": 7.56692783447921e-09,
+      "loss": 0.6527,
+      "step": 29285
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.851906925953283,
+      "learning_rate": 7.406485752284864e-09,
+      "loss": 0.6225,
+      "step": 29290
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.8402884513015625,
+      "learning_rate": 7.247761616750449e-09,
+      "loss": 0.6469,
+      "step": 29295
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.4351542588978035,
+      "learning_rate": 7.0907554824922734e-09,
+      "loss": 0.6303,
+      "step": 29300
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.64055372052512,
+      "learning_rate": 6.935467403536567e-09,
+      "loss": 0.6229,
+      "step": 29305
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.894591370143973,
+      "learning_rate": 6.781897433317808e-09,
+      "loss": 0.7172,
+      "step": 29310
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.798009572789954,
+      "learning_rate": 6.630045624678727e-09,
+      "loss": 0.6643,
+      "step": 29315
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.366269183216473,
+      "learning_rate": 6.479912029872526e-09,
+      "loss": 0.6645,
+      "step": 29320
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.12951703426055,
+      "learning_rate": 6.331496700558437e-09,
+      "loss": 0.6074,
+      "step": 29325
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.854777420180117,
+      "learning_rate": 6.1847996878072745e-09,
+      "loss": 0.5839,
+      "step": 29330
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.656105622679166,
+      "learning_rate": 6.039821042096439e-09,
+      "loss": 0.6215,
+      "step": 29335
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.39358766372186,
+      "learning_rate": 5.896560813313801e-09,
+      "loss": 0.6232,
+      "step": 29340
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.391871972235465,
+      "learning_rate": 5.755019050754373e-09,
+      "loss": 0.6248,
+      "step": 29345
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.191262489667674,
+      "learning_rate": 5.6151958031230855e-09,
+      "loss": 0.6542,
+      "step": 29350
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.596109269185588,
+      "learning_rate": 5.477091118532562e-09,
+      "loss": 0.637,
+      "step": 29355
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.630244059530306,
+      "learning_rate": 5.34070504450479e-09,
+      "loss": 0.6157,
+      "step": 29360
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.727987744073469,
+      "learning_rate": 5.2060376279700065e-09,
+      "loss": 0.6494,
+      "step": 29365
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.508158304269665,
+      "learning_rate": 5.073088915267255e-09,
+      "loss": 0.647,
+      "step": 29370
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.8063412094916815,
+      "learning_rate": 4.941858952143275e-09,
+      "loss": 0.6464,
+      "step": 29375
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.577388459120201,
+      "learning_rate": 4.812347783755278e-09,
+      "loss": 0.6468,
+      "step": 29380
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.693548063025151,
+      "learning_rate": 4.6845554546676164e-09,
+      "loss": 0.6603,
+      "step": 29385
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.680480106578478,
+      "learning_rate": 4.558482008852894e-09,
+      "loss": 0.6685,
+      "step": 29390
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.4956655850957965,
+      "learning_rate": 4.434127489693629e-09,
+      "loss": 0.6422,
+      "step": 29395
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.334691883331565,
+      "learning_rate": 4.311491939979484e-09,
+      "loss": 0.666,
+      "step": 29400
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.838608087764427,
+      "learning_rate": 4.190575401910036e-09,
+      "loss": 0.661,
+      "step": 29405
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 7.5438973303204495,
+      "learning_rate": 4.071377917091446e-09,
+      "loss": 0.7081,
+      "step": 29410
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.435445651075055,
+      "learning_rate": 3.95389952654035e-09,
+      "loss": 0.6411,
+      "step": 29415
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.562288771628546,
+      "learning_rate": 3.838140270680524e-09,
+      "loss": 0.638,
+      "step": 29420
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 6.483748229835458,
+      "learning_rate": 3.7241001893451035e-09,
+      "loss": 0.6452,
+      "step": 29425
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.857673991541536,
+      "learning_rate": 3.611779321774367e-09,
+      "loss": 0.6119,
+      "step": 29430
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.737987373281746,
+      "learning_rate": 3.501177706618508e-09,
+      "loss": 0.6412,
+      "step": 29435
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.5811841822383785,
+      "learning_rate": 3.392295381935418e-09,
+      "loss": 0.6375,
+      "step": 29440
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.978688056473059,
+      "learning_rate": 3.2851323851906812e-09,
+      "loss": 0.6638,
+      "step": 29445
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.967654656326927,
+      "learning_rate": 3.179688753259802e-09,
+      "loss": 0.671,
+      "step": 29450
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.528084776273715,
+      "learning_rate": 3.075964522425423e-09,
+      "loss": 0.6668,
+      "step": 29455
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.7681723490814,
+      "learning_rate": 2.9739597283789946e-09,
+      "loss": 0.6578,
+      "step": 29460
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.237788531018208,
+      "learning_rate": 2.8736744062202173e-09,
+      "loss": 0.6294,
+      "step": 29465
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.369903138483945,
+      "learning_rate": 2.775108590457598e-09,
+      "loss": 0.6641,
+      "step": 29470
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.349803726300211,
+      "learning_rate": 2.6782623150067857e-09,
+      "loss": 0.6181,
+      "step": 29475
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.870894509436404,
+      "learning_rate": 2.583135613193344e-09,
+      "loss": 0.6202,
+      "step": 29480
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.561984308798503,
+      "learning_rate": 2.4897285177494234e-09,
+      "loss": 0.6406,
+      "step": 29485
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.970137297751924,
+      "learning_rate": 2.3980410608170914e-09,
+      "loss": 0.6224,
+      "step": 29490
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.930025831692994,
+      "learning_rate": 2.3080732739455545e-09,
+      "loss": 0.6585,
+      "step": 29495
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 6.05574857651503,
+      "learning_rate": 2.2198251880922726e-09,
+      "loss": 0.6353,
+      "step": 29500
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.581775862440681,
+      "learning_rate": 2.1332968336240654e-09,
+      "loss": 0.6416,
+      "step": 29505
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.370277333475543,
+      "learning_rate": 2.0484882403148943e-09,
+      "loss": 0.6805,
+      "step": 29510
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.3428214713148945,
+      "learning_rate": 1.965399437347526e-09,
+      "loss": 0.6417,
+      "step": 29515
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.8639952216571185,
+      "learning_rate": 1.8840304533118694e-09,
+      "loss": 0.6714,
+      "step": 29520
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.333033103133333,
+      "learning_rate": 1.8043813162083036e-09,
+      "loss": 0.631,
+      "step": 29525
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.919789592117598,
+      "learning_rate": 1.7264520534426844e-09,
+      "loss": 0.6757,
+      "step": 29530
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.901987244613265,
+      "learning_rate": 1.6502426918313387e-09,
+      "loss": 0.6216,
+      "step": 29535
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.979375934899607,
+      "learning_rate": 1.575753257597734e-09,
+      "loss": 0.6487,
+      "step": 29540
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.651082696096523,
+      "learning_rate": 1.502983776373035e-09,
+      "loss": 0.6217,
+      "step": 29545
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.466242260270571,
+      "learning_rate": 1.431934273197766e-09,
+      "loss": 0.6561,
+      "step": 29550
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.667264339911075,
+      "learning_rate": 1.3626047725195934e-09,
+      "loss": 0.6647,
+      "step": 29555
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.900916909243711,
+      "learning_rate": 1.2949952981949899e-09,
+      "loss": 0.6604,
+      "step": 29560
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.553663231238028,
+      "learning_rate": 1.2291058734881234e-09,
+      "loss": 0.6462,
+      "step": 29565
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 6.09318450504494,
+      "learning_rate": 1.1649365210714137e-09,
+      "loss": 0.6316,
+      "step": 29570
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.814946633549386,
+      "learning_rate": 1.1024872630255311e-09,
+      "loss": 0.6211,
+      "step": 29575
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.257440542630988,
+      "learning_rate": 1.0417581208393979e-09,
+      "loss": 0.6522,
+      "step": 29580
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.669418086862083,
+      "learning_rate": 9.827491154096315e-10,
+      "loss": 0.6647,
+      "step": 29585
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.696641863133449,
+      "learning_rate": 9.254602670416557e-10,
+      "loss": 0.6464,
+      "step": 29590
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.579217604058113,
+      "learning_rate": 8.698915954480358e-10,
+      "loss": 0.619,
+      "step": 29595
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.632195703141335,
+      "learning_rate": 8.160431197495878e-10,
+      "loss": 0.6176,
+      "step": 29600
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.773687529065707,
+      "learning_rate": 7.639148584764889e-10,
+      "loss": 0.653,
+      "step": 29605
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.402195898389152,
+      "learning_rate": 7.135068295649472e-10,
+      "loss": 0.6091,
+      "step": 29610
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.683534156836418,
+      "learning_rate": 6.648190503610874e-10,
+      "loss": 0.6599,
+      "step": 29615
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.25776544357083,
+      "learning_rate": 6.178515376181748e-10,
+      "loss": 0.6392,
+      "step": 29620
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.690548389862812,
+      "learning_rate": 5.726043074977262e-10,
+      "loss": 0.6222,
+      "step": 29625
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.384653894013326,
+      "learning_rate": 5.290773755689537e-10,
+      "loss": 0.6298,
+      "step": 29630
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.766865932934473,
+      "learning_rate": 4.872707568093216e-10,
+      "loss": 0.6399,
+      "step": 29635
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 6.264078290847827,
+      "learning_rate": 4.471844656050994e-10,
+      "loss": 0.6358,
+      "step": 29640
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.789134124948843,
+      "learning_rate": 4.088185157491431e-10,
+      "loss": 0.6618,
+      "step": 29645
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.066132528261486,
+      "learning_rate": 3.7217292044422484e-10,
+      "loss": 0.6361,
+      "step": 29650
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.259020902234293,
+      "learning_rate": 3.372476922991474e-10,
+      "loss": 0.6293,
+      "step": 29655
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.152132238576465,
+      "learning_rate": 3.040428433315201e-10,
+      "loss": 0.6497,
+      "step": 29660
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 7.046910457321705,
+      "learning_rate": 2.7255838496775824e-10,
+      "loss": 0.6266,
+      "step": 29665
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.462013754364072,
+      "learning_rate": 2.42794328041418e-10,
+      "loss": 0.6296,
+      "step": 29670
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.547167003537467,
+      "learning_rate": 2.1475068279430688e-10,
+      "loss": 0.6,
+      "step": 29675
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.4896715999147565,
+      "learning_rate": 1.884274588759283e-10,
+      "loss": 0.6708,
+      "step": 29680
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.0500035500995555,
+      "learning_rate": 1.6382466534459184e-10,
+      "loss": 0.6508,
+      "step": 29685
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.448348736657129,
+      "learning_rate": 1.4094231066574816e-10,
+      "loss": 0.6573,
+      "step": 29690
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.191738441045034,
+      "learning_rate": 1.1978040271309887e-10,
+      "loss": 0.6245,
+      "step": 29695
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.598097897991266,
+      "learning_rate": 1.0033894876859685e-10,
+      "loss": 0.6464,
+      "step": 29700
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.963954418786347,
+      "learning_rate": 8.261795552189089e-11,
+      "loss": 0.6265,
+      "step": 29705
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.68072279151534,
+      "learning_rate": 6.661742907088098e-11,
+      "loss": 0.6927,
+      "step": 29710
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.287053203230731,
+      "learning_rate": 5.23373749217182e-11,
+      "loss": 0.6326,
+      "step": 29715
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.513974310633034,
+      "learning_rate": 3.9777797987139435e-11,
+      "loss": 0.6449,
+      "step": 29720
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.508111634160351,
+      "learning_rate": 2.893870258979803e-11,
+      "loss": 0.6469,
+      "step": 29725
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.394271709957407,
+      "learning_rate": 1.982009245948824e-11,
+      "loss": 0.6508,
+      "step": 29730
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.741623800409336,
+      "learning_rate": 1.2421970733145217e-11,
+      "loss": 0.6351,
+      "step": 29735
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.517673242269964,
+      "learning_rate": 6.744339957065471e-12,
+      "loss": 0.6477,
+      "step": 29740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.693274488472275,
+      "learning_rate": 2.7872020846864135e-12,
+      "loss": 0.6142,
+      "step": 29745
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.337435074057215,
+      "learning_rate": 5.505584776965833e-13,
+      "loss": 0.6744,
+      "step": 29750
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.1189379692077637,
+      "eval_runtime": 25.4202,
+      "eval_samples_per_second": 31.707,
+      "eval_steps_per_second": 3.973,
+      "step": 29754
+    },
+    {
+      "epoch": 3.0,
+      "step": 29754,
+      "total_flos": 1024319199117312.0,
+      "train_loss": 1.0088004046184542,
+      "train_runtime": 70733.587,
+      "train_samples_per_second": 6.731,
+      "train_steps_per_second": 0.421
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 29754,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 1024319199117312.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}